This letter proposes a deep learning-based channel estimation method for reconfigurable intelligent surface (RIS)-assisted multiple-input multiple-output (MIMO) full-duplex (FD) systems under hardware impairments (HI). Unlike traditional model-based estimators that rely on ideal assumptions and require complex analytical models for non-idealities, the proposed approach leverages data-driven learning to capture nonlinear distortions. Learning directly from data allows the model to adapt to practical impairments without explicit modeling, resulting in enhanced robustness and low-