International audience

International audience

The upper mid-band balances coverage and capacity for the future cellular systems and also embraces extremely large-scale multiple-input multiple-output (XL-MIMO) systems, offering enhanced spectral and energy efficiency. However, these benefits are significantly degraded under mobility due to channel aging, and further exacerbated by the unique near-field (NF) and spatial non-stationarity (SnS) propagation in such systems. To address this challenge, we propose a novel channel prediction approach that incorporates dedicated channel modeling, probabilistic representations, and Bayesian inference algorithms for this emerging scenario. Specifically, we develop tensor-structured channel models in both the spatial-frequency-temporal (SFT) and beam-delay-Doppler (BDD) domains, which capture the NF and SnS propagation effects and leverage temporal correlations among multiple snapshots for channel prediction. In this model, the factor matrices of multi-linear transformations are parameterized by BDD domain grids and SnS factors, where beam domain grids are jointly determined by angles and slopes under spatial-chirp based NF representations. To enable tractable inference, we replace these environment-dependent BDD domain grids with uniformly sampled ones, and introduce perturbation parameters in each domain to mitigate grid mismatch.We further propose a hybrid beam domain strategy that integrates angle-only sampling with slope hyperparameterization to avoid the computational burden of explicit slope sampling. On this basis, we develop tensor-structured bi-layer inference (TS-BLI) algorithm under the expectation-maximization (EM) framework, which reduces the computational complexity by leveraging the inherent separation across different domains. In the E-step, we develop the bi-layer factor graph representation to isolate the bilinear mixing in the spatial domain induced by SnS propagation, thus facilitating bi-layer iterations using approximate inference techniques. In the M-step, we leverage an alternating strategy for hyperparameter learning, with closed-form rules derived by the quadratic approximation of objective functions. Numerical simulations based on a near-practical channel simulator developed upon QuaDRiGa with SnS extensions demonstrate the superior channel prediction performance of the proposed algorithm.

Integrated Sensing and Communication (ISAC) systems combine sensing and communication functionalities within a unified framework, enhancing spectral efficiency and reducing costs by utilizing shared hardware components. This paper investigates multipath component power delay profile (MPCPDP)-based joint range and Doppler estimation for Affine Frequency Division Multiplexing (AFDM)-ISAC systems. The path resolvability of the equivalent channel in the AFDM system allows the recognition of Line-of-Sight (LoS) and Non-Line-of-Sight (NLoS) paths within a single pilot symbol in fast time-varying channels. We develop a joint estimation model that leverages multipath Doppler shifts and delays information under the AFDM waveform. Utilizing the MPCPDP, we propose a novel ranging method that exploits the range-dependent magnitude of the MPCPDP across its delay spread by constructing a Nakagami-m statistical fading model for MPC channel fading and correlating the distribution parameters with propagation distance in AFDM systems. without requiring additional ranging-specific time synchronization beyond standard receiver coarse synchronization, since the method exploits relative delays and MPCPDP statistics. We also transform the nonlinear Doppler estimation problem into a bilinear estimation problem using a First-order Taylor expansion. Moreover, we introduce the Expectation Maximization algorithm to estimate the hyperparameters and leverage the Expectation Consistent algorithm to cope with high-dimensional integration challenges. Extensive numerical simulations demonstrate the effectiveness of our MPCPDP-based joint range and Doppler estimation in ISAC systems.

This paper investigates channel estimation for linear time-varying (LTV) wireless channels underdouble sparsity, i.e., sparsity in both the delay and Doppler domains. An on-grid approximation is first considered, enabling rigorous hierarchical-sparsity modeling and compressed sensing-based channel estimation. Guaranteed recovery conditions are provided for affine frequency division multiplexing (AFDM), orthogonal frequency division multiplexing (OFDM) and single-carrier modulation (SCM), highlighting the superiority of AFDM in terms of doubly sparse channel estimation. To address arbitrary Doppler shifts, a relaxed version of the on-grid model is introduced by utilizing multiple elementary Expansion Models (BEM) each based on Discrete Prolate Spheroidal Sequences (DPSS). Next, theoretical guarantees are provided for the precision of this off-grid model before further extending it to tackle channel prediction by exploiting the inherent DPSS extrapolation capability. Finally, numerical results are provided to both validate the proposed off-grid model for channel estimation and prediction purposes under the double sparsity assumption and to compare the corresponding mean squared error (MSE) and overhead performance when different wireless waveforms are used.

In moderate- to high-mobility scenarios, channel state information (CSI) varies rapidly and becomes temporally non-stationary, leading to severe performance degradation in the massive multiple-input multiple-output (MIMO) transmissions. To address this issue, we propose a tensor-structured approach to dynamic channel prediction (TS-DCP) for massive MIMO systems with temporal non-stationarity, exploiting both dual-timescale and cross-domain correlations. Specifically, due to inherent spatial consistency, non-stationary channels over long-timescales can be approximated as stationary on short-timescales, decoupling complicated temporal correlations into more tractable dual-timescale ones. To exploit such property, we propose the sliding frame structure composed of multiple pilot orthogonal frequency-division multiplexing (OFDM) symbols, which capture short-timescale correlations within frames by Doppler domain modeling and long-timescale correlations across frames by Markov/autoregressive processes. Building on this, we develop the Tucker-based spatial-frequency-temporal domain channel model, incorporating angle-delay-Doppler (ADD) domain channels and factor matrices parameterized by ADD domain grids. Furthermore, we model cross-domain correlations of ADD domain channels within each frame, induced by clustered scattering, through the Markov random field and tensor-coupled Gaussian distribution that incorporates high-order neighborhood structures. Following these probabilistic models, we formulate the TS-DCP problem as variational free energy (VFE) minimization, and unify different inference rules through the structure design of trial beliefs. This formulation results in the dual-layer VFE optimization process and yields the online TS-DCP algorithm, where the computational complexity is reduced by exploiting tensor-structured operations. Numerical simulations demonstrate the significant superiority of the proposed algorithm over benchmarks in terms of channel prediction performance.

This paper investigates the integration of coded caching (CC) into fluid antenna (FA) multiple access (FAMA) systems to overcome their fundamental performance limitations. While FAMA has demonstrated strong interference suppression capabilities without relying on precoding, its delivery rate saturates in the high signal-to-noise ratio (SNR) regime due to residual multi-user interference, and it requires at least as many transmit antennas as served users. On the other hand, CC eliminates multi-user interference by jointly designing caching and delivery phases, but suffers from the well-known worst-user bottleneck, especially under wireless fading channels and low SNR conditions. To overcome the interference-limited nature of FAMA and the worst-user bottleneck in CC, we consider a CC-enabled FAMA framework. The proposed approach enables interference-free transmission to multiple users using a single transmit antenna, and leverages adaptive port selection at the users to combat channel fading. We analyze the average rates and effective gains of XOR-based CC and the recently developed aggregated CC (ACC) schemes under the FAMA framework, and derive simple closed-form approximations that accurately characterize the performance. Furthermore, we rigorously prove that, in the limit where the number of FA ports grows without bound while maintaining sufficient spatial diversity, the effective multiplexing gains in the low-SNR limit of both XCC- and ACC-enabled FAMA asymptotically achieve the nominal gain that is only attainable under high-SNR conditions with traditional antennas. Simulation results show that the proposed architecture outperforms both conventional FAMA and traditional-antenna CC schemes, achieving significant spectral efficiency gains in a wide range of SNR regimes.

Grid-based methods in sparse signal reconstruction (SSR) are well-regarded for their efficacy in direction-of-arrival (DoA) estimation. This paper presents the EP (Expectation Propagation)-SURE (Stein's Unbiased Risk Estimate)-SBL (Sparse Bayesian Learning) algorithm, designed for single snapshot DoA estimation. The algorithm divides DoA estimation into two parts: grid-on estimation and off-grid error estimation, employing first-order and second-order Taylor expansions. In grid-on estimation, sparse Bayesian learning is employed for sparse modeling. To tackle hyperparameter estimation challenges within sparse Bayesian learning, the algorithm adopts SURE estimator instead of the commonly-used expectation-maximization (EM) approach. For off-grid error estimation, the algorithm utilizes the EP technique to handle high-dimensional, non-tractable integration in posterior mean calculations. The feasibility and effectiveness of the proposed algorithm are validated through extensive simulations.

In this work, we study uplink communication in cell-free (CF) massive multiple-input multiple-output (MaMIMO) systems, a promising architecture for next-generation networks. To address the challenge of pilot contamination, we employ semi-blind transmission structures that enable joint channel and data symbol estimation. However, Bayesian estimation in such semi-blind frameworks leads to intractable bilinear problems. To tackle this, we propose a simplified, distributed method based on Expectation Propagation (EP) for efficient semi-blind channel estimation. Notably, we identify that if the data constellation set can be decomposed into multiple sub-constellation sets with identical amplitudes, this structure can be leveraged to significantly reduce computational complexity. This approach is particularly advantageous for managing large constellation sizes, ensuring scalability and efficiency in practical systems. Additionally, approximations based on the Central Limit Theorem are incorporated to further simplify computations.

When derived from the Bethe Free Energy (BFE) of the Generalized Linear Model (GLM), Approximate Message Passing (AMP) algorithms combine two asymptotic Large System Limit (LSL) simplifications which are asymptotic Gaussianity of extrinsics and large random matrix theory based asymptotic variance computations. In the provably convergent AMBGAMP algorithm, a LSL version of the BFE is derived. In Expectation Propagation (EP) style minimization, the LSL BFE cost function is augmented with Lagrangian terms for mean and variance consistency constraints, augmented with a quadratic version of the mean constraints as in the Method of Multipliers (MM). The mean Lagrange multipliers then get updated ADMM-style (Alternating Direction of MM). In this approach, the weights of the MM terms need to be carefully chosen, which is not part of the MM philosophy, and the Lagrange multipliers have no particular meaning. On the other hand, AMP can be derived by directly introducing LSL simplifications in the Belief Propagtion (BP) algorithm that minimizes the original GLM BFE. This allows to relate extrinsic messages to posterior pdfs by first-order Taylor series expansion based perturbations. We also apply LSL approximations to the variances of the various Gaussians involved, which in fact leads to a rederivation of a fundamental LSL theorem describing the deterministic limit of posterior variances. We show that this LSL version of BP leads to BFE modifications that correspond to the augmented Lagrangian of the LSL BFE, explaining its weights and Lagrange Multipliers. These insights should facilitate the extension of AMP to more complex settings such as bilinear models.

We propose a new class of coded caching schemes for wireless networks, which we term as non-orthogonal multiple access (NOMA) aided aggregated coded caching (NACC), which manages to alleviate both the uneven channel bottleneck as well as the shared-cache limitation. In particular, NACC uses an aggregation principle to efficiently serve users with unequal channel strengths, as well as uses for the first time the NOMA principle to simultaneously transmit to multiple users that share the same exact cache state. This transmission strategy represents an efficient utilization of NOMA within the coded caching framework. We analyze the high-SNR transmission performance of the proposed scheme and derive analytical expressions for the achievable rate and the effective gain, providing a qualitative understanding of its spectral efficiency. Our numerical results for urban Micro-cell environments abiding by$5\mathrm{G}$standards, demonstrate that NACC achieves - with a single transmit antenna - the same spectral efficiency as a multi-user (MU) multicasting system with 32 transmit antennas, while it also matches the spectral efficiency of traditional MU unicasting system with 8 transmit antennas.

In this paper, the weighted fractional Fourier transform (WFRFT) signal domain is introduced to enhance the security performance of two-way trusted relay systems at the signal level. The proposed scheme, which requires only a single relay node, leverages the multi-component energy distribution characteristics of WFRFT signals to improve security with low complexity and high power efficiency. The inherent security mechanism of WFRFT analyzed in this paper can be simply summarized as follows: the superposition of components in WFRFT signals that do not satisfy specific constraints will result in the inability to perfectly reconstruct the message signal. Based on this, confidential information is encoded into WFRFT signals with private transform orders, allowing legitimate users to achieve perfect decoding. Since WFRFT signals exhibit energy concentration only in specific transformation domains, mismatched transform orders adopted by the eavesdropper cause energy loss in the information-bearing signal, leading to inter-component interference that further degrades the quality of the recovered signal. The advantages of the proposed scheme in limiting information leakage and improving the achievable secrecy sum rate (SSR) are analyzed. Numerical results validate the theoretical analysis and demonstrate the secrecy performance of the proposed scheme.

Affine Frequency Division Multiplexing (AFDM) is a chirp-transform modulation technique that has shown reliable performance in high-mobility scenarios, making it an attractive option for next generation communication systems. Recent literature suggests that under chirp parameter adjustment, AFDM can achieve optimal diversity performance in delay-doppler channels with maximum likelihood (ML) detection. However, the performance of AFDM with minimum mean square error equalization (MMSE-Eq) has not been extensively investigated in the existing literature. In this paper, we analyze the performance of AFDM with MMSE-Eq, derive a lower bound for the theoretical bit error rate (BER) of the AFDM system, and discuss the relationship between chirp parameters and performance degradation. To optimize BER performance, we propose two distinct chirp parameter selection strategies for frequency selective and doubly selective channels, respectively. These strategies offer the advantage of avoiding extensive computations. Additionally, we propose a low-complexity and high-performance iterative MMSE-Eq algorithm based on time-domain channel matrix operations. The algorithm resolves the issue encountered in existing low-complexity methods, where different chirp parameter selections significantly impact the complexity. Simulation results demonstrate the efficacy of our proposed parameter selection strategies and the outstanding BER performance achieved by the iterative MMSE-Eq algorithm.

In the context of the expectation-maximization (EM) algorithm, which often faces challenges due to intractable posterior distributions, this study explores an innovative approach by integrating the EM algorithm with expectation consistent (EC) approximate inference. Our method involves the incorporation of the EC algorithm into the M-step of the EM algorithm, resulting in the EM-EC algorithm. We demonstrate that the fixed points of the proposed EM-EC algorithm correspond to stationary points of a specific constrained auxiliary function, thereby providing a variational interpretation of the algorithm. Through simulations, we showcase the effectiveness and robustness of this novel approach, highlighting its potential for advancing the field of Bayesian network estimation.

Many signal processing problems involve a Generalized Linear Model (GLM), which is a type of linear model where the unknowns may be non-identically independently distributed (n.i.i.d.). Vector Approximate Message Passing for Generalized Linear Models (GVAMP) is a computationally efficient belief propagation technique used for Bayesian inference. However, the posterior variances obtained from GVAMP with limited complexity are only exact under the assumption of an independent and identically distributed (i.i.d.) prior, owing to the averaging operations involved. In numerous problems, it is beneficial not just to estimate the unknowns but also to obtain accurate posterior distributions. While VAMP, and especially AMP, are applicable to high-dimensional problems, many applications involve dimensions that are not excessively high, allowing for more complex operations. Furthermore, in finite dimensions, the asymptotic regime that leads to correct variances under certain measurement matrix model assumptions is not applicable. To overcome these challenges, we propose a revised version of GVAMP, named reGVAMP. This method provides a multivariate Gaussian posterior approximation, which includes inter-parameter correlations, and yields accurate posterior marginals requiring only the extrinsic distributions to become Gaussian.

This paper explores uplink communication in cell-free (CF) massive multiple-input multiple-output (MaMIMO) systems, employing semi-blind transmission structures to mitigate pilot contamination. We propose a simplified, decentralized method based on Expectation Propagation (EP) for semi-blind estimation challenges. By utilizing orthogonal pilots, we pre-process the received signals to establish a simplified equivalent factorization scheme for the transmission process. Moreover, this study integrates Central Limit Theory (CLT) with EP, eliminating the need for new auxiliary variables in the factorization scheme. We also refine the algorithm by assessing the variable scales involved. A decentralized approach is proposed to significantly reduce the computational demands on the Central Processing Unit (CPU).

We introduce an energy-efficient downlink rate splitting multiple access (RSMA) scheme, employing a simple matched filter (MF) for precoding. We consider a transmitter equipped with multiple antennas, serving several single-antenna users at the same frequency-time resource, each with distinct message requests. Within the conventional 1-layer RSMA framework, requested messages undergo splitting into common and private streams, which are then precoded separately before transmission. In contrast, we propose a novel strategy where only an MF is employed to precode both the common and private streams in RSMA, promising significantly improved energy efficiency and reduced complexity. We demonstrate that this MF-precoded RSMA achieves the same delivery performance as conventional RSMA, where the common stream is beamformed using maximal ratio transmission (MRT) and the private streams are precoded by MF. Taking into account imperfect channel state information at the transmitter, we proceed to analyze the delivery performance of the MF-precoded RSMA. We derive the ergodic rates for decoding the common and private streams at a target user respectively in the massive MIMO regime. Finally, numerical simulations validate the accuracy of our analytical models, as well as demonstrate the advantages over conventional RSMA.

Sparse signal reconstruction (SSR) involves tackling large underdetermined systems of linear equations while incorporating constraints or regularizers. Expectation propagation (EP) emerges as a robust method for SSR, converting these constraints into prior information. However, the cubic complexity of matrix inversion per EP cycle hinders its implementation in large systems without approximation. In various applications like direction of arrival estimation (DoA), radar imaging etc., the signal to be recovered exhibits sparsity in the Fourier dictionary. To address this, we present a fast EP algorithm based on the Gohberg-Semencul (G-S) formula and Levinson-Durbin (L-D) type algorithm, boasting only quadratic complexity. Notably, no approximation operations or random measurement matrices are required for matrix inversion compared to approximate message passing (AMP) and other message passing based algorithms. Furthermore, it is compatible with non-identically and independently distributed (n.i.i.d.) priors. Numerical simulations conclusively demonstrate the efficacy of fast EP.

In this paper, we prove the convergence of the simplified information geometry approach (SIGA), which was proposed for massive MIMO-OFDM channel estimation. For a general Bayesian inference problem, we first show that the iteration of the common second-order natural parameter (SONP) is separated from that of the common first-order natural parameter (FONP). Hence, the convergence of the common SONP can be checked independently. We show that with the initialization satisfying a specific but large range, the common SONP is convergent regardless of the value of the damping factor. For the common FONP, we establish a sufficient condition of its convergence and prove that the convergence of the common FONP relies on the spectral radius of a particular matrix related to the damping factor. We give the range of the damping factor that guarantees the convergence in the worst case. Further, we determine the range of the damping factor for massive MIMO-OFDM channel estimation by using the specific properties of the measurement matrices. Simulation results are provided to confirm the theoretical results.

This paper addresses channel estimation for linear time-varying (LTV) wireless propagation links under the assumption of double sparsity i.e., sparsity in both the delay and the Doppler domains. Affine frequency division multiplexing (AFDM), a recently proposed waveform, is shown to be optimal (in terms of pilot overhead) for this problem. With both mathematical analysis and numerical results, the minimal pilot and guard overhead needed for achieving a target mean squared error (MSE) while performing channel estimation is shown to be the smallest when AFDM is employed instead of both conventional and recently proposed waveforms.

Generalized Approximate Message Passing (GAMP) allows for Bayesian inference in linear models with non-identically independently distributed (n.i.i.d.) priors and n.i.i.d. measurements of the linear mixture outputs. It represents an efficient technique for approximate inference, which becomes accurate when both rows and columns of the measurement matrix can be treated as sets of independent vectors and both dimensions become large. It has been shown that the fixed points of GAMP correspond to the extrema of a large system limit of the Bethe Free Energy (LSL-BFE), which represents a meaningful approximation optimization criterion regardless of whether the measurement matrix exhibits the independence properties. However, the convergence of (G)AMP can be notoriously problematic for certain measurement matrices, and the only sure fixes so far are damping (by a difficult-to-determine amount) or perform a double ADMM. In this paper, we revisit the GAMP algorithm (as e.g. for sparse Bayesian learning (SBL)) by more rigorously applying an alternating constrained minimization strategy to an appropriately reparameterized LSL BFE. This guarantees convergence, at least to a local optimum. We furthermore introduce a natural extension of the BFE to integrate the estimation of (the SBL) hyperparameters via Variational Bayes, leading to Variational AMBGAMP or VAM-BGAMP.

In this paper we consider cell-free (CF) massive MIMO (MaMIMO) systems, which comprise a very large number of geographically distributed access points (APs) serving a much smaller number of users. We exploit channel sparsity to tackle pilot contamination, which originates from the reuse of pilot sequences. Specifically, we consider semi-blind methods for channel estimation in the presence of unknown Gaussian i.i.d. data to resolve the pilot contamination. This task is further aided by exploiting prior channel information in a Bayesian formulation. We develop Bayesian Maximum a Posteriori (MAP) channel estimators and we also provide various Cramer-Rao Bounds to characterize performance limits. The main contribution is the derivation of an original type of Bayesian CRB for the semi-blind problem at hand, in which a certain expectation operation is facilitated by the asymptotics of the large system dimensions considered here. Whereas Bayesian CRBs lead to fairly useless lose bounds, corresponding to unrealistic genie-aided scenarios, the proposed variation turns out to be quite tight as illustrated by performance comparisons with various estimation algorithms.

Generalized Approximate Message Passing (GAMP) allows for Bayesian inference in linear models with non-identically independently distributed (n.i.i.d.) priors and n.i.i.d. measurements of the linear mixture outputs. It represents an efficient technique for approximate inference, which becomes accurate when both rows and columns of the measurement matrix can be treated as sets of independent vectors and both dimensions become large. It has been shown that the fixed points of GAMP correspond to the extrema of a large system limit of the Bethe Free Energy (LSL-BFE), which represents a meaningful approximation optimization criterion regardless of whether the measurement matrix exhibits the independence properties. However, the convergence of (G)AMP can be problematic for certain measurement matrices. In this paper, we revisit the GAMP algorithm by applying a simplified version of the Alternating Direction Method of Multipliers (ADMM) to minimizing the LSL-BFE. We show convergence of the mean and variance subsystems in AMBGAMP and in the Gaussian case, convergence of mean and LSL variance to the Minimum Mean Squared Error (MMSE) quantities.

The Channel State Information (CSI) of the orthogonal frequency division multiplexing (OFDM) comprises data pertaining to the attenuation of multipath propagation. In this paper, we assume that the amplitude fading of both line-of-sight (LoS) and non-line-of-sight (NLoS) paths conforms to the Nakagami-m distribution. Via establishing a relationship between the distribution parameters and the propagation distance, we propose a CSI-based ranging method utilizing the Expectation Maximization (EM)-Revisited Approximate Message Passing (reVAMP) algorithm. This algorithm is not only applicable to the CSI-based ranging estimation but can also be extended to other parameter estimation scenarios. It effectively tackles challenges associated with generalized linear models (GLMs) that involve hidden random variables and the intractability of posterior distributions during the EM iterations.

In this paper, we propose to remove the analog stage of hybrid beamforming (HYBF) in the millimeter wave (mmWave) full-duplex (FD) systems. Such a solution is highly desirable as the analog stage suffers from high insertion loss and high power consumption. Consequently, the mmWave FD nodes can operate with a fewer number of antennas, instead of relying on a massive number of antennas, and to tackle the propagation challenges of the mmWave band we propose to use near-field intelligent reflecting surfaces (NF-IRSs). The objective of the NF-IRSs is to simultaneously and smartly control the uplink (UL) and downlink (DL) channels while assisting in shaping the SI channel: this to obtain very strong passive SI cancellation. A novel joint active and passive beamforming design for the weighted sum-rate (WSR) maximization for the NF-IRSs-assisted mmWave point-to-point FD system is presented. Results show that the proposed solution fully reaps the benefits of the IRSs, only when they operate in the NF, which leads to considerably higher gains compared to the conventional massive MIMO (mMIMO) mmWave FD and half duplex (HD) systems.

In this paper, we investigate the channel estimation for massive multiple-input multiple-output orthogonal frequency division multiplexing (MIMO-OFDM) systems. We revisit the information geometry approach (IGA) for massive MIMO-OFDM channel estimation. By using the constant magnitude property of the entries of the measurement matrix and the asymptotic analysis, we find that the second-order natural parameters (SONPs) of the distributions on all the auxiliary manifolds (AMs) are equivalent to each other at each iteration of IGA, and the first-order natural parameters (FONPs) of the distributions on all the AMs are asymptotically equivalent to each other at the fixed point. Motivated by these results, we simplify the iterative process of IGA and propose a simplified IGA for massive MIMO-OFDM channel estimation. It is proved that at the fixed point, the a posteriori mean obtained by the simplified IGA is asymptotically optimal. The simplified IGA allows efficient implementation with fast Fourier transformation (FFT). Simulations confirm that the simplified IGA can achieve near the optimal performance with low complexity in a limited number of iterations.

In Dynamic Time Division Duplex (DynTDD), downlink/uplink (DL/UL) slot allocation is adaptive with traffic load. DynTDD systems have received a lot of attention for 5th generation (5G) mobile communication systems, as the spectrum efficiency of wireless communication networks is improved by the flexible and dynamic duplex operation. However when using DynTDD, a different DL/UL slot configuration is likely to be selected by neighboring cells, leading to Cross Link Interference (CLI) between the Base Stations (BS), which is known as BS-to-BS or DL-to-UL interference, and between User Equipment (UE) which is known as UE-to-UE or UL-to-DL interference. Rank deficient channels are frequently encountered in Multi-Input Multi-Output (MIMO) networks, due to poor scattering and keyhole effects, or when using Massive MIMO and moving to mmWave. While the implications of rank deficient channels are well understood for the single user (SU) point to point setting, less is known for interference networks. In this paper, we extend a MIMO Interference Alignment (IA) feasibility investigation framework to rank deficient channels and we investigate the IA feasibility in DynTDD considering rank deficient MIMO interfering channels, by establishing the simultaneously necessary and sufficient conditions. These conditions allow us to evaluate the Degrees-of-Freedom (DoF) of centralized designs more precisely compared to the loose proper conditions. We then compare the achievable DoF of centralized designs with those of various distributed designs, allowing us to get a better idea of the DoF price to pay for distributedness (but not accounting for the gain in information exchange reduction that distributed designs permit).

Full Duplex (FD) communication can revolutionize wireless communications as it avoids using independent channels for bi-directional communications. This work generalizes the point-to-point FD communication in millimeter wave (mmWave) band consisting of K-pairs of massive MIMO FD nodes operating simultaneously. We present a novel joint hybrid beamforming (HYBF) and combining scheme for weighted sum-rate (WSR) maximization to enable the coexistence of massive MIMO FD links cost-efficiently. The proposed algorithm relies on alternative optimization based on the minorization-maximization method. Moreover, we present a novel SI and massive MIMO interference channel aware power allocation scheme to include the optimal power control. Simulation results show significant performance improvement compared to a traditional bidirectional fully digital half-duplex (HD) system.

This paper considers a bidirectional (BD) full-duplex (BD-FD) communication system design under the joint sum-power and per-antenna power constraints. The sum-power constraints are naturally imposed by regulation to limit the total transmit power, and the per-antenna power constraints consider the physical limits of the power amplifiers (PAs). We propose a novel beamforming design to maximize the weighted sum-rate (WSR) with alternating optimization under the limited dynamic range (LDR) noise model. At each iteration, we use minorization-maximization approach to optimize the beamformers and power allocation. Simulation results show significant performance gain compared to a half-duplex BD system or FD system with only sum-power constraints. However, the gains are limited by the maximum of the thermal noise variance or the LDR noise variance.

In this paper we consider cell-free (CF) massive MIMO (MaMIMO) systems, which comprise a very large number of geographically distributed access points (APs) serving a much smaller number of users. We exploit channel sparsity to tackle pilot contamination, which originates from the reuse of pilot sequences. Specifically, we consider semi-blind methods for joint channel estimation and data detection. Under the challenging assumption of deterministic parameters, we determine sufficient conditions and necessary conditions for semi-blind identifiability, which guarantee the non-singularity of the Fisher Information Matrix (FIM) and the existence of the Cramer-Rao bound (CRB). We propose a message passing (MP) algorithm which determines the exact channel coefficients in the case of semiblind identifiability. We show that the system is identifiable if the Karp-Sipser algorithm yields an empty core. Additionally, we propose a Bayesian semi-blind approach which results in an effective algorithm for joint channel estimation and multi-user detection. This algorithm alternates between channel estimation and linear multi-user detection. Numerical simulations verify the analytical derivations.

In this paper, we consider the problem of user rate balancing in the downlink of multi-cell multi-user (MU) Multiple-Input-Multiple-Output (MIMO) systems with partial Channel State Information at the Transmitter (CSIT). With MIMO leading to multiple streams per user, user rate balancing involves both aspects of balancing and sum rate optimization. We linearize the problem by introducing a rate minorizer and by formulating the balancing operation as constraints leading to a Lagrangian, allowing to transform rate balancing into weighted sum minimization with Perron Frobenius theory. We provide original analytical expressions for the Lagrange multipliers for the multiple power constraints which can also handle the case in which some power constraints are satisfied with inequality, as can arise in a multi-cell scenario. We introduce two partial CSIT formulations. One is based on the ergodic rate Mean Squared Error (EMSE) relation, the other involves an original rate minorizer in terms of the received interference plus noise covariance matrix, in the partial CSIT case applied to the Expected Signal and Interference Power (ESIP) rate. The simulation results exhibit the improved performance of the proposed techniques over naive partial CSIT beamforming based on perfect CSIT algorithms, and in particular illustrate the close to optimal performance of the ESIP approach.

In-band Full-Duplex (FD) is a promising wireless transmission technology allowing to increase data rates by up to a factor of two, via simultaneous transmission and reception, but with a potential to increase system throughput even much more in cognitive radio and random access systems thanks to simultaneous transmission and sensing. In this work, we consider a practical hybrid beamforming design for a bidirectional massive MIMO FD system under the joint per-antenna and sum-power constraints. Moreover, we consider non-ideal circuitry in the transmit and receive chains, which is modelled with the limited dynamic range (LDR) noise model. The per-antenna power constraints take into account the actual physical limits of the power amplifiers and the sum-power constraints are imposed to limit the total transmit power. The precoders are optimized with alternating optimization by using the minorization-maximization approach. Simulation results show significant performance improvement compared to a traditional bidirectional half-duplex system.

Cell-free MIMO, employing distributed antenna systems (DAS), is a promising approach to deal with the capacity crunch of next generation wireless communications. In this paper, we consider a wireless network with transmit and receive antennas distributed according to homogeneous point processes. The received signals are jointly processed at a central processing unit. We study if the favorable propagation properties, which enable almost optimal low complexity detection via matched filtering in massive MIMO systems, hold for DAS with line of sight (LoS) channels and general attenuation exponent. Making use of Euclidean random matrices (ERM) and their moments, we show that the analytical conditions for favorable propagation are not satisfied. Hence, we propose multistage detectors, of which the matched filter represents the initial stage. We show that polynomial expansion detectors and multistage Wiener filters coincide in DAS and substantially outperform matched filtering. Simulation results are presented which validate the analytical results.

In-band Full Duplex (FD) is a wireless communication technology which has the potential to transmit and receive simultaneously in the same frequency band. Self-interference cancellation (SIC) is the key enabler to achieve FD operation. As the SI is severe, some SIC is required at the antenna level and in the analog domain before analog to digital converter (ADC) in the receiver chain because the ADC has only a limited dynamic range. Here we consider deliberately scaling up the received signal, provoking ADC saturation due to the SI signal. This leads to missing samples which we propose to reconstruct under the assumptions that the receive signal of interest is a low pass bandlimited signal with known spectrum (mask), oversampling and (perfect) digital SIC after the ADC. The missing samples are estimated by fixed lag Kalman smoothing. More upscaling leads to fewer available samples but with less quantization noise. Hence an optimum compromise arises. We provide an approximate resulting MSE analysis based on large random matrix theory, replacing randomly selected Fourier transform vectors by vectors of i.i.d. variables. Simulation results show the improvement in reconstruction Signal to Noise Ratio (RSNR) and the optimal compromise behavior.

In many applications such as massive multi-input multi-output (MIMO) radar, massive MIMO channel estimation, speech processing, image and video processing, the received signals are tensors. In such applications, utilizing techniques from tensor algebra can be beneficial since it retains the tensorial structure in the received signal compared to processing on the matricized version of the same signal. Furthermore, the underlying parameters or states to be estimated are sparse in many of the above-said applications compared to the large system dimensions. In this paper, we propose techniques which allow handling the extension of sparse Bayesian learning (SBL) to time-varying states. Adding the parameters of the autoregressive process which is used to the model the time-varyings of the state leads to a non-linear (at least bilinear) state-space model. Belief propagation (BP) is a promising method to compute the minimum mean squared error (MMSE) or maximum a posteriori (MAP) estimates, but at the the expense of a high computational burden. However, inspired by a previous work on a combined BP and variational Bayes (VB) technique, we noted that using a combination of BP, VB, and expectation propagation (EP) can help to alleviate the computational complexity.

We consider a cell-free MIMO system in uplink, comprising a massive number of distributed transmit and receive antennas. In our distributed antenna system (DAS), transmit and receive antennas are distributed according to homogeneous point processes (PP) and the received signals are processed jointly at a central processing unit (CPU). In centralized massive MIMO systems, the phenomenon of favorable propagation has been observed: when the number of receive antennas tends to infinity while the number of transmit antennas remains finite, the users' channels become almost orthogonal and low complexity detection via matched filtering is almost optimal. We analyze the properties of DASs in asymptotic conditions when the network dimensions go to infinity with given intensities of the transmit and receive antenna PPs. We study the analytical conditions of favorable propagation in DASs with two kinds of channels, namely, channels with path loss and transmit and receive antennas in line of sight (LoS) or in multipath Rayleigh fading. We show that the analytical conditions of favorable propagation are satisfied for channels impaired by path loss and Rayleigh fading while they do not hold in the case of LoS channels, motivating the use and analysis of multi-stage receivers. Simulation results of the favorable propagation conditions and the performance of multi-stage detectors for finite systems validate the asymptotic analytical results.

In this paper, we investigate the problem of user rate balancing for the downlink transmission of multiuser multicell Multiple-Input-Multiple-Output (MIMO) systems with per cell power constraints. Due to the multiple streams per user, user rate balancing involves both aspects of balancing and sum rate optimization. We exploit the rate Mean Squared Error (MSE) relation, formulate the balancing operation as constraints leading to Lagrangians in optimization duality, allowing to transform rate balancing into weighted MSE minimization with Perron Frobenius theory. The Lagrange multipliers for the multiple power constraints can be formulated as a single weighted power constraint in which the weighting can be optimized to lead to the satisfaction with equality of all power constraints. Actually, various problem formulations are possible, including single cell full power transmission leading to a dual norm optimization problem, and per cell rate balancing which breaks the balancing constraint between cells. Simulation results are provided to validate the proposed algorithms and demonstrate their performance improvement over e.g. unweighted MSE balancing.

This paper considers a bidirectional full-duplex Multi-Input Multi-Output (MIMO) OFDM system. The limited dynamic range (LDR) noise model takes into account the hardware impairments in the radio frequency (RF) chain and is thus more practical. Hence we propose a beamforming (BF) design which takes into account the LDR noise characteristics and also is robust to imperfections in the estimated channel. At the transmit side, we introduce a two stage beamformer (BF) with an inner BF of lower dimension and an outer BF of higher dimension, both BFs being at the digital (baseband) side. The inner BF in OFDM domain handles directive transmission, while the outer BF in time domain handles self interference (SI). At the receive side, we propose a hybrid combiner which involves an analog phase shifter based BF, with fewer RF chains compared to the number of receive antennas and a digital (baseband) BF in OFDM domain. The analog BF helps reduce SI before analog-to-digital conversion (ADC). All the BFs are optimized using maximization of the expected weighted sum rate (WSR) which is solved using an alternating minorization approach. The proposed multi-stage BF architecture has multiple advantages including SI reduction during OFDM cyclic prefixes, with uplink (UL) or downlink (DL) possibly using different numerology or being asynchronous, allowing proper ADC operation.

In this paper, we address the fundamental problem of Sparse Bayesian Learning (SBL), where the received signal is a high-order tensor. We furthermore consider the problem of dictionary learning (DL), where the tensor observations are assumed to be generated from a Kronecker structured (KS) dictionary matrix multiplied by the sparse coefficients. Exploiting the tensorial structure results in a reduction in the number of degrees of freedom in the learning problem, since the dimensions of each of the factor matrices are significantly smaller than the matricized dictionary if we vectorize the observations. We propose a novel fast algorithm called space alternating variational estimation with dictionary learning (SAVED-KS), which is a version of variational Bayes (VB)-SBL pushed to the scalar level. Similarly, as for SAGE (space-alternating generalized expectation maximization) compared to EM, the component-wise approach of SAVED-KS compared to SBL renders it less likely to get stuck in bad local optima and its inherent damping (more cautious progression) also leads to typically faster convergence of the non-convex optimization process. Simulation results show that the proposed algorithm has a faster convergence rate and lower mean squared error (MSE) compared to the alternating least squares (ALS) based method for tensor decomposition.

We consider coordinated beamforming (BF) for the Multi-Input Single-Output (MISO) Interfering Broadcast Channel (IBC). The beamformers are optimized for the Ergodic Weighted Sum Rate (EWSR) or various approximations and bounds thereof, for the case of Partial Channel State Information at the Transmitters (CSIT). Gaussian (posterior) partial CSIT can optimally combine channel estimate and channel covariance information. With Gaussian partial CSIT, the beamformers only depend on the means (estimates) and (error) covariances of the channels. We extend a recently introduced large system analysis for optimized beamformers with partial CSIT, by a stochastic geometry inspired randomization of the channel covariance eigen spaces, leading to much simpler analytical results, which depend only on some essential channel characteristics. In the Massive MISO (MaMISO) limit, we obtain deterministic approximations of the signal and interference plus noise powers at the receivers for various BFs, which are tight as the number of BS antennas and the total user subspace dimension tend to infinity at fixed ratio. Simulation results exhibit the correctness of the large system results and the performance superiority of optimal BF designs based on both the MaMISO limit of the EWSR and using Linear Minimum Mean Squared Error (LMMSE) channel estimates.

We present a novel and efficient approach for estimating the maximum likelihood (ML) estimates of the angles-of-arrival (AoAs) of multiple sources. The approach is iterative and is based on orthogonal projections in order to optimise the ML cost function, thus the name OPML. As will be shown, the advantage of using an orthogonal basis of the signal manifold would allow solving the ML cost function in an iterative manner. In fact, we propose two algorithms based on OPML, i.e. OPML-1 and OPML-2, which exhibit lower computational complexity and faster convergence than existing ML algorithms. In this paper, we discuss the idea of OPML and its two implementations, followed by simulation results to demonstrate their performance.

Gains from Massive MIMO are crucially dependent on the availability of channel state information at the transmitter which is far too costly if it has to estimated directly. Hence, for a time division duplexing system, this is derived from the uplink channel estimates using the concept of channel reciprocity. However, while the propagation channel is reciprocal, the overall digital channel in the downlink also involves the radio frequency chain which is non-reciprocal. This calls for calibration of the uplink channel with reciprocity calibration parameters to derive the downlink channel estimates. Initial approaches towards estimation of the reciprocity calibration parameters [1], [2] were all based on least squares. An ML estimator and a CRB for the estimators was introduced in [3]. This paper presents a more elegant and accurate CRB expression for a general reciprocity calibration framework. An optimal algorithm based on Variational Bayes is presented and it is compared with existing algorithms.

Optimal BeamFormers (BFs) that maximize the Weighted Sum Rate (WSR) for a Multiple-Input Multiple-Output (MIMO) interference broadcast channel (IBC) remains an important research area. Under practical scenarios, the problem is compounded by the fact that only partial channel state information at the transmitter (CSIT) is available. Hence, a typical choice of the optimization metric is the Expected Weighted Sum Rate (EWSR). However, the presence of the expectation operator makes the optimization a daunting task. On the other hand, for the particular, but significant, special case of massive MIMO (MaMIMO), the EWSR converges to Expected Signal covariance Expected Interference covariance based WSR (ESEI-WSR) and this metric is more amenable to optimization. Recently, [1] considered a multi-user Multiple-Input Single-Output (MISO) scenario and proposed approximating the EWSR by ESEI-WSR. They then derived a constant bound for this approximation. This paper performs a refined analysis of the gap between EWSR and ESEI - WSR criteria for finite antenna dimensions.

Massive multiple input multiple output (MaMIMO) is key to enabling the 1000 fold capacity improvement promised by 5G, thanks to the linear increase in system capacity with the number of base station (BS) antennas. In this work, we consider a MaMIMO interfering broadcast channel (IBC) with no coordination among the BSs under Time Division Duplexing (TDD). Each BS only has partial channel state information (CSI) of its own UE. Under this setting, we propose a duality based approach to combine the received covariance information at the BS and the estimated uplink (UL) channel from the UE to derive the downlink (DL) beamformers. We also take into account the fact that the channel reciprocity is impacted by the effect of the transmit and receive chain and correct it using the estimated reciprocity calibration factors. Essentially, using reciprocity, we transform the partial CSI at the receiver in the UL to a partial CSI at the transmitter in the downlink (DL). The technique is evaluated on the Eurecom OpenAirInterface Massive MIMO hardware testbed [1].

This work deals with hybrid beamforming for the MIMO Interfering Broadcast Channel (IBC), i.e. the Multi-Input Multi-Output (MIMO) Multi-User (MU) Multi-Cell (MC) downlink (DL) channel. Hybrid beamforming (HBF) is a low complexity alternative for fully digital precoding in Massive MIMO systems. Hybrid architectures involve a combination of digital and analog processing that enables both beamforming and multiplexing gains. We consider BF design by maximizing the Weighted Sum Rate (WSR) for the case of Perfect Channel State Information at the Transmitter (CSIT). We optimize the WSR using minorization and alternating optimization, the result of which is observed to converge fast. The design is proposed for both fully and partially connected analog BF architectures. Moreover, we consider the BF design under realistic scenarios with per-RF (radio frequency) chain or per-antenna power constraints, leading to novel interference leakage aware water filling procedures. Simulation results illustrate the good WSR performance of the designs and the gains over naive constraint satisfaction approaches.

One of the biggest challenges in operating massive multiple-input multiple-output systems is the acquisition of accurate channel state information at the transmitter. To take up this challenge, time division duplex is more favorable thanks to its channel reciprocity between downlink and uplink. However, while the propagation channel over the air is reciprocal, the radio-frequency front-ends in the transceivers are not. Therefore, calibration is required to compensate the RF hardware asymmetry. Although various over-the-air calibration methods exist to address the above problem, this paper offers a unified representation of these algorithms, providing a higher level view on the calibration problem, and introduces innovations on calibration methods. We present a novel family of calibration methods, based on antenna grouping, which improves accuracy and speeds up the calibration process compared to existing methods. We then provide the Cramér-Rao bound as the performance evaluation benchmark and compare maximum likelihood and least squares estimators. We also differentiate between the coherent and non-coherent accumulation of calibration measurements, and point out that enabling non-coherent accumulation allows the training to be spread in time, minimizing the impact to the data service. Overall, these results have special value in allowing the design of reciprocity calibration techniques that are both accurate and resource-effective.

The problem of joint Angle-of-Arrival (AoA) and calibration parameters in a wideband scenario is addressed. The system consists of multiple sources transmitting from different directions and in certain subcarriers. Sources in the same beam are regarded as transmitted from the same AoA and their signals are allocated to different subcarriers while users in different beams may transmit on the same subcarrier, i.e. the signals are multiplexed in space and frequency. Then, signals from a given beam not necessarily occupy the full bandwidth but only specific subcarriers. In addition, due to the wideband of the signal, each RF chain introduces frequency dependent gain and phase shift that need to be calibrated to perform a subsequent demultiplexing in a digital beamforming matrix. We propose a novel blind algorithm that (i) estimates the AoAs of all present sources in the bandwidth of interest and (ii) estimates the different gains/phases at each antenna per frequency up to an unknown impairment at a reference antenna. We provide identifiability conditions that ensure a successful parameter estimation. Finally, the potential of the proposed algorithm compared to the case of known calibration parameters is assessed by simulations.

The problem of Angle-of-Arrival estimation of multiple sources in the presence of mutual coupling is addressed. In this paper, we derive a Mean-Squared-Error (MSE) expression of a recently proposed algorithm that could estimate the Angles-of-Arrival of multiple sources in the presence of more mutual coupling parameters, compared to traditional methods. The MSE expression is compared with the MSE of MUSIC with known mutual coupling parameters and to the Cramer-Rao bound (CRB) of any unbiased estimator that estimates the Angles-of-Arrival in the presence of mutual coupling. It is shown that the proposed method is asymptotically unbiased. In addition, it is also shown that the method attains CRB for large number of antennas with fixed coupling parameters and uncorrelated sources. For high SNR, the CRB is not necessarily attained, however, we study the gap between the derived MSE and the CRB.

The problem of Angle-of-Arrival estimation of multiple sources in the presence of mutual coupling is addressed. The presence of unknown mutual coupling between antenna array elements is known to degrade the performance of direction-finding algorithms. We first present a result explaining why some traditional methods, that estimate Angles-of-Arrival (AoAs) of multiple sources in the presence of mutual coupling, suffer from an identifiability issue, when the number of coupling parameters exceeds a certain level. Then, we present a first method that estimates AoAs of sources when more coupling parameters are present, namely when the number of coupling parameters exceeds that certain level. Finally, we propose a refinement of the proposed algorithm, which could further enhance the AoA estimates. Simulation results have demonstrated the potential of the proposed method and its refined version, for different scenarios, as it enjoys better performance than existing methods. A better description of the paper could be found in the Conclusions section.

Performance of OFDM (Orthogonal Frequency Division Multiplexing) systems is limited by inter carrier interference (ICI) under high Doppler scenarios such as that encountered in high speed trains like TGV. The use of multiple receive antennas is known to be a very effective way to combat ICI. In a recent publication, the authors explored the use of transmit (Tx) antennas for ICI mitigation. It considered a MIMO (multiple input multiple output) scenario with perfect channel state information at transmitter (CSIT) and iteratively designed a transmit beamformer to maximize the sum capacity across all the subcarriers in the presence of ICI. In this paper, we make the design more robust by considering only partial CSIT knowledge. The beamformer is designed by optimizing the expected weighted sum rate (EWSR) under large MIMO asymptotics regime. The convergence of the beamformer follows easily from the design.

This paper addresses Particle Filter (PF)-based hybrid Cooperative Localization (CLoc) strategies consisting of fusing absolute position information from embedded Global Navigation Satellite System (GNSS) with relative distance-dependent estimates using Impulse Radio - Ultra WideBand (IR-UWB) technology. Such hybrid GNSS/IR-UWB CLoc yet cannot benefit from the high precision estimates from the IR-UWB due to the disparity between GNSS position and IR-UWB V2V ranging noises, leading to a divergence in CLoc accuracy. This paper first investigates the source of such counter-intuitive effect, and second proposes a novel adaptive Bayesian dithering technique to improve the efficiency of GNSS/IR-UWB fusion-based CLoc. This strategy increases the probability to reach a 20 cm accuracy from 50% (conventional IR-UWB and WiFi PF) to 90%.

The Interfering Broadcast Channel (IBC) applies to the downlink of (cellular and/or heterogeneous) multi-cell networks, which are limited by multi-user (MU) interference. The interference alignment (IA) concept has shown that interference does not need to be inevitable. In particular spatial IA in the MIMO IBC allows for low latency transmission. However, IA requires perfect and typically global Channel State Information at the Transmitter(s) (CSIT), whose acquisition does not scale well with network size. Also, the design of transmitters (Txs) and receivers (Rxs) is coupled and hence needs to be centralized (cloud) or duplicated (distributed approach). CSIT, which is crucial in MU systems, is always imperfect in practice. We consider the joint optimal exploitation of mean (channel estimates) and covariance Gaussian partial CSIT. Indeed, in a Massive MIMO (MaMIMO) setting (esp. when combined with mmWave) the channel covariances may exhibit low rank and zero-forcing might be possible by just exploiting the covariance subspaces. But the question is the optimization of beamformers for the expected weighted sum rate (EWSR) at finite SNR. We propose explicit beamforming solutions and indicate that existing large system analysis can be extended to handle optimized beamformers with the more general partial CSIT considered here.

This work deals with beamforming for the MIMO Interfering Broadcast Channel (IBC), i.e. the Multi-Input MultiOutput (MIMO) Multi-User Multi-Cell downlink (DL). The novel beamformers are here optimized for the Expected Weighted Sum Rate (EWSR) for the case of Partial Channel State Information at the Transmitters (CSIT). Gaussian (Posterior) partial CSIT can optimally combine channel estimate and channel covariance information. We introduce the first large system analysis for optimized beamformers with partial CSIT, here for the Massive MISO (MaMISO) case. In the case of Gaussian partial CSIT, the beamformers only depend on the means and covariances of the channels. The large system analysis furthermore allows to predict the EWSR performance on the basis of the channel statistics only.

In an effort to improve positioning accuracy in Vehicular Ad hoc NETworks (VANETs), Cooperative Localization (CLoc) has been proposed to fuse relative observations from Vehicle-to-Vehicle (V2V) communication devices with absolute observations from on-board resources such as Global Navigation Satellite Systems (GNSS), Inertial Measurement Units (IMU), and Wheel Speed Sensors (WSSs). In challenging but common tunnel environments, prolonged GNSS outages and unsustainable error accumulation of inertial sensors over time (e.g., gyroscopes) lead to the fast divergence of position estimates. In this paper, we aim at resolving this problem by relying on additional Vehicle-to-Infrastructure (V2I) measurements, making use of RoadSide Units (RSUs) internally to the tunnel. In particular, we explore practical trade-offs between V2I technology (i.e., Impulse Radio - Ultra WideBand (IR-UWB) or ITS-G5) and RSUs deployment costs (i.e., in terms of density and geometric configuration), while improving CLoc absolute accuracy. We also consider combining this solution with lane detection capabilities (e.g., camera-based) and associated road information, while comparing it with a more conventional approach based on GNSS repeaters.

A natural extension of the "Spatial" smoothing preprocessing technique is presented and analysed. It is well known that subspace methods do not work properly in the presence of coherent sources. In this paper, a "Spatio-Frequential" smoothing technique is described when the transmit OFDM symbol is received through multiple coherent signals using a uniform linear antenna array. After this preprocessing technique, one could efficiently apply any 2-dimensional subspace method to jointly estimate the angles and times of arrival of the incoming coherent signals. Simulation results demonstrate the potential of the proposed 2D smoothing method over existing separate spatial or frequential smoothing techniques.

The problem of estimating the number of superimposed signals using noisy observations from N antennas is addressed. In particular, we are interested in the case where a low number of snapshots L = O(N) is available. We focus on the Minimum Description Length (MDL) estimator, which is revised herein. Furthermore, we propose a modified MDL estimator, with the help of random matrix tools, which improves the estimation of the number of sources. Simulation results demonstrate the potential of the modified MDL estimator over the traditional one, in the case where L = O(N).

Two algorithms for the problem of joint angles and delays of arrival (JADE) of multiple paths are presented. The algorithms are based on a generalisation of the Matrix Pencil algorithm to the two dimensional case, i.e. 2D Matrix Pencils. Matrix pencil algorithms offer estimation of signal parameters, i.e. angles of arrival (AoA) or times of arrival (ToA), of multiple sources using a single snapshot. We focus on a scenario, where the OFDM symbol is transmitted in a rich multipath channel, which is the case of an indoor environment, and received through multiple antennas. The first algorithm seems more interesting than the second one, since it's motivated from an idea that most Wi-Fi systems use a large number of subcarriers compared to the number of antennas. Simulation results demonstrate the potential of the first algorithm and its performance as a function of Signal-to-Noise ratio (SNR).

Precise location services are seen as key enablers to future Intelligent Transport Systems (ITSs). Relying on Vehicle-to-Vehicle (V2V) communication links, one promising solution consists in performing distributed Cooperative Positioning (CP). More specifically, Cooperative Awareness Message (CAM) broadcasts from neighboring vehicles (seen as “virtual anchors”) are used to exchange positional information and to measure V2V radiolocation metrics such as the Received Signal Strength Indicator (RSSI). For the sake of fusing these nonlinear hybrid data, Particle Filters (PFs) represent the required positional information by a set of particles with associated weights. However, in a jointly cooperative and distributed context, the transmission of explicit particle clouds (required by receiving neighbors to update their own location estimates) is hardly affordable under limited V2V channel capacity with typical numbers of particles. In this paper we thus combine and compare several solutions in terms of message representation and adaptive transmission policy so as to reduce simultaneously CAM overhead, channel congestion and computational complexity. Proposals are made at both signal processing level (parametric density approximation) and protocol level (jointly adaptive transmission payload, power and rate), showing no impact on channel load in congested scenarios and negligible CP accuracy degradation in comparison with standard CAM transmission at critical rates.

The weighted sum rate (WSR) maximizing linear precoder algorithm is studied in large correlated multiple-input single output (MISO) interference broadcast channels (IBC). We consider an iterative WSR design which exploits the connection with Weighted sum Minimum Mean Squared Error (WMMSE) designs as in [1], [2] and [3], focusing on the version in [1]. We propose an asymptotic approximation of the signal- tointerference plus noise ratio (SINR) at every iteration. We also propose asymptotic approximations for Matched Filter (MF) precoders. Simulations show that the approximations are accurate, especially when the channels are correlated.

Uncoded transmission of a bandlimited signal over fading channels is considered. Each K signal samples are transmitted over N channels, thus increasing the sample numbers by a factor of N/K. The signals are pre-filtered to acquire certain statistical or spectral properties and then post-filtered at the receiver to obtain the best estimate. We incorporate the distortion outage probability as a benchmark to evaluate the quality of estimation. The main focus of this work is the scenario when the N fading channels are parallel, while the linear minimum mean square filter is used at the receiver. We show that the distortion outage probability vanishes inversely polynomially with SNR, with the exponent N - K + 1 for the high SNR regime, thus achieving a diversity order of N - K + 1 for the proposed oversampled uncoded transmission system.

In the setting of the two-user single-input single-output X channel, recent works have explored the degrees-offreedom (DoF) limits in the presence of perfect channel state information at the transmitter (CSIT), as well as in the presence of perfect-quality delayed CSIT. Our work shows that the same DoF-optimal performance - previously associated to perfect-quality current CSIT - can in fact be achieved with current CSIT that is of imperfect quality. The work also shows that the DoF performance previously associated to perfect-quality delayed CSIT, can in fact be achieved in the presence of imperfect-quality delayed CSIT. These follow from the presented sum-DoF lower bound that bridges the gap - as a function of the quality of delayed CSIT - between the cases of having no feedback and having delayed feedback, and then another bound that bridges the DoF gap - as a function of the quality of current CSIT - between delayed and perfect current CSIT. The bounds are based on novel precoding schemes that are presented here and which employ imperfect-quality current and/or delayed feedback to align interference in space and in time.

Ergodic interference alignment (IA) is a simple yet powerful tool that not only achieves the optimal K/2 degrees of freedom (DoF) of the K-user single-input single-output (SISO) interference channel (IC), but also allows each user to achieve at least half of its interference-free capacity at any SNR. By considering more general message sets, Nazer et al. also covered the MISO case. In this paper, we consider first the SIMO interference channel and extend ergodic IA techniques to this setting with Nrreceive antennas. Our scheme achieves KNr/(Nr+ 1), which is the DoF yielded by (standard) IA and is also the DoF of the channel when K > Nr. Moreover, this technique exhibits spatial scale invariance. By combining the existing MISO and the new SIMO results, we can also cover MIMO with Nttransmit antennas for the cases where either Nt/Nror Nr/Ntis an integer R, yielding DoF =3D min(Nt, Nr)KR/(R + 1) which is optimal for K > R.

Most techniques designed for the multiple-input multiple-output (MIMO) Interference Channel (IC) require accurate current channel state information at the transmitter (CSIT) which is not a realistic assumption because of feedback delay. We evaluate the net degrees of freedom (DoF) that different schemes can be expected to reach in a realistic system by taking into account the time and the cost of CSIT acquisition (training and feedback). A recent variant of ergodic interference alignment (IA) clearly outperforms the other schemes as its robustness to feedback delays proves to be advantageous in terms of net DoF.

This paper presents an innovative spectrum sensing scheme for Orthogonal Frequency Division Multiplexing (OFDM) signals based on enhancing the performance of the popular autocorrelation detectors (AD) using non-linear image processing methods. These methods improve the detection accuracy of the AD under particular false-alarm constraints. The proposed scheme is used in the detection of two OFDM systems, Long Term Evolution (LTE) and DVB terrestrial digital TV (DVB-T) under low signal to noise ratio (SNR) channel conditions. Results obtained show significant improvement in correct signals detection/classification up to 18% and 48% at a false-alarm of 5% and low SNR conditions equal to -18dB, using the combined AD and image processing scheme for the detection of LTE and DVB-T signals, respectively.

Most techniques designed for the the multi-input multiple-output (MIMO) Broadcast Channel (BC) andMIMO Interference Channel (IC) require accurate current and instantaneous channel state information at the transmitter (CSIT). This is not a realistic assumption because of feedback delay. A novel approach by Lee and Heath, space-time interference alignment (STIA), proves that in the underdetermined (overloaded) multi-input single-output (MISO) BC with Nttransmit antennas and K = Nt+ 1 users Nt(sum) Degrees of Freedom (DoF) are achievable if the feedback delay is not too big, thus disproving the conjecture that any delay in the feedback necessarily causes a DoF loss. However the feedback delay needs to remain less or equal to Tcover Nt+1, where Tcis the coherence time. We consider the MIMO BC and show that the use of multi-antenna receivers allows to achieve full (sum) DoF with bigger feedback delay, up to equation. We also extend this result to the MIMO IC.

Channel State Information at the Transmitter (CSIT) is of utmost importance in multi-user wireless networks, in which transmission rates at high SNR are characterized by Degrees of Freedom (DoF, the rate prelog). In recent years, a number of ingenious techniques have been proposed to deal with delayed and imperfect CSIT. However, we show that the precise impact of these techniques in these scenarios depends heavily on the channel model. We introduce the use of linear Finite Rate of Information (FRoI) signals to model time-selective channel coefficients, a model which turns out to be well matched to DoF analysis. Both the block fading model and the stationary bandlimited channel model are special cases of the FRoI channel model (CM). However, the fact that FRoI CMs model stationary channel evolutions allows to exploit one more dimension: arbitrary time shifts. In this way, the FroI CM allows to maintain the DoF unaffected in the presence of CSIT feedback (FB) delay, by increasing the FB rate. We call this Foresighted Channel Feedback (FCFB). We then consider netDoF, by accounting also for the DoF consumed in training overhead and feedback. We work out the details for the MISO broadcast channel (BC), including optimization of the number of users, and exhibit unmatched netDoF performance compared to existing approaches.

Multiple antennas facilitate the coexistence of multiple users in wireless communications, leading to spatial multiplexing and spatial division and to significant system capacity increase. However, this comes at the cost of very precise channel state information at the transmitters (CSIT). We advocate the use of channel propagation models to transform location information into (possibly incomplete) CSIT. We investigate the resulting multi-user sum rate from a DoF (Degree of Freedom, high SNR rate prelog, spatial multiplexing factor) point of view. For single-cell multi-user communications, we argue for a revival of SDMA (Spatial Division Multiple Access). In the MIMO case, the receive antennas can suppress the Non Line of Sight (NLoS) channel components to transform the MIMO channel into a MISO LoS channel, allowing the CSIT to be limited to LoS information. For the multi-cell problem, we consider the feasibility of interference aligment in the case of reduced rank MIMO channels. We then focus on the LoS components. Whereas in general MIMO multi-cell coordinated beamforming, the transmitters require global CSIT due to the coupling between transmit and receive filters, in the LoS case decoupling arises, permitting location based transmit beamforming. Location aided techniques may furthermore exploit location prediction through mobility trajectory information. This would allow slow fading (and even connectivity) predictibility, something that is difficult to achieve without location information.

Most techniques designed for the multi-input single-output (MISO) Broadcast Channel (BC) require accurate current channel state information at the transmitter (CSIT) which is not a realistic assumption because of feedback delay. A novel approach by Lee and Heath, space-time interference alignment, proves that in the underdetermined (overloaded) MISO BC with Nttransmit antennas and K = Nt+1 users Nt(sum) Degrees of Freedom (DoF) are achievable if the feedback delay is not too big, thus disproving the conjecture that any delay in the feedback necessarily causes a DoF loss. We explain this approach a bit more succinctly and evaluate the net DoF that this scheme can be expected to yield in a realistic system by taking into account the cost of CSIT acquisition (training and feedback). We term the resulting scheme ST-ZF, referring to the use of Space-Time Zero Forcing precoding. The net DoF comparison with TDMA-ZF, MAT-ZF and MAT shows that ST-ZF is also of interest in practice.

The performance of channel estimation is often assessed by deriving the proper Cramér-Rao Bound (CRB). Depending on how to treat the symbols and the channel, we have previously derived different versions of CRB. Specifically, we have dealt with the cases where the symbols and/or the channel are assumed to be either deterministic unknowns or random. Moreover, the symbols have been considered to be either jointly estimated with the channel or marginalized. All in all, we have derived six different versions of Bayesian and deterministic CRBs. However, we have shown that many of these CRBs are too optimistic in the sense that they are not strict enough to be attained by any deterministic or Bayesian estimator. In this paper we propose modified versions of those loose CRBs in the context of SIMO FIR system that are valid at least in the moderate and high SNR regimes. The analytical formulas for the lower bounds introduced are validated by some Monte-Carlo simulations.

Although most of the conventional localization algorithms rely on Line of Sight (LOS) conditions, fingerprinting allows positioning in multipath and even in Non-LOS (NLOS) environments. In contrast to the traditional Received Signal Strength (RSS), the Power Delay Profile (PDP) fingerprint may allow positioning on the basis of a single link if the multipath is rich enough. Fingerprinting is a pattern matching technique for which a performance analysis may be difficult in general. In this paper we focus on a global performance indicator, in the form of Pairwise Error Probability (PEP). Similarly to PEP analysis in communication over fading channels, we find that the PEP for PDP fingerprinting exhibits a certain diversity order, linked to the number of paths. We investigate and show the results for Gaussian Maximum Likelihood (GML) based approaches for the Rayleigh fading path amplitude case.

This paper presents a novel technique in spectrum sensing based on a new characterization of primary users signals in wideband communications. First, we have to remind that in cognitive radio networks, the very first task to be operated by a cognitive radio is sensing and identification of spectrum holes in the wireless environment. This paper summarizes the advances in the algebraic approach. Initial results have been already disseminated in few other conferences. This paper aims at finalizing and presenting the last results and the complete framework of the proposed technique based on algebraic spectrum discontinuities detection. The signal spectrum over a wide frequency band is decomposed into elementary building blocks of subbands that are well characterized by local irregularities in frequency. As a powerful mathematical tool for analyzing singularities and edges, the algebraic framework is employed to detect and estimate the local spectral irregular structure, which carries important information on the frequency locations and power spectral densities of the sensed subbands. In this context, a wideband spectrum sensing techniques was developed based on an analog decision function to multi-scale wavelet product. The proposed sensing techniques provide an effective sensing framework to identify and locate spectrum holes in the signal spectrum.

Cognitive radio is a smart wireless communication concept that is able to promote the efficiency of the spectrum usage by exploiting its free frequency bands, namely spectrum holes. Detection of spectrum holes is one of the first steps of implementing a cognitive radio system. Another step towards the feasibility and a real implementation of a cognitive radio network is the problem of location awareness. This problem arises when we do consider a realistic scenario in hybrid overlay/underlay systems, when these spectrum opportunities permit cognitive radios to transmit below the primary users tolerance threshold. In this case, the cognitive radio, have to estimate robustly the primary users locations in the network in order to adjust its transmission power function of the estimated location in the network. Adding to this the fact that in wideband radio one may not be able to acquire signals at the Nyquist sampling rate due to the current limitations in Analog-to-Digital Converter (ADC) technology, we end up with a system that should, at a sub-Nyquist rate, properly recover the bands over which the primary users transmit and estimate their location in the network. In this paper 1, we proposed to analyze all these arisen problems. During the problem formulation and when analyzing more deeply the equations related to each question apart, we will make the link between the formulation of spectrum sensing, location awareness and the hardware limitation by describing those problems in a unique compressed sensing formalism. Via the proposed framework, we made it possible to overcame a challenging postulate of fixed frequency spectrum allocation by also estimating the spectrum usage boundaries in a blind way.

In this paper, we study the sum rate performance of zero-forcing (ZF) and regularized ZF (RZF) precoding in large MISO broadcast systems under the assumptions of imperfect channel state information at the transmitter and per-user channel transmit correlation. Our analysis assumes that the number of transmit antennas M and the number of single-antenna users K are large while their ratio remains bounded. We derive deterministic approximations of the empirical signal-to-interference plus noise ratio (SINR) at the receivers, which are tight as M, K → ∞. In the course of this derivation, the per-user channel correlation model requires the development of a novel deterministic equivalent of the empirical Stieltjes transform of large dimensional random matrices with generalized variance profile. The deterministic SINR approximations enable us to solve various practical optimization problems. Under sum rate maximization, we derive 1) for RZF the optimal regularization parameter; 2) for ZF the optimal number of users; 3) for ZF and RZF the optimal power allocation scheme; and 4) the optimal amount of feedback in large FDD/TDD multiuser systems. Numerical simulations suggest that the deterministic approximations are accurate even for small M, K.

Power Delay Profile-Fingerprinting (PDP-F) al lows to do positioning in multipath and even in NLOS environments. Although many algorithms for position fingerprinting have been developed, analytical investigation in this area is still not matured. In this paper, we derive Cramer-Rao bounds (CRBs) for location dependent parameters (LDPs) when they are finite and perform local identifiability analysis under different path amplitude assumptions. We show that local identifiability of the position vector can be accomplished if a condition for the pulse shape is satisfied even with one path under the assumption that path amplitude is a genuine function of position (anisotropic path attenuation). On the other hand at least two paths are required to achieve local identifiability for a distance dependent attenuation model (isotropic path attenuation) for path amplitudes. In order to simplify the analysis we assume that pulses from different paths are non overlapping. Fisher Information Matrix (FIM) for LDPs and the position vector is derived to prove the statements.

We consider prefiltering for a single-carrier transmission over frequency-selective channels, where reduced-state trellis-based equalization is employed at the receiver, such as delayed decision-feedback sequence estimation (DDFSE) or reduced-state sequence estimation (RSSE). While previously proposed prefiltering schemes are based on the optimum filters of decision-feedback equalization (DFE), the prefiltering scheme introduced in this paper is designed according to a signal- to-interference-plus-noise ratio (SINR), whose definition takes into account explicitely the subsequent trellis-based equalizer and its complexity. In addition to the prefilter, a finite-length target impulse response for DDFSE/RSSE and an infinite- length feedback filter for state-dependent decision feedback in DDFSE/RSSE, respectively, is optimized. The developed solutions lend themselves to an interpretation of the tasks of the optimum filters. The presented numerical results show that noticeable gains can be achieved compared to state-of-the-art prefilters.

This paper considers the MISO broadcast channel with different spatial correlations of the user vector channels. The base station implements regularized zero-forcing (RZF) precoding based on an imperfect channel estimation. We derive a deterministic equivalent of the signal-to-interference plus noise ratio (SINR) by applying novel results from the field of large dimensional random matrices. Based on this deterministic equivalent, we compute the sum rate maximizing RZF precoder which is given in closed form for independent and identically distributed channels. Simulations show that the accuracy of the approximated SINR extends well into finite dimensions.

A to maximize sum rate (SR). Some of the solutions, especially the iterative ones, are capable of achieving very high performance. Nevertheless, they suffer from convergence problems. Closed form (CF) solutions on the other hand offer poor performances. This paper proposes a novel CF precoder design algorithm for the MU-MIMO BC, facilitated by stream selection. The proposed algorithm constructs the precoders per stream in a recursive manner. It contains two complementary steps: The first step consists in a recursive selection of the best available streams to minimize interference between users. This selection procedure is based on a null space criterion combined with the eigenvalues for each available stream. In the second step, the precoders are calculated according to the chosen precoding technique with eventually a power distribution optimization. Several precoding techniques can be applied, namely ZF (Zero Forcing) or MMSE (Minimum Mean Squared Error) linear beamforming (BF) and DPC (Dirty Paper Coding). We compare the proposed algorithm to some of the best CF algorithms available in the literature such as an all user SVD based recursive solution with ZF DPC precoders (ZFDPC-SUS). The obtained results demonstrate better performances for the proposed algorithm, closer to the sum capacity. We also show through a brief theoretical study that the SR of the proposed algorithm converges toward the DPC performance when the number of users grows to infinity.

In this work, we analyze the optimum expected end-to-end distortion (EED) in delay-constrained wideband multiple-input mulitple-output (MIMO) systems. We prove that the existence of frequency diversity benefits the EED in delay-constrained systems though it does not impact the ergodic channel capacity. For further analysis, we derive the closed-form expression of the optimum asymptotic expected EED, comprised of the optimum distortion exponent and the multiplicative optimum distortion factor. We present that, the optimum asymptotic expected EED decreases monotonically with frequency diversity order, but EED does not vanish with infinite frequency diversity. We also investigate the impact of spatial correlation on EED in wideband systems. The theoretical results in this paper can be guidelines for practical wideband system design.

In this paper we address the problem of Weighted Sum Rate (WSR) maximization for a K-user Multiple-Input Single-Output (MISO) cognitive Interference Channel (IFC) with linear transmit beamforming (BF) vectors in an underlay cognitive radio setting. We consider a set of L single-antenna Primary receivers to which the cognitive system can causes a limited amount of interference. We thus propose an iterative algorithm to determine the BF vectors for the secondary transmission. The optimization of the Lagrange multipliers involved in the optimization problem is based on the subgradient method. The expression of the BF vector can be interpreted as dual Uplink (UL) MMSE receiver that takes into account the interference caused by a fictitious link between the primary user and secondary base station. Finally Deterministic Annealing is applied to make the convergence of the algorithm easier.

In this paper a K user multi-input single-output (MISO) interference channel (IFC) is considered where the interference at each receiver is treated as an additional Gaussian noise contribution (Noisy IFC). We address the MISO downlink (DL) beamformer design and power allocation for maximizing the minimum SINR with per base station power constraints and imposing a minimum quality of service (QoS) requirement for each receiver. We study a distributed iterative algorithm for solving the given beamforming problem based on a combination of duality principles and the property that maxmin SINR problem is strictly related to the total power minimization problem. Finally we show that it is possible to characterize the entire Pareto boundary of the SINR (Rate) region for a K-user MISO IFC solving a sequence of maxmin SINR imposing different set of QoS constraints.

Traditionally, the performance of different semi-blind channel estimation algorithms has been assessed and compared to a certain lower bound. One of these famous lower bounds that has been extensively used in the literature is the Cramer Rao Bound (CRB). Depending on how we treat the symbols and the channel, different versions of CRB have been derived. There are two possible cases on how to treat the symbols and/or the channel namely, deterministic unknowns or random. Moreover, the symbols are either jointly estimated with the channel or eliminated. In other words, we have six different cases to be handled. In this paper we present the CRBs that exist in the literature and fit to some of these cases and derive the others in the context of SIMO single carrier cyclic prefix systems (SC-CP). On the top of that we present a unified framework that permits to derive all versions of CRBs in a concrete manner. All the derived CRBs are validated numerically by conducting limited Monte-Carlo simulations.

Many algoritms have been proposed for precoders design in a multiuser MIMO system (MU-MIMO). Nevertheless, the proposed solutions showed to have better results for some SNRs (signal to noise ratio) regions and degrade in some other parts. This paper proposes a new double iterative procedure for sum-rate maximization. The proposed algorithm is based on jointly optimizing the precoders and decoders using two different decoding schemes. The solution here is supporting multi streams per user. The algorithm is based on a WMMSE (weighted minimum mean square error) precoder combined with two iterative receivers namely the MF (matched filter) and MMSE (minimum mean square error) decoders. The resulting precoding matrices from the first algorithm (WMMSE/MF) are used as an initialization for the second one (WMMSE/MMSE). The choice of these decoders and their combination has been done according to their properties. Another crucial point in this proposal is the decision on the switching point between these two algorithms. A dynamic algorithm introducing very low extra complexity is proposed here. To validate our proposed solution we compare it with an existing MMSE and WMMSE based iterative optimization algorithms. The obtained results demonstrate significant gains without introducing supplementary complexity. Comparison with DPC (dirty paper coding) performances shows how close our proposed solution is to the BC (broadcast channel) channel capacity.

Although most of the conventional localization algorithms rely on LOS conditions, it is possible to do positioning with Power Delay Profile-Fingerprinting (PDP-F) in multipath and even in NLOS environments. Many algorithms for position fingerprinting have been developed, but analytical investigation in this area is still not matured yet. In this paper we aim to find the pairwise error probability (PEP) for PDP-F based localization systems. The objective is to see the performance of PDP-F algorithms under different cost functions and also under different path amplitude assumptions. By PEP, what is meant is the same as in the PEP analysis in digital communication channels. Hence the approach is similar for PDP-F. However its analysis is not as straightforward as it is for the digital communication channel case. We investigate and show the results for least squares (LS) based algorithm under deterministic path amplitude modeling and Gaussian Maximum Likelihood (GML) based algorithm for the Rayleigh fading modeling of the path amplitudes.

This paper considers a large multi-user time-division duplex (TDD) system, where the base station (BS) acquires channel state information via pilot signaling from the users. In the downlink the BS employs zero-forcing (ZF) and regularized zero-forcing (RZF) precoding. We derive the optimal sum rate maximizing amount of channel training using sum rate approximations from the large system analysis of MISO downlink channels under (R)ZF precoding. Moreover, in the regime of high signal-to-noise ratio (SNR), we derive approximate solutions of the optimal amount of training for both schemes that are of closed-form. By comparing the two schemes, we find that RZF requires less training than ZF, but the training interval of both schemes is equal for asymptotically high SNR. Furthermore, simulations are carried out which demonstrate the accuracy of our approximate solutions.

The attractiveness of OFDM decreases with the rising of inter-carrier interference in quickly time-varying channels. Classical OFDM low-complex detection is impaired and more elaborated techniques are required to mitigate the need for full matrix equalization. We present here a fresh approach to this subject, introducing novel fast-converging iterative techniques based on preconditioning. Moreover, we interpret windowing under a new perspective in association with the Basis Expansion Modeling of the time-varying channel. We discuss the complexity of the proposed methods, showing that they are still linear to the OFDM block size. We conclude by illustrating their competitive performance by means of numerical simulations.

Traditionally, the performance of blind SIMO channel estimates has been characterized in a deterministic fashion, by identifying those channel realizations that are not blindly identifiable. In this paper, we focus instead on the performance of Linear Equalizers for fading channels when they are based on blind channel estimates. Our analysis shows that with Zero Forcing Linear Equalizer (ZF-LE) at least one order of the diversity is lost depending on the way by which the scalar ambiguity that results from the blind channel estimation is resolved. However, in some Tx scenarios we are able to recover the diversity with MMSE-LE. Various Tx scenarios are considered in detail.

A refined estimation and tracking of the instantaneous frequency variations is desirable for a variety of audio applications (audio coding, singer segregation, music transcription and transformations, etc). In the present paper, we extend the periodic modeling with global amplitude and frequency modulation approach. We introduce a first order approximation producing an additive term involving the derivative of the `normalized waveform' multiplied by the instantaneous FM signal. The variations of the global FM get expressed through a subsampled representation and estimated using a simple least-squares scheme.

The performance of the classical low complexity OFDM detection is known to rapidly degrade with the raising of inter-carrier interference in the presence of fast time-varying channels. Advanced equalization techniques to mitigate the impact of the inter-carrier interference under those circumstances, generally involve significantly higher complexity. In this paper we address a class of reduced-complexity fast-converging iterative equalization algorithms yielding nearly-optimal performance with respect to other well-known methods. The complexity is optimized by suitable pre-conditioning in order to drastically reduce the number of required iterations necessary to achieve reliable signal detection. A detailed analysis supported by numerical results under realistic scenarios shows the nearly-optimal performance achievable by the proposed techniques with a very limited complexity, when comparing with classical zero-forcing and minimum-mean square error linear equalization.

In this paper we compare general unitary beamforming to constant modulus unitary beamforming in terms of achievable sum-rate for MIMO broadcast channels. We give a complete analysis of both schemes in the 2 × 2 multi-user MISO configuration. Furthermore, we study the regime of asymptotically high signal-to-noise ratio. For growing number of transmit antennas the sum-rate gap between both schemes is significantly increasing. Simulations are carried out that corroborate our analytical results.

The work establishes the asymptotic rate of decay for the probability of node isolation in bounded wireless sensor networks, in the high density regime. In this regime, the exposition reveals the role of the most isolated neighborhoods of the bounding region in exponentially increasing the average probability of isolation. The problem is treated for a large family of random spatial distributions of nodes, random shapes of node coverage areas, and random topography of the network's bounding region. Different examples are presented to insightfully describe the detrimental effect of boundedness in network isolation. Finally we address different aspects relating to extremely isolating bounding regions, and densities that vary exponentially in time.

This paper proposes a new double iterative procedure for sum-rate maximization in a Multiuser MIMO system (MU-MIMO). The proposed algorithm is based on joint precoder and decoder optimization involving two different decoding schemes. For that we considered a precoding algorithm namely the iterative SJNR (Signal to Jamming and Noise Ratio) precoder combined with two iterative receivers. The first receiver is the MF (Matched Filter) determining the best direction maximizing the received power for each user. The resulting receiving vector from the first algorithm will be used as an initialization for the second one. The second receiver is the MSR (Maximum Sum Rate) receiver. The selection of the switching point between these two receivers is determined and performed by a dynamic algorithm introducing very low extra complexity. To link the precoder and the selected receiver through the iterations, we use an iterative procedure based on a virtual channel calculation evolving with the system towards convergence. Finally to validate our proposed solution we compare it with an existing MMSE based iterative optimization algorithm. This algorithm is based on MMSE approach for both the transmitting and receiving side. The obtained results demonstrate significant gains without introducing supplementary complexity.

A base station (BS) transmits (Tx) and receives (Rx) signals to and from multiple mobile users (MU) through a two-way amplify and forward (AF) relay station (RS) using a two-phase protocol. The BS and the RS are both equipped with multiple antennas. In the first phase (time or frequency), the BS and all MU transmit their signals to the RS. In the second phase, the relay transmits towards the BS and MU a transformed signal in a broadcast (BC) fashion. We present a Weighted Sum Rate (WSR) maximizing approach. The optimization problem is similar as e.g., the multi-input multi-output (MIMO) BC where instead of alternating between Tx and Rx filters, one now alternates between Tx filters at BS and RS. Furthermore, we show that in the two-phase relaying considered here, there is not only a rate region for the MU in the downlink, but the coupled optimization of the BS transmitter and the RS receiver/transmitter leads in fact to an uplink/downlink rate region. Different UL and DL rates in this region can be achieved through rate tradeoffs across individual users.

Centralized algorithms for weighted sum rate (WSR) maximization for the K-user frequency-flat MIMO Interference Channel (MIMO IFC) with full channel state information (CSI) are considered. Maximization of WSR is desirable since it allows the system to cover all the rate tuples on the rate region boundary for a given MIMO IFC. First, we propose an iterative algorithm to design optimal linear transmitters and receivers. The transmitters and receivers are optimized to maximize the WSR of the MIMO IFC. Subsequently, we propose a greedy user selection algorithm based on the maximum WSR algorithm that can be applied to select a subset of transmit-receive pairs that cooperate in the interest of maximizing the sum-rate of the resulting cooperative network. To the best of our knowledge this is the first time user selection has been proposed in the context of the MIMO IFC.

For existing localization algorithms, Non-Line-of-Sight (NLOS) propagation introduces some problems for the determination of the mobile position. This is because most of these algorithms depend on the information extracted from the Line-of-Sight (LOS) path such as Time-of-Arrival (ToA) or Time-Difference-of-Arrival (TDoA) received by either one or more Base Stations. On the contrary, algorithms based on Power Delay Profile Fingerprinting (PDP-F) take advantage of the uniqueness of the multipath channel between the Base Station (BS) and the Mobile Station (MS) over the geographical region of interest. The fingerprinting approach as its name implies performs a matching between a simulated database quantity and its corresponding measured quantity and indicates a match based on some cost function between the two. In this paper, we introduce an extension to this approach by including the effects of the Doppler shifts of the paths which we call the method as Power Delay Doppler Profile Fingerprinting (PDDP-F). With the inclusion of this extra information, we aim to increase the localization accuracy by resolving the paths not only in delay dimension but also in Doppler dimension.

This paper addresses the problem of linear beamforming design in MIMO broadcast channels. An iterative optimization method for unitary beamforming is proposed, based on successive optimization of Givens rotations. Under the assumption of perfect channel state information at the transmitter (CSIT) and for practical average signal-to-noise ratios (SNR), the proposed technique provides higher sum rates than zero-forcing (ZF) beamforming while performing close to minimum-mean- squared-error (MMSE) beamforming when the number of transmit antennas equals the number of scheduled users. Moreover, it is shown to achieve linear sum-rate growth with the number of transmit antennas. Interestingly, the proposed unitary beamforming approach proves to be very robust to channel estimation errors. In the simulated scenarios, it provides better sum rates than ZF beamforming and even MMSE beamforming as the variance of the estimation error increases. When combined with simple vector quantization techniques for CSIT feedback in systems with multiuser scheduling, the proposed technique proves to be well suited for limited feedback scenarios with practical number of users, exhibiting performance gains over existing techniques.

The communication between a multiple-antenna transmitter and multiple receivers (users) with either a single or multiple-antenna each can be significantly enhanced by providing the channel state information at the transmitter (CSIT) of the users, as this allows for scheduling, beamforming and multiuser multiplexing gains. The traditional view on how to enable CSIT has been as follows so far: In time-division duplexed (TDD) systems, uplink (UL) and downlink (DL) channel reciprocity allows for the use of a training sequence in any given uplink slot, which is exploited to obtain an uplink channel estimate. This estimate is in turn recycled in the next downlink slot. In frequency-division duplexed (FDD) systems, which lack the UL and DL reciprocity, the CSIT is provided via the use of a dedicated feedback link of limited capacity between the receivers and the transmitter. In this paper, we focus on TDD systems and show that the traditional TDD CSIT acquisition fails to fully exploit the channel reciprocity in its true sense. In fact, we show that the system can benefit from a combined CSIT acquisition strategy mixing the use of limited feedback and that of a training sequence. We demonstrate the potential of our approach in terms of improved CSIT quality under a global training and feedback resource constraint.

Concerning the uncertainty of channels and peak power constraint, we give a new practical layering scheme to do reliable transmission. In our scheme, Walsh matrix is employed to do layer-time coding. Regarding columns of a layer-time coding matrix as layers and rows as time, after Walsh layer-time coding, interference among layers can be removed or diminished by adding rows up. When there are layers decoded successfully after the previous transmission, only not-yet-decoded layers will be retransmitted. Simulation results show that our Walsh layering scheme with hybrid automatic repeat request (HARQ) performs much better than the traditional single-layer ARQ sequential transmission with respect to average time delay.

Since the introduction of the diversity-multiplexing tradeoff (DMT) by Zheng and Tse for ML reception in frequency-flat MIMO channels, some results have been obtained also for the DMT of frequency-selective MIMO channels and for the DMT of suboptimal receivers such as linear (LEs) and decision-feedback equalizers (DFEs) for frequency-selective SIMO channels or frequency-flat MIMO channels. We have recently extended these results to the case of linear receivers for frequency-selective MIMO channels. However, the diversity properties of linear receivers turn out to be fairly catastrophic. In this paper we show that full diversity can be restored by the introduction of a convolutive linear MIMO precoding scheme that we showed earlier to allow to attain the optimal DMT for ML or DFE detection (in the frequency-flat case). The precoder needs to be used with a moderate amount of redundancy in the form of zero-padding, and with a MMSE design for the linear equalizer. A MMSE-ZF design also benefits substantially from the precoding. The proposed scheme is a significant extension of an earlier SISO result by Tepelenlioglu to the MIMO case.

Common localization approaches require a large amount of information to be available in order to achieve identifiability, when the signal propagates in a strictly Non-Line-of-Sight (NLOS) environment. Furthermore, even if they achieve identifiability, they usually perform poorly. In this contribution we investigate the conditions that must be met for identifiability to be feasible and for performance to be adequate. Through basic theorems and simple numerical examples, we study the benefit of exploiting additional information, if such is available and we provide intuitive conclusions that can be proved useful for any localization scheme.

A multiple fundamental frequency estimator is a key building block in music transcription and indexing operations. However, systems trying to perform this task tend to be very complex. Indeed, music transcription requires an analysis accounting for both physical and psycho-acoustical matters. In this work, we propose a physically-motivated audio signal analysis followed by an auditory-based selection. The audio signal model allows for a better time/frequency resolution tradeoff, while the auditory distance discards the redundant/non-relevant information. No prior information on the musical instrument, musical genre, and/or maximum polyphony are needed. Simulations show that the proposed technique achieves good transcription results for a variety of string and wind instruments. The proposed scheme is also shown to be robust in the presence of noise, percussive sounds and in unbalanced signal-to-interference ratio (SIR) situations.

Closed loop transmit diversity has already been adopted by 3GPP for MIMO HSDPA in the form of TxAA and its dual stream counterpart, D-TxAA. While both these transmission techniques provide performance gains for single user (SU) scenarios, they both introduce multi-user interference in the downlink in multi-user (MU) scenarios. In this paper, we study the extension of these transmission techniques to the multiuser case which entail minimal changes to the existing standard. To this end, we consider the classical MMSE chip equalizer receivers that feed back beamforming weights so as to maximize the receive SINR at each user equipment (UE). Given that the base station (BS) has to use these weights to transmit data to the UEs, we compare practical and realistic strategies that BS can employ in order to maximize downlink capacity. We derive the SINR expression for MMSE chip equalizer receivers for the general case of MU-TxAA which is used at the receivers to select optimum feedback weights. We investigate different multiuser schemes for HSDPA in the downlink (DL), compare their performance and suggest optimal strategies for single and dual stream transmission for both single and multi-antenna receivers and corroborate our arguments with simulation results. We show that for the case of single antenna receivers, scheduling users with same beamforming weights maximizes downlink capacity in TxAA. For the D-TxAA with multiple antennas at receivers (MIMO) we show that SDMA outperforms spatial multiplexing in terms of maximizing DL capacity.

Accurate location estimation of the Mobile Terminal in a strictly Non-Line-of-Sight propagation environment is still a challenging problem. Existing techniques that attempt to tackle this problem, either perform poorly or are not very practical due to their high computational complexity. In this work we present a low-complexity 2-step approach for the joint estimation of the location and the speed in the presence of nuisance parameters. This hybrid method effectively extracts the location-related information contained in the Doppler Shift, without explicit knowledge of the AOA. It then combines this information with the AOD and the TOA and outputs a Least-Square estimate. Despite its simplicity, it can achieve an accuracy of 10cm in 75% of the cases for sufficiently high SNR1.

The technologies of Multiple Input Multiple Output (MIMO) in relay protocols are considered as one of the most promising candidates for next generation wireless communication systems to extend cell coverage and enhance system capacity. However, since a relay station (RS) cannot transmit and receive a signal simultaneously when RS operates in half duplex mode, spectral efficiency of the system is deteriorated. In this paper, we consider multi-user MIMO two-way relay protocol which can improve the spectral efficiency of the relay systems. In our scenario, base station (BS) and mobile stations (MSs) exchange their own messages via RS with two-phase (two time slots), where RS does not have enough number of antennas for decoding the received signal in the time slots of multiple access channel. RS broadcast the received signal and each MS receives the signal includes unknown interferences in this case. To suppress the unknown interferences for MSs, we propose simple and effective beamforming scheme for our multi-user MIMO two-way relaying. We evaluate bit error rate (BER) performance of the proposed beamforming scheme to confirm that the proposed scheme can work well in our two-way relay scenario.

The ESPRIT algorithm is an attractive solution to many parameter estimation problems due to its low computational cost. In this paper we apply ESPRIT to the estimation of the angle of arrivals (AoA), the angle of departures (AoD), the delays and the Doppler shifts of different components of the received signal. Due to the structure of the channel impulse response matrix of a MIMO-OFDM system, these four sets of parameters can be jointly estimated via a 4-dimensional algorithm, thus the need for pairing them is eliminated. The estimates of these parameters can essentially be utilized in localization algorithms applicable to non-line-of-sight (NLoS) environments.

The absolute gain of multi-user diversity in the context of a downlink channel is the focus of this contribution. Enjoying multi-user diversity gain first requires the channel information availability at the transmitter (CSIT) in a normal downlink (DL) system. Although multi-user diversity gains have been specified well, but the burden of exchange of information required is not rigorously accounted for. We analyze a time-division duplex (TDD) broadcast channel with initial assumption of channel information neither at the base station (BS) nor at the users' side. We propose two different but simple transmission strategies which make necessary channel state information (CSI) available at both communicating ends. We give approximate analytical expressions for both strategies which capture well the resource utilization (the cost) and the gain associated to multi-user diversity. This helps us to analyze the absolute gain of the multi-user diversity. The two schemes, the one with oblivious users and the other with informed users, are compared for the sum rate and interesting conclusions are drawn.

We address the issue of low complexity linear equalization for cyclic-prefix (CP) and zero-padded (ZP) block transmissions (OFDM). While cyclic-prefix (CP) OFDM does not exploit frequency diversity offered by multipath fading, use of appropriate redundant linear preceding permits a linear equalizer (LE) to benefit from full diversity in the frequency selective channel. We exploit excess time of block transmission (TX) for low-complexity (and full-diversity) linear equalization. Relying on specific structure in the TX scheme, it is shown that excess time can take different forms and impacts diversity order of linear (block) equalization at the receiver (RX). We focus on frequency selective channels and revisit full-diversity low- complexity LE for the ZP case. We discuss special cases of ZP-OFDM with excess time that admit full-diversity LE and draw parallels with the standard ZP case.

In this paper, we investigate the behaviors of the optimum end-to-end distortion of spatially correlated, multiple-input-multiple-output (MIMO) systems. Assuming Rayleigh fading channel and the transmitter perfectly knows the instantaneous channel rate, we derive an analytic expression of the tight lower bound of the end-to-end mean quadratic distortion at any SNR for transmitting a white thermal noise source, in terms of the spatial correlation matrix, antenna numbers, the ratio of source-bandwidth to channel-bandwidth, the ratio of signal power to noise power (SNR) and the source power. By analyzing the expression, we obtain the SNR exponent and the corresponding factor at the asymptotically high SNR. Also, we show that higher correlation brings higher distortion lower bound, which corresponds to our intuition.

Location estimation in multipath environments (ME) can lead to significantly higher accuracy if the information contained in the Non-Line-of-Sight (NLOS) signal components is exploited thoroughly with the use of multiple antennas at both sides of a communication system and with the aid of appropriate channel modeling. In this contribution, this information is exploited and a Maximum-Likelihood (ML) estimator for the location of the Mobile Terminal is proposed, which is based on the reception of the signal in only one Base Station (BS). The method described herein can be characterized as direct location estimation since, in contrast to the traditional two-step approaches, the solution is formulated in terms of the received signal and not some channel-dependent parameters whose values have been estimated a-priori. The high accuracy of the method is validated using the Cramer-Rao Bound (CRB). Intuitive conclusions are reached through the comparison of different systems and for different propagation environments, which is performed by simulations.

We revisit CDMA downlink receivers (RX) based on linear minimum mean-square error (LMMSE) chip-equalizer front-end followed by a Walsh code correlator for single-input-single (multi)- output (SISO or SIMO) channels with the purpose of highlighting the non-trivial question of bias in the output of the equalizer. In a linear time-invariant channel, this bias is constant at chip-equalizer output, but evolves over time at code correlator output impacting Signal-to-Interference-plus-noise ratio (SINR) and thus achievable rates in such receivers. In principle, this bias must be taken into account in further RX/decoding stages even if its impact is small. It is shown that a new class of maximum-likelihood (ML) RX leading to potential performance gains is obtained when properly accounting for symbol-level bias across a set of user codes. These results are extended to the multi-input-multi-output (MIMO) case of UMTS high-speed downlink packet access (HSDPA).

For transmitting a continuous-amplitude source, quadratic distortion is the primary performance metric. Its lower bound decreases with channel capacity which can be increased by multiple antennas. In this paper, assuming the transmitter perfectly knows the instantaneous channel rate by feedback from the receiver and a white thermal noise source is transmitted over a quasi-static Rayleigh fading AWGN channel, we give the compact analytic expression of the tight lower bound on the optimal end-to-end mean quadratic distortion at any SNR, in terms of antenna numbers, the ratio of source-bandwidth to channel-bandwidth (SCBR), the ratio of signal power to noise power (SNR) and the source power. We analyze the expression of the lower bound to obtain the SNR exponent and the SNR coefficient at the asymptotically high SNR. Results indicate that the optimal distortion does necessarily decrease with both antenna numbers for any SCBR, the commutation between the transmit antenna number and the receive antenna number can affect the optimal distortion, and the frequency-selectivity can benefit the optimal distortion.

Optimal linear receivers for MIMO HSDPA (as for SISO/SIMO) are symbol-level (deterministic) multiuser receivers, known unfortunately to be time-varying in nature and thus prohibitively complex. Traditional less complex alternative is dimensionality-reducing linear chip-equalization followed by further non-linear (interference canceling) or joint detection stages to improve symbol estimates. Well-known versions of former include inter-stream Successive Interference Canceling (SIC) involving all codes while the latter leads to per-code joint spatial maximum-likelihood (ML) receiver. We investigate the class of MIMO HSDPA receivers based upon LMMSE chip- level MIMO (equalizer) front end, and introduce two (one static and the other time-varying) models of the resulting spatial channel, a consequence of treating the scrambler as random or deterministic. It is shown that in the random case, statistical properties can be exploited to design MIMO receivers while the deterministic point-of-view leads to another set of reduced- dimensionality linear receivers or interference cancelers.

Music transcription refers to extraction of a human readable and interpretable description from a recording of a music performance. Automatic music transcription remains, nowadays, a challenging research problem when dealing with polyphonic sounds or when removing certain constraints. Some instruments like guitars and violins add ambiguity to the problem as the same note can be played at different positions. When dealing with guitar music tablature are, often, preferred to the usual music score, as they present information in a more accessible way. Here, we address this issue with a system which uses the visual modality to support traditional audio transcription techniques. The system is composed of four modules which have been implemented and evaluated: a system which tracks the position of the fretboard on a video stream, a system which automatically detects the position of the guitar on the first fret to initialize the first system, a system which detects the position of the hand on the guitar, and finally a system which fuses the visual and audio information to extract a tablature. Results show that this kind of multimodal approach can easily disambiguate 89% of notes in a deterministic way.

In this paper, we propose to use complex orthogonal space-time block coding (COSTBC) in analog transmission with application to channel feedback. We prove that an equivalent complex orthogonal channel can be generated by COSTBC and then the matched filter bounds on the signal-to-noise ratio via multiple-input multiple-output channels are achieved by maximal ratio combining (MRC). Simulation shows that COSTBC-MRC analog schemes outperforms spatial-multiplexing oriented analog schemes and uncoded random vector quantization schemes with respect to mean-squared errors (MSE).

We consider the blind multichannel dereverberation problem for a single source. We have shown before [5] that the single-input multi-output (SIMO) reverberation filter can be equalized blindly by applying MIMO Linear Prediction (LP) to its output (after SISO input pre-whitening). In this paper, we investigate the LP-based dereverberation in a noisy environment, and/or under acoustic channel length underestimation. Considering ambient noise and late reverberation as additive noises, we propose to introduce a postfilter that transforms the MIMO prediction filter into a somewhat longer equalizer. The postfilter allows to equalize to non-zero delay. Both MMSE-ZF and MMSE design criteria are considered here for the postfilter.We also focus here on computationally efficient (FFT based) block Toeplitz covariance matrix enhancement that enforces the SIMO filtered source plus white noise structure before applying MIMO LP. A second suggested refinement is an iterative refinement between SISO and MIMO LP. Simulations show that the proposed scheme is robust in noisy environments, and performs better compared to the classic Delay-&-Predict equalizer and the Delay-&-Sum beamformer.

A key building block in music transcription and indexing operations is the decomposition of music signals into notes. We model a note signal as a periodic signal with (slow) frequency-selective amplitude modulation and global time warping. Time-varying frequency-selective amplitude modulation allows the various harmonics of the periodic signal to decay at different speeds. Time-warping allows for some limited global frequency modulation. The bandlimited variation of the frequency-selective amplitude modulation and of the global time warping gets expressed through a subsampled representation and parametrization of the corresponding signals. Assuming additive white Gaussian noise, a maximum likelihood approach is proposed for the estimation of the model parameters and the optimization is performed in an iterative (cyclic) fashion that leads to a sequence of simple least-squares problems.

In this paper, we study the optimum expected end-to-end distortion on a reproduced white thermal noise source conveyed over a flat Rayleigh fading multi-input multi-output (MIMO) channel with time-interleaving. Assuming no outage event happens, perfect channel information at the receiver and ideal interleaving, we derive the analytical expression of the tight lower bound on the expected quadratic end-to-end distortion for general signal-to-noise ratios (SNR) and analyze its asymptotic form at high SNR. Straightforwardly, the tight lower bound for no-outage cases is also a lower bound for outage cases although loose. Our results expose the mechanism of how time diversity branches benefit the end-to-end distortion for a MIMO system.

In this paper we provide a general framework for the performance analysis o pilot-aided linear channel estimators class including the general interpolation, least squares (LS), regularized LS, minimum-mean-squared-error (MMSE) and approximated MMSE estimators. The analysis is performed from the perspective of Long Term Evolution Orthogonal Frequency Division Multiple Access (LTE OFDMA) down-link systems. We also propose two novel modified MMS schemes, an Exponential Mismatched MMSE and a Simplified MMSE, to overcome the high implementation complexity of the MMSE and offering improvements to other known approximated methods. At the end, we verify the analytical results by means of Monte-Carlo simulations in terms of Normalized Mean Square Error (NMSE) and coded Bit Error Rate (BER).

In this paper, we consider stationary time- and frequency-selective MIMO channels. No channel knowledge neither at the transmitter nor at the receiver is assumed to be available. We investigate the capacity behavior of these doubly selective channels as a function of one of the system parameters, the number of transmit antennas and channel parameters as delay spread, Doppler bandwidth and channel spread factor (the product of the previous two parameters). For critically spread channels (channel spread factor of 1), it is widely believed that the dominant term of high-SNR expansion of the capacity is log(log(SNR)) or in other words, the pre-log (the coefficient of log(SNR)) is zero. We provide a very simple scheme showing that for critically spread and mildly overspread channels a non-zero pre-log exists under certain conditions. We specify these conditions in terms of the Doppler bandwidth and the delay spread. We reason that for nearly critically spread channels, MIMO systems exhibit same degrees of freedom as that of a SISO system. At higher channel spread factor (overspread case), the log(SNR) term vanishes and log(log(SNR)) term becomes the dominant capacity term. We specify the range of existence for log(SNR) regime.

The asymptotic capacity for the non-coherent MIMO stationary channels having nttransmit and nrreceive antennas with flat fading is the focus of this paper. Fading processes of concern are bandlimited. These non-coherent MIMO channels were studied by Etkin and Tse and lower bound of capacity was shown to grow with min(nt, nr)[1 - min(nt, nr)mu] log(SNR) where mu is the normalized Doppler bandwidth. The contribution of this paper is to specify the pre-log for MIMO channels and this is done by giving matching upper and lower bounds of the pre-log. Moreover min(nt, nr) factor in the pre-log term bears a small modification. The actual pre-log is min(nt, nr, 1/2mu)[1 - min(nt, nr, 1/2mu)mu] and takes into account the optimal number of streams that should be activated as a function of the Doppler bandwidth.

We consider SINR maximizing receivers based on the concept of chip-level filtering and symbol level equalization for WCDMA downlink. In this contribution we propose a new class of receivers based on channel sparsifying linear pre-processing at chip-rate followed by time-varying symbol level equalizers. Due to a sparse structure imposed on the channel (i.e. sparsification) by a chip level pre-equalizer filter which we call channel sparsifier, the effective channel after despreading presents itself as a symbol-level ISI channel. Time-varying equalization at symbol level is necessitated by the presence of aperiodic scrambler which is treated as deterministic. The optimal channel sparsifier maximizes SINR at the output of symbol level equalizer. We focus here on downlink channels that have significant dispersion in the temporal domain. Expressions for post processing SINRs (after channel sparsification, despreading and subsequent symbol level equalization) are derived for all receivers and used for performance evaluation. We show that improved receivers for WCDMA downlink can be designed benefiting from a combination of generalized channel sparsification, deterministic treatment of scrambler and non-linear equalization.

Location estimation under non-line-of-sight (NLOS) propagation conditions has been recognized as a very difficult task. In this contribution, a method that employs only one base station (BS) is proposed for tackling this problem. Its efficiency mainly stems from two factors. On one hand it exploits the information available in signal components traveling through different paths by considering proper channel modeling. On the other hand it combines the aforementioned spatial information with temporal information that is available in a dynamic channel. The source of the latter form of information is the Doppler shift. The performance of the method is further improved by considering the direct position and speed estimation from the received signal, rather than the common two-step approaches that are based on estimating channel-dependent parameters such as the angle of arrival (AOA) and/or the time of arrival (TOA), prior to localizing.

2times2 unitary preceding based on receiver feedback is applied alongside spatial multiplexing at the base station in HSDPA (D-TxAA) when the mobile terminal supports MIMO transmissions [1]. This precoding will influence achievable sum- rate of the MIMO channel if it influences the Signal-to- Interference-plus-Noise Ratio (SINR) of streams at the receiver (RX) output. We propose a set of MIMO HSDPA receivers, all based upon a LMMSE chip-level matrix filter (equalizer) front end, and introduce the notion of joint bias for the MIMO chip equalizer. Statistical properties of the spatial model thus obtained are exploited to analyze the performance of proposed MIMO receivers. It is shown that precoding choice depends upon the MIMO receiver and the extent of its impact depends on the MIMO RX.

The fifteen articles in this special issue are devoted to new equalization techniques for wireless communications, including new the latest theories and applications.

We consider least-squares (LS) downlink channel estimation for LTE OFDMA receivers. The 3GPP Long Term Evolution (LTE) project aims at continuing the competitiveness of the 3GPP standard reached with HSPA. OFDMA was chosen as multiple-access scheme for the downlink and 10 MHz bandwidth is the minimum capability for user equipments (UE). In such a context, an ill conditioning problem of the classical LS approach was found due to the un-excitation of a large portion of available sub-carriers of the LTE OFDMA symbol (due to the guard band). Two methods are investigated and compared to solve this problem: the first solution is regularization and the second one is down-sampling of the channel impulse response. The latter solution proves to perform better and allows some computational savings.

We consider a downlink channel where a base station equipped with M transmit antennas communicates with K ≥ M single-antenna receivers and has partial channel knowledge obtained via a limited rate feedback channel. We propose scalar feedback metrics that provide an estimate of the received signal-to-noise plus interference ratio (SINR), which are combined with efficient user selection algorithms and zero-forcing beamforming. The asymptotic system sum rate for large K is analyzed and numerical results are provided, showing the performance of each metric in different scenarios.

The linear prediction algorithm estimates a zero-forcing (ZF) equalizer from the SIMO channel output's second order statistics. Linear prediction can be easily extended in the presence of an additive white noise, since the white noise variance can be easily identified and compensated for in the reverberant signal covariance matrix. However, the presence of the additive noise has so far not been considered for the design of the ZF equalizer, and the resulting equalizer is not optimal. In this paper, we consider two issues in the design of the LP-based equalizer in the presence of additive white noise. First, we investigate the effect of relative subchannel delay compensation on the output SNR. We show that such relative delay can reduce considerably the output SNR. Then, we optimize the transformation of the multivariate prediction filter to a longer equalizer filter using the SNR criterion. The optimization corresponds to MMSE-ZF design, and the filter length increase allows for the introduction of some equalization delay, that can also be optimized.

System performance can be quite improved by timely full channel state information at the transmitter (CSIT), especially for multidimensional channels as MIMO. For obtaining full CSIT, it is necessary to feed back the transmit channel knowledge from the receiver to the transmitter. There are two schemes to feed back the transmit channel knowledge: One is that channel estimation is done at the receiver with feeding back the channel estimate result subsequently; the other is that the received training signals are fed back to the transmitter with channel estimation subsequently. In this paper, we assume that the channel is quasi-static Rayleigh-fading, the channel knowledge is fed back from the receiver to the transmitter in a time-discrete uncoded linear analog way without any quantization on the information, and the least-square estimator is used. We give explicit expressions of estimate error of two feedback schemes. The two schemes are compared with respect to mean-square error. It comes out that the channel-estimate-based feedback scheme is less imperfect than the received-signal-based feedback scheme with respect to mean-square error. Nevertheless, the difference is trivial at high SNR.

Non-line-of-sight and multipath propagation conditions pose significant problems for most mobile terminal positioning approaches. In contrast, power delay profile fingerprinting (PDP-F) thrives on multipath propagation. This multipath extension of T(D)oA is based on matching an estimated power delay profile from one or several base stations (BSs) (or other transmitters (broadcast, ...)) with a memorized power delay profile map for a given cell. It is obvious that the overall location accuracy depends strongly of the quality of the PDP estimation. We propose exploiting the prior knowledge on the received signal structure to enhance the PDP estimation and increase the localization accuracy. Depending on the propagation environment, we propose a deterministic and Bayesian framework for PDP estimation. And, we investigate the application to fingerprinting based mobile localization.

The problem of joint linear beamforming and scheduling in a MIMO broadcast channel is considered. We show how orthogonal linear beamforming (OLBF) can be efficiently combined with a low-complexity user selection algorithm to achieve a large portion of the multiuser capacity. The use of orthogonal transmission enables the transmitter to calculate exact signal-to-interference plus noise ratio (SINR) values during the user selection process. The knowledge of multiuser interference proves to be of particular importance for user scheduling as both the number of users in the cell and the average signal-to-noise ratio (SNR) decrease. The sum capacity of our scheme is characterized in the low-SNR regime, providing analytical results on the performance gain over zero-forcing beamforming (ZFBF). Numerical results show gains over both suboptimal and optimal ZFBF techniques in different scenarios.

Near-orthogonal transmission over MIMO broadcast channels with limited feedback is addressed, identifying the intrinsic tradeoffs that govern the system performance in terms of sum rate. A low-complexity scheme is proposed for joint scheduling and beamforming, which employs feedback in the form of quantized channel directions and lower bounds on each user's received SINR, given a maximum orthogonality factor epsi between transmit beamforming vectors (epsi = 0 if orthogonal beamforming). The authors assume a simple power allocation, which consists of distributing the available power equally over the active beams. In a system with K active users and a given average SNR, the authors show that the sum rate is a function of the number of antennas, the number of active beams (considered less than or equal to the number of antennas), the orthogonality factor epsi and the quantization codebook size. A practical rate function is derived for the described system which approximates the average sum rate accurately as validated through computer simulations. Based on the proposed rate function, optimality of SDMA vs. TDMA systems with asymptotically large number of users is shown. In addition, the authors provide a performance comparison between the proposed algorithm with limited feedback and other approaches with different level of channel state information at the transmitter.

In this paper, we consider the blind multichannel dereverberation problem for a single source. The multichannel reverberation impulse response is assumed to be stationary enough to allow estimation of the correlations it induces from the received signals. It is well-known that a single-input multi-output (SIMO) filter can be equalized blindly by applying multichannel linear prediction (LP) to its output when the input is white. When the input is colored, the multichannel linear prediction will both equalize the reverberation filter and whiten the source. We exploit the channel spatial diversity, and the speech signal non-stationarity to estimate the source correlation structure, which can hence be used to determine a source whitening filter. Multichannel linear prediction is then applied to the sensor signals filtered by the source whitening filter, to obtain source dereverberation. Particular attention is paid to the alignment of the received signals on the various microphones. This leads to an increase in the prediction performance, and allows the use of shorter predictor. The proposed approach represents hence a paradigm shift from the delay-and-sum beamformer to the delay-and-predict equalizer.

We consider a chip level decision-directed NLMS equalization scheme which targets estimating the total transmitted base station chip sequence in a decision-directed manner and using it as the desired response for equalizer adaptation. For this purpose, we explicitly use only the knowledge of the user-assigned HSPDSCH codes in order to obtain reliable signal components by hard decisions. By exploiting the equivalence between the actual multirate transmission in the sense of containing multiple spreading factors and the multicode pseudo-transmission at the single HSDPA spreading level we use also the estimated pseudo-symbols of other codes via LMMSE weightings. In addition to its reasonable complexity and Max-SINR achieving performance in realistic HSDPA working regimes, the proposed scheme also has the advantage of not requiring the channel parameters. We evaluate its performance by extensive simulations vis-à-vis the Griffiths equalizer which requires channel parameters.

Since the introduction of the diversity-rate tradeoff by Zheng and Tse for ML reception in frequency-flat MIMO channels, some results have been obtained also for the diversity behavior of suboptimal receivers such as linear and decision-feedback equalizers for frequency-selective SIMO channels. However, these results are limited to infinite length equalizers. Furthermore, so far attention has focused mostly on just diversity order aspects of diversity. In this paper we analyze the diversity of more practical FIR equalizers. We show in particular that in the case of multiple subchannels, the diversity of infinite length filters can also be attained by FIR equalizers of sufficient length. Increasing the filter lengths improves the coding gain though. Whereas the diversity order determines the slope of the asymptote at high SNR, the coding gain determines its position

We consider the issue of terminal reconfigurability in the downlink of WCDMA. For the purpose of optimizing power consumption in mobile terminals, we propose an adaptive- complexity equalization algorithm, which adapts the equalization length to the environment. A simple approach in WCDMA systems consists of computing the equalizer coefficients in frequency domain, and carry out channel equalization in time domain. The equalizer can be easily computed in frequency domain from channel estimates, which are generally obtained through pilot symbols. In our work, we decouple the task of adaptive complexity equalization in two parallel operations: variable length equalization and equalization length control. We propose a practical scheme to reduce equalization complexity by prewindowing in frequency domain and performing IFFT of variable length. An element of length control monitors the equalizer coefficients, updating the equalizer length at each stage. Both simulation and experimental results in outdoor-to- indoor scenarios show good performance and significant power savings with respect to full length equalization.

Non-Line-of-Sight and multipath propagation conditions pose significant problems for most mobile terminal positioning approaches. In contrast, power delay profile fingerprinting (PDP-F) thrives on multipath propagation. This multipath extension of T(D)oA is based on matching an estimated power delay profile from one or several base stations (BSs) (or other transmitters (broadcast, ...)) with a memorized power delay profile map for a given cell. With a single BS, an absolute time reference is required as for ToA. With multiple BSs, that requirement can be dropped as in TDoA. We propose a validation of PDP-F via simulations that can easily be reproduced. The multicellular environment consists of a big box in which multipath arises by reflection off the six sides. The resulting PDP depends on the positions of BS and terminal, the attenuation mechanism and the reflection coefficients of the six sides. We also propose an extension of the PDP-F (PSDP-F) taking into account spatial information available with an multi-antenna reception. PSDP-F can be considered as a multipath extension of the combined T(D)oA and AoA methods, without explicit requirement for antenna array calibration.

The use of multiple transmit (Tx) and receive (Rx) antennas allows to transmit multiple signal streams in parallel and hence to increase communication capacity. We have previously introduced simple convolutive linear precoding schemes that spread transmitted symbols in time and space, involving spatial spreading, delay diversity and possibly temporal spreading. In this paper we show that the use of the classical multiple-input-multiple-output (MIMO) decision feedback equalizer (DFE) (but with joint detection) for this system allows to achieve the optimal diversity-versus-multiplexing tradeoff introduced in Zheng and Tse, "Diversity and multiplexing: A fundamental tradeoff in multiple-antenna channels," IEEE Trans. Inf. Theory, May 2003, when a minimum mean squared error (MMSE) design is used. One of the major contributions of this work is the diversity analysis of a MMSE equalizer without the Gaussian approximation. Furthermore, the tradeoff is discussed for an arbitrary number of transmit and receive antennas. We also show the tradeoff obtained for a MMSE zero forcing (ZF) design. So, another originality of this paper is to show that the MIMO optimal tradeoff can be attained with a suboptimal receiver, in this case a DFE, as opposed to optimal maximum likelihood sequence estimation (MLSE)

The use of multiple transmit (TX) and receive (RX) antennas allows multiple signal streams to be transmitted in parallel and hence communication capacity to be increased. We have previously introduced simple convolutive linear precoding schemes that spread transmitted symbols in time and space, involving spatial spreading, delay diversity and possibly temporal spreading. We show that the use of the classical MIMO DFE for this system allows the optimal diversity versus multiplexing tradeoff, introduced by L. Zheng and D. Tse (see IEEE Trans. Info. Theory, vol.49, no.5, p.1073-96, 2003), to be achieved.

A key building block in music transcription and indexing operations is the decomposition of the music signal into notes. We model a note signal as a periodic signal with (slow) global variation of amplitude (reflecting attack, sustain, decay) and frequency (limited time warping). The bandlimited variation of global amplitude and frequency is expressed through a subsampled representation and parameterization of the corresponding signals. Assuming additive white Gaussian noise, a maximum likelihood approach is proposed for the estimation of the model parameters and the optimization is performed in an iterative (cyclic) fashion that leads to a sequence of simple least-squares problems. Particular attention is paid to the estimation of the basic periodic signal, which can have a non-integer period, and the estimation of the amplitude signal with guaranteed positivity.

In this paper we derive the optimal diversity versus multiplexing tradeoff for a frequency selective i.i.d. Rayleigh MIMO channel. This tradeoff is shown to be better than the one of the frequency flat MIMO channel. Traditional approaches for frequency selective channels use OFDM techniques in order to exploit the diversity gain due to frequency selectivity. We show that although coding in OFDM over a sufficient subset of subcarriers allows to exploit full diversity as such (at fixed rate), such an approach leads to a suboptimal diversity vs multiplexing tradeoff

We consider high speed packet data access service (HSDPA) which is introduced with the Release-5 of the UMTS-FDD standard. Signal to interference plus noise ratio (SINR) and throughput bounds from the usage of channel matched filter (RAKE in FIR form) and LMMSE equalizer plus correlator type mobile terminal receiver structures are obtained for the high speed downlink shared channels (HSDSCH) under certain residual intracell interference levels which represent the situations after the possible usage of front end intracell interference cancellers. Exact orthogonality factor expression is obtained which is valid for any type of linear receiver. The distributions of radio channel parameters and received powers from own and surrounding base stations are modeled under correlated shadowing w.r.t the mobile position, the cell radius and the type of environment. From such modeling, more realistic performance figures might be obtained as compared to fixing them to certain values.

The performances of the LDU (lower-diagonal-upper) factorization transform were recently shown to be equivalent to those of the Karhunen-Loeve transform (KLT), which is optimal for Gaussian sources, in the limit of high rates (Phoong, S.-M. and Lin, Y.-P., 2000; Mary, D. and Slock, D.T.M., 2001; Lahouti, F. and Khandani, A.K., 2001). We further investigate the performances of the LDU for actual transform coding (TC) schemes. Our previous results (Mary and Slock, 2001) showed that the LDU should be implemented in closed loop around the quantizers, though this leads to a noise feedback effect, similar to that occurring in DPCM systems. We develop novel analyses of these effects on the distortion-rate functions and coding gains. The proposed analyses compare our previous results, obtained for a hypothetical TC system for which the bit allocation is optimal and the rate is high, to those obtained for practical TC systems whose bit allocation is nearly optimal. By means of a theorem and numerical results, evidence is given that ordering the subsignals in the source vector by order of decreasing variance minimizes the quantization noise feedback. For the investigated practical systems, we show that deviations from the high rate assumptions arise below /spl sim/3 b/s. The effects of the noise feedback become non negligible below /spl sim/2 b/s. The LDU competes with the KLT above /spl sim/2.5 b/s.

The use of multiple transmit (TX) and receive (RX) antennas allows the transmission of multiple signal streams in parallel and hence to increase communication capacity. We have previously introduced simple convolutive linear precoding schemes that spread transmitted symbols in time and space, involving spatial spreading, delay diversity and possibly temporal spreading. Such linear precoding allows us to attain full diversity without loss in ergodic capacity. Linear precoding however cannot provide coding gain. Hence practical transmission systems have to involve channel coding. Threading is an example of a MIMO transmission system in which spatial diversity gets exploited via channel coding only. Practical symbol constellations however only allow the exploitation of a limited diversity order by channel coding. Hence, powerful yet simple MIMO TX schemes can be obtained by combining the coding gain and diversity exploitation of classical channel codes with linear precoding to exploit the remaining diversity degrees. A typical design would use channel coding to exploit temporal fading with linear precoding to exploit spatiofrequential fading.

This paper considers a SIMO or SISO communication link. In the case of white noise, the Matched filter bound (MFB) is proportional to the total channel energy. Hence all diversity sources present in the channel show up in the MFB. The MFB usually is a close approximation for the performance of maximum likelihood sequence detection (MLSD) and represents an upper bound for the performance of any receiver (Rx). In this paper we consider the diversity performance of suboptimal Rx's of the decision feedback equalization (DFE) type. Two DFE designs are considered: minimum mean squared error (MMSE) or MMSE zero-forcing (MMSE-ZF). The SNR at the detection point of a MMSE(-ZF) DFE exhibits a performance loss w.r.t. the MFB, a loss that is determined by the energy in the feedback filter. It is shown that there exists an upperbound for this loss that is channel independent. Hence the DFE enjoys as much diversity as the MFB.

Orthogonal transforms are compared with the casual transform in lossless transform coders. For single-stage lossless coding, it was previously shown that the integer-to-integer implementation of the best orthogonal decorrelating transform, the KLT, leads to lower compression performance than its casual counterpart. The analysis in the framework of a multi-stage lossless coding scheme, which yields a lossy coded signal, and an error signal was pursued. This scheme allows one to choose the respective bit rates of both complementary signals, depending for example on the bandwidth of the transmission link. It was shown that the casual approach presents several advantages w.r.t. its orthogonal counterparts. For orthogonal transforms, the price paid for the multiresolution approach is a bit rate penalty of 0.25 bit per sample. This excess bit rate is due to a "gaussianization effect" of the transforms. Firstly, it was shown under the assumptions of smooth p.d.f.s for the sources, and of high resolution for the lossy coded signal, that the casual approach allows one to code the data without causing any excess bit rate as compared with a single-stage coder. Secondly, the approach based on the casual transform allows one to easily switch between a single- or a multi-stage compressor. Thirdly, in the framework of interchannel redundancy removal, this approach allows one to easily fixed the distortion and rate for both the low resolution and the error signal of each channel, by using different stepsizes in the quantization stage. Any of the channels may, as a particular case, be chosen to be directly losslessly coded. Finally, a side advantage of the casual approach is that entropy coding of the error signal is made very simple since for odd quantization stepsizes, the discrete error sources are uniformly distributed, so that the optimal codewords have the same length, and fixed rate coding is optimal.

We address the problem of OFDM transmission over a time-varying, frequency-selective channel with high Doppler spread. This creates situations where the channel significantly evolves over the time span of one OFDM symbol. We analyze the CP-OFDM transmission mechanism, and the impairments due to channel variations, using a decomposition of these variations over a base of sinusoid functions sampling the Doppler spectrum at subcarrier frequencies. On the one hand, this leads to a fairly parsimonious parameterization of the time-varying channel impulse response. On the other hand we show that, considering a whole CP-OFDM symbol, this leads to a duality between equalization of the delay spread of a time-varying channel in the time domain, and the equalization of the Doppler spread of a frequency selective channel in the frequency domain. Using this duality, we show that equalization in the frequency domain can benefit from all known time domain equalization methods.

The main advantage of backward over forward adaptive coding schemes is to update the coding parameters with the data available at the decoder, avoiding thereby any excess bit rate. The performances of two practical backward adaptive transform coding schemes are analyzed in terms of rate and distortion for two transforms: the KLT (Karhunen-Loeve transform) and the LDU transform (based on a lower-diagonal-upper factorization of the covariance matrix, R, of the data). For both algorithms, we model the expected distortion w.r.t. the number of vectors available at the decoder. Our analysis shows that, for an algorithm using Sheppard's correction on the second order moment estimates, the distortion should converge to the target distortion. Without this correction, the effects of backward adaptation are shown to move the actual r(D) point of the system from the target point by the same term for both transforms. Simulation results confirming the theoretic analysis are presented.

For a transmitter that has a perfect knowledge of the MIMO channel, the maximum achievable capacity corresponds to the waterfilling solution. In practice, the available knowledge may only be partial due to the time selectivity of the channel, and the delay or absence of feedback from the receiver. However, exploiting the partial knowledge leads to a significant improvement when compared to the capacity without any channel knowledge. We analyze the MIMO capacity with various types of partial knowledge of the channel under practical frequency flat channel models.

The use of multiple transmit and receive antennas allows to transmit multiple signal streams in parallel and hence to increase communication capacity. Apart from capacity, the MIMO channel also offers potentially a large number of diversity sources. To exploit these diversity degrees, and hence enhance outage capacity, bit interleaved coded modulation is now a classical solution. In this paper we propose to exploit the diversity sources by linear precoding, to turn the fading channel into a non-fading one. Additional channel coding then only serves to enhance robustness against noise. To streamline the processing and analysis, the linear precoding considered here is convolutional instead or blockwise. We particularly focus in this paper on two non-iterative receiver strategies. Performance improvements are shown over conventional VBLAST.

In a transform coding framework, we compare the optimal causal approach (LDU, Lower-Diagonal-Upper) to the optimal unitary approach (Karhunen-Loeve Transform, KLT). The criterion of merit used for this comparison is the coding gain, defined for a transformation T as the ratio of the average distortion obtained with the identity transformation over the average distortion obtained with T. Both transforms are known to yield the same gain when they are computed on the signal covariance matrix R. The purpose of this paper is to compare the behavior of these two transformations when the ideal transform coding scheme gets perturbed, that is, when only an estimate R + ΔR of R is known. In this case, not only the transformation itself will be perturbated, but also the bit allocation mechanism. We compare the two approaches in two cases. Firstly, ΔR is caused by a quantization noise: the coding scheme is based on the statistics of the quantized data. We find that the coding gain in the unitary case is higher than in the causal case. In a second case, ΔR corresponds to an estimation noise: the coding scheme is based on an estimate of R based on a finite amount of available data. In this case, both causal and unitary approaches are strictly equivalent, because of the unimodularity and decorrelating properties of the transformations. Simulations results confirming the predicted behavior of the coding gains with perturbations are reported.

The use of multiple transmitter and receiver antennas allows to transmit multiple signal streams in parallel and hence to increase communication capacity. To distribute the multiple signal streams over the MIMO channel, linear space-time codes have been shown to be a convenient way to reach high capacity gains with a reasonable complexity. The space-time codes that have been introduced so far are block codes, leading to the manipulation of possibly large matrices. To reduce complexity, we propose a flexible spatial spreading and scrambling framework which allows to transmit an aribtrary number of streams. The number of streams would in practice be adjusted to fit the channel rank. Special cases of partial scrambling, in the case in which Nrxis an integer fraction of Ntx, or no scrambling, when Nrx≥ Ntx, are also considered.

For the case of white uncorrelated inputs, most of the blind multichannel identification techniques are not very robust and only allow one to estimate the channel up to a number of ambiguities, especially in the MIMO case. On the other hand, all current standardized communication systems employ some form of known inputs to allow channel estimation. The channel estimation performance in those cases can be optimized by a semiblind approach which exploits both training and blind information. When the inputs are colored and have sufficiently different spectra, the MIMO channel may become blindly identifiable up to one constant phase factor per input, and this under looser conditions on the channel. For the case of spatial multiplexing, possible cooperation between the channel inputs allows for more complex MIMO source prefiltering that may allow blind MIMO channel identification up to just one global constant phase factor. We introduce semiblind criteria that are motivated by the Gaussian ML approach. They combine a training based weighted least-squares criterion with a blind criterion based on linear prediction. A variety of blind criteria are considered for the various cases of source coloring.

In this paper iterative blind estimation of the complex amplitudes of the users is considered. A Gaussian mixture model formulation of the problem is introduced and the expectation maximization (EM) algorithm for estimation of parameters for a Gaussian mixture observation model is used. Simulation results compare the performance of the proposed algorithm with the Cramer-Rao bound.

The expectation maximization (EM) algorithm is popular in estimating the parameters of the statistical models. We consider application of the EM algorithm to maximum likelihood estimation. A hidden Markov model (HMM) formulation is used and the EM algorithm is applied to estimate the parameters of the HMM which, in turn, are used to estimate received amplitudes of the users. The proposed method is compared with that of Khan (see PIMRC, 2002) and is found to be superior.

We consider multicode and multirate transmission scenarios in a DS-CDMA system operating in an asynchronous fashion in a multipath environment. Oversampling wrt the chip rate is applied to the cyclostationary received signal and multisensor reception is considered, leading to a linear multichannel model. Channels for different users are considered to be finite-impulse response (FIR) and of possibly different lengths, depending upon their processing gains. We consider an individualized linear MMSE-ZF or projection receiver for a given user, exploiting its spreading sequence and timing information. In the multicode case, a certain user is considered to use several spreading codes in order to transmit at a higher rate. Considering different code sequences to be issuing from different virtual users, the propagation channel impulse responses of all these users are the same. However, the total channel impulse response which includes spreading sequences is different for all users. On the other hand, in the multirate case, a periodically varying set of periodic spreading codes spread successive symbols of a certain user. Symbols spread by different codes can therefore by considered to be issuing from different virtual users. The problem therefore boils down to classical multiuser detection with time-invariant interference canceling filters for each virtual user. A blind channel estimate is also obtainable through Capon's method (first used by Tsatsanis) as a by-product of the MMSE-ZF receiver algorithm.

This paper deals with optimal coding for vectorial signals by means of a decorrelating transform such as DPCM. We show that the optimal causal transform corresponds to a (lower-diagonal-upper) triangular factorization of the autocorrelation matrix of the signal : the transformation matrix is triangular and unit diagonal. Each one of its rows is the optimal prediction filter for the corresponding component of the vector to be coded. We analyze the effect on the coding gain of the perturbation due to backward adaptation (prediction based on the quantized signal), as for DPCM coders. We then show that two previously introduced transformations, in the context of subband coding, appear as special cases of vectorial DPCM coding, and we compare these two transformations when perturbations occur on the reference signal. Finally, we apply some results of vectorial DPCM coding to wideband speech coding.

In DS-CDMA communications, the conventional receiver is the RAKE receiver. In the downlink (base station to mobile) signalling with cell-dependent scrambling, orthogonal codes and a common channel for all the users, this receiver does not maximize the signal-to-interference-plus-noise ratio (SINR) at its output. Another receiver, with the same structure as the RAKE receiver, is suitable for downlink DS-CDMA communications; the one in which the channel matched filter gets replaced by a filter that is designed to maximize the SINR at the receiver output. We analyze the use of three different transmission diversity (TD) techniques, namely space-time TD (STTD), orthogonal TD (OTD) and delay TD (DTD). All of them are compared for the two receiver structures: RAKE and max-SINR receivers. The max-SINR receiver structures proposed here for the three TD modes are new and are shown to usually significantly outperform the RAKE schemes. We also discuss the relative performance merits of the three TD schemes for one or the other receiver structure.

The conventional receiver for DS-CDMA communications is the RAKE receiver which is a matched filter (MF), matched to the operations of spreading, pulse shape filtering and channel filtering. The RAKE receiver assumes a sparse/pathwise channel model so that the channel matched filtering gets done pathwise, with delay adjustment and decorrelation per path and maximum-ratio combining of path contributions at the symbol rate. Original RAKE receivers work with continuous delays, which are tracked by an early-late scheme. This requires signal interpolation and leads to suboptimal treatment of diffuse portions in the channel impulse response. These disadvantages can be avoided by a discrete-time RAKE, operating at a certain oversampled rate. Proper sparse modeling of the channel is an approximation problem that requires exploitation of the limited bandwidth of the pulse shape. We propose and simulate a number of sparse channel approximation algorithms along the lines of matching pursuit, of which the recursive early-late (REL) approach appears most promising. We also analyze and simulate the effect of channel estimation on the RAKE output SINR.

For the case of white uncorrelated inputs, most of the blind multichannel identification techniques are not very robust and only allow to estimate the channel up to a number of ambiguities, especially in the MIMO case. On the other hand, all current standardized communication systems employ some form of known inputs to allow channel estimation. The channel estimation performance in those cases can be optimized by a semiblind approach which exploits both training and blind information. When the inputs are colored and have sufficiently different spectra, the MIMO channel may become blindly identifiable up to one constant phase factor per input, and this under looser conditions on the channel. For the case of spatial multiplexing, possible cooperation between the channel inputs allows for more complex MIMO source prefiltering that may allow blind MIMO channel identification up to just one global constant phase factor. We introduce semiblind criteria that are motivated by the Gaussian ML approach. They combine a training based weighted least-squares criterion with a blind criterion based on linear prediction. A variety of blind criteria are considered for the various cases of source coloring.

Certain blind channel estimation techniques allow the identification of the channel up to a scale or phase factor. This results in singularity of the Fisher information matrix (FIM). The Cramer-Rao bound, which is the inverse of the FIM, is then not defined. To regularize the estimation problem, one can impose constraints on the parameters. In general, many sets of constraints are possible but are not always relevant. We propose a constrained CRB, the pseudo-inverse of the FIM, which gives, for a minimum number of constraints, the lowest bound on the mean squared estimation error.

The purpose of semi-blind channel identification methods is to exploit the information used by blind methods and the information coming from known symbols. The main focus of this paper is the study of deterministic quadratic semi-blind algorithms which are of particular interest because of their low computational complexity. The associated criteria are formed as a linear combination of a blind and a training sequence based criterion. Through the examples of subchannel response matching and subspace fitting based semi-blind criteria, we study how to construct properly such semi-blind criteria and how to choose the weights of the linear combination. We provide a performance study for these algorithms and give theoretical conditions for the semi-blind performance to be independent of the weights.

We explore the channel estimation problem in the case of quasi-synchronous users in a DS-CDMA system. Knowledge of the transmit (TX) filter is assumed, and the anti-aliasing low-pass front end receive (RX) filter is designed for critical sampling at the Nyquist rate for the TX filter. It is shown that when the sampling frequency is larger than the Nyquist frequency, the discrete-time representation of the channel is not unique. However, all representations can be treated in a similar fashion once the Nyquist rate is satisfied. On the other hand, fractionally sampling the channel leads to a scenario in which the cut-off frequency can be approached arbitrarily close to the Nyquist rate. In the case of sparse channels, sampling the channel at any rate lends to a small number of non-zero coefficients in the finite-impulse response(FIR) representation of the channel. The structured channel estimation algorithm presented in this paper exploits the sparseness of this model. Results are compared with those of other previously proposed structured methods.

We address the problem of downlink interference rejection in a DS-CDMA system. Periodic orthogonal Walsh-Hadamard sequences spread different users' symbols followed by scrambling by a symbol aperiodic base-station specific overlay sequence. The point-to-point propagation channel from the cell-site to a certain mobile station is the same for all downlink signals (desired user as well as the intracell interference). Orthogonality of the underlying Walsh-Hadamard sequences is destroyed by multipath propagation, resulting in multiuser interference if a coherent combiner (the RAKE receiver) is employed. In this paper, we propose a blind linear equalization algorithm which equalizes for the common downlink channel, thus rendering the user signals orthogonal again. A simple code matched filter subsequently suffices to cancel the multiple access interference (MAI) from intracell users. It is shown that the receiver maximizes the signal-to-interference plus noise ratio (SINR) at its output.

One of the main problems with linear multiuser detectors for DS-CDMA systems with large spreading factors and time-varying multipath propagation is that typically not enough data is available to estimate the detector's parameters well. Pathwise processing is an approach that allows a separation between rapidly varying and slowly varying parameters. In this approach, which was introduced by Matti Latva-aho (see PhD thesis, Oulu University, Finland, 1998), the scarce training data are used to estimate the few rapidly varying parameters while the whole received signal can be used to estimate the slowly varying parameters. We present some refinements to the original pathwise processing approach to avoid signal cancellation due to correlation between paths in slowly varying environments. We also consider the extension to spatio-temporal processing and propose the introduction of structural constraints in the detector filters to reduce complexity and facilitate the practical implementation.

We address the problem of blind multiuser multichannel identification in a spatial division multiple access (SDMA) context. Using a stochastic model for the input symbols and only second order statistics, we develop a simple algorithm, based on the generalized Schur algorithm to apply LDU decomposition of the covariance matrix of the received data. We show that this method leads to identification of the channel, up to a unitary mixture matrix. Furthermore, the identification algorithm is shown to be robust to channel length overestimation and approaches the performance of the weighted linear prediction (WLP) method, at low computational cost.

We consider p users in a DS-CDMA system operating asynchronously in a multipath environment. Oversampling w.r.t. the chip rate is applied to the cyclostationary received signal and multi-antenna reception is considered, leading to a linear multichannel model. Channels for different users are considered to be FIR and of possibly different lengths. We consider an individualized linear MMSE receiver for a given user, exploiting its spreading sequence and timing information. The blind determination of the receiver boils down to the blind channel identification. We explore blind channel identifiability requirements. Sufficiency of these requirements is established and it is shown that if zero-forcing conditions can be satisfied, then the CDMA channel (and hence the receiver) is identifiable with probability 1. It is also shown that linear MMSE receivers obtained by different criteria (including a new one) have the same identifiability requirements asymptotically in SNR.

We investigate maximum likelihood (ML) methods for blind and semi-blind estimation of multiple FIR channels. Two blind deterministic ML (DML) strategies are presented. In the first one, we propose to modify the iterative quadratic ML (IQML) algorithm in order to "denoise" it and hence obtain consistent channel estimates. The second strategy, called pseudo-quadratic ML (PQML), is naturally asymptotically denoised. Links between these two approaches are established and their global convergence is proved. Furthermore, we propose semi-blind ML techniques combining PQML with two different training sequence estimation methods and compare their performance. These semi-blind techniques, exploiting the presence of known symbols, outperform their blind version. They also allow channel estimation in situations where blind and training sequence methods fail separately. Simulations are presented to demonstrate the performance of all the proposed algorithms, and comparisons between them are discussed in a blind and/or semi-blind context.

We derive a new adaptive filtering algorithm called the instrumental variable affine projection (IVAP) algorithm and give its fast version (FIVAP algorithm). The IVAP algorithm departs from the AP algorithm and uses an IV. The IV process is generated in a way such that the new algorithm combines between the AP and the fast Newton transversal filter (FNTF) algorithms. Simulations show that the IVAP algorithm is more robust to noise than the AP algorithm. With the IV, the sample covariance matrix loses its Hermitian property and its displacement structure is different from the one of the AP algorithm. Consequently, the derivation of a fast version is done by deriving the IV sliding window covariance fast transversal filter (IV SWC FTF) algorithm. Using this and other ingredients, we derive the FIVAP algorithm whose computational complexity is nearly the same as the FAP algorithm.

We present two approaches to stochastic maximum likelihood identification of multiple FIR channels, where the input symbols are assumed Gaussian and the channel deterministic. These methods allow semi-blind identification, as they accommodate a priori knowledge in the form of a (short) training sequence and appears to be more relevant in practice than purely blind techniques. The two approaches are parameterized both in terms of channel coefficients and in terms of prediction filter coefficients. Corresponding methods are presented and some are simulated. Furthermore, Cramer-Rao Bounds for semi-blind ML are presented: a significant improvement of the performance for a moderate number of known symbols can be noticed.

We pursue our iterative quadratic maximum likelihood (IQML) approach to blind estimation of multiple FIR channels. We use a parameterization of the noise subspace in terms of linear prediction quantities. This parameterization is robust w.r.t. a channel length mismatch. Specifically, when the channel length is overestimated, no problems occur. Underestimation leads to a reduced-order channel estimate. We introduce two matched filter bounds (MFBs) to characterize the performance of receivers using reduced-order channel models. The first one (MFB1) uses the channel model to perform the spatio-temporal matched filtering that yields data reduction from multichannel to single-channel form. The rest of the processing remains optimal. MFB2 on the other hand bounds the performance of the Viterbi algorithm with the reduced channel model. It is shown that the reduced model provided by IQML is the one that maximizes MFB1. We also propose some low complexity techniques for obtaining consistent estimates with which to initialize IQML.

We consider mobile radio communications with one user of interest and possibly interfering users and noise, over several discrete-time channels obtained either by oversampling or from multiple antennas. The optimal receiver structure for one signal of interest plus spatially and temporally correlated noise is MLSE equalization with an appropriately weighted metric for vector signals. We show however that we can alternatively pass the vector received signal through both a MISO (multi-input single output) matched filter and a MIMO blocking equalizer. The blocking equalizer output is independent of the signal of interest and is used as the input to a MISO Wiener filter that reduces the noise in the matched filter output. The training sequence of the signal of interest can be used to estimate the corresponding channel, from which the matched filter and blocking equalizer can be determined. The remaining quantities can be adapted from the available signals.

Blind channel identification methods based on the oversampled channel output is a problem of theoretical and practical interest. It is first demonstrated that the subspace methods developed in Moulines are not robust to errors in the determination of the model order. An alternative solution is then proposed, based on a linear prediction approach. The effect of overestimating the channel order is investigated by simulations: it is demonstrated that the prediction error method is "robust" to over-determination.

Slock and Papadias (1994) showed that in the case of multiple antennas and/or oversampling, FIR ZF equalizers exist for FIR channels and can be obtained from the noise-free linear prediction (LP) problem. The LP problem also leads to a minimal parameterization of the noise subspace, which was used to solve the deterministic maximum likelihood (DML) channel estimation problem. The present authors provide further contributions along two lines. One is a number of blind equalization techniques of the adaptive filtering type. They also present some robustifying modifications of the DML problem.

We present a new class of adaptive filtering algorithms for blind equalization of constant modulus signals. The algorithms are first derived in a classical system identification context by minimizing at each iteration a deterministic criterion and then their counterpart for blind equalization is derived by modifying this criterion taking into account the constant-modulus property of the transmitted signal. The algorithms impose more constraints than the classical constant modulus algorithm (CMA) and as a result achieve faster convergence. An asymptotic analysis has provided useful parameter bounds that guarantee the algorithms' stability. A priori knowledge of these bounds helps the algorithms escape from undesirable local minima of their cost function thus giving them a potential advantage over the classical CMA. An efficient computational organization for the derived algorithms is also proposed and their behaviour has been tested by means of computer simulations.>

Equalization for digital communications constitutes a very particular blind deconvolution problem in that the received signal is cyclostationary. Oversampling (OS) (w.r.t. the symbol rate) of the cyclostationary received signal leads to a stationary vector-valued signal (polyphase representation (PR)). OS also leads to a fractionally-spaced channel model and equalizer. In the PR, channel and equalizer can be considered as an analysis and synthesis filter bank. Zero-forcing (ZF) equalization corresponds to a perfect-reconstruction filter bank. We show that in the OS case FIR ZF equalizers exist for a FIR channel. In the PR, the multichannel linear prediction of the noiseless received signal becomes singular eventually, reminiscent of the single-channel prediction of a sum of sinusoids. As a result, the channel can be identified from the received signal second-order statistics by linear prediction in the noise-free case, and by using the Pisarenko method when there is additive noise. In the given data case, MUSIC (subspace) or ML techniques can be applied.>

Equalization for digital communications constitutes a very particular blind deconvolution problem in that the received signal is cyclostationary. Oversampling (OS) (w.r.t. the symbol rate) of the cyclostationary received signal leads to a stationary vector-valued signal (polyphase representation (PR)). OS also leads to a fractionally-spaced channel model and equalizer. In the PR, channel and equalizer can be considered as an analysis and synthesis filter bank. Zero-forcing (ZF) equalization corresponds to a perfect-reconstruction filter bank. We show that in the OS case FIR ZF equalizers exist for a FIR channel. In the PR, the noise-free multichannel power spectral density matrix has rank one and the channel can be found as the (minimum-phase) spectral factor. The multichannel linear prediction of the noiseless received signal becomes singular eventually, reminiscent of the single-channel prediction of a sum of sinusoids. As a result, a ZF equalizer can be determined from the received signal second-order statistics by linear prediction in the noise-free case, and by using a Pisarenko-style modification when there is additive noise. In the given data case, Music (subspace) or ML techniques can be applied. We also present some Cramer-Rao bounds and compare them to the case of channel identification using a training sequence.>

The authors consider the problem of designing multiresolution transforms that are adapted to the given image signal, in the sense that they maximize the coding gain at each resolution level. A simple alternating optimization algorithm is derived for solving this problem in the framework of the lattice realization of para-unitary quadrature mirror filters. The resulting large coding scheme is discussed in some detail, and its performance is compared with that of the discrete cosine transform (JPEG) technique and with that of some nonadapted multiresolution transforms.>

Consideration is given to the problem of designing a signal-adapted two-band filtering system that has the perfect reconstruction property for application to multiresolution signal coding. A natural optimization criterion for this problem is introduced, and an iterative algorithm for computing optimal filters in lattice form is described. Some simulation results for the application of the technique to image coding are presented.>

The topic of adaptive filtering is addressed for the case when the input and desired-response signals are split into several subbands, leading to a decomposition into multiple smaller adaptive filtering problems. The use of fractionally spaced adaptive filters to overcome the aliasing problem is proposed. This concept is shown to apply to the subband filtering setup. The fractional spacing idea is applied to several existing alternative approaches to the aliasing problem, leading to more economical solutions. Tree structures of subband filters are introduced to obtain multiresolution filters, and it is proposed to apply these filters to the demanding problem of acoustic echo cancellation. A three-fold gain, involving the modeling, tracking, and criterion (perception) issues, is obtained.>

Some preliminary results are presented on a novel approach to the analysis of the propagation of round-off errors in the fast transversal filter (FTF) recursive least squares (RLS) algorithm. This approach is based on the concept of backward consistency which can be applied to any recursive algorithm, e.g. to the class of Kalman filtering algorithms. The backward consistency concept is applied to the FTF algorithm. This application leads to the introduction of the FTF state variables that are backwardly consistent. In other words, each point on the FTF manifold represents a value for the FTF state variables that corresponds exactly to the solution of a prewindowed shift-invariant least-squares (LS) problem. The advantage of this approach is that the error propagation on the FTF manifold corresponds exactly (without averaging or even linearization) to the propagation of a perturbation on the input data in the LS problem. The dynamics of this perturbation are analyzed.>

Traditionally, there have been two groups of fast recursive least squares (RLS) algorithms, the fixed-order fast transversal filter (FTF) algorithms and the order-recursive fast lattice (FLA) algorithms. More recently, a third group of fast RLS algorithms has been introduced, the so-called fast QR RLS (FQR) algorithms. Although this group has been introduced as a third independent group of fast RLS algorithms, it is shown that the FQR algorithms and the FLA algorithms are essentially the same group of algorithms and that it is basically only the way in which these algorithms are derived that makes them appear to be different. However, the FQR algorithms are not identical to any particular member of the FLA group; although the same identities are used to update the same quantities, the way in which these identities are tied together to form a complete algorithm is different. However, various members within the FLA group itself also display such differences. In this way, the reconciliation brings out several interesting (e.g. numerical) aspects. Various new algorithms are discussed.>

The authors present scalar implementations of multichannel fast recursive least squares algorithms in transversal filter form (so-called FTF). By processing the different channels sequentially, i.e one at a time, the processing of any channel reduces to that of the single-channel algorithm. This sequential processing decomposes the multichannel algorithm into a set of intertwined single-channel algorithms. Geometrically, this corresponds to a modified Gram-Schmidt orthogonalization of multichannel error vectors. Algebraically, this technique corresponds to matrix triangularization of multichannel error covariance matrices and converts matrix operations into a regular set of scalar operations. Algorithm structures that are amenable to VLSI implementation on arrays of parallel processors follow naturally from this approach. Numerically, the resulting algorithm benefits from the advantages of triangularization techniques in block processing. Stabilization techniques for control of numerical error propagation in the update recursions are incorporated.>

The problem of numerical stability of fast recursive least-squares transversal filter (FTF) algorithms is addressed. The prewindowing case with exponential weighting is considered. A framework for the analysis of the error propagation in these algorithms is developed. Within this framework, it is shown that the computationally most efficient 7N form (dealt with by G. Carayanmis et al. (1983) and by J.M. Cioffi (1984)) is exponentially unstable. By introducing redundancy in this algorithm, feedback of numerical errors becomes possible. This leads to a numerically stable FTF algorithm with complexity 9N. The results are presented for the complex multichannel joint-process filtering problem.>

The important problem of Adaptive Line Enhancing (ALE) is addressed in this paper. Its solution involves an Adaptive Notch Filter (ANF) proposed in [1],[2] using a minimal parameter constrained infinite impulse response (IIR) model in conjunction with the Recursive Prediction Error Method (RPEM) [3]. A Fast Transversal Filter (FTF) algorithm for the adaptive RLS-type updating of the linear phase filter is presented.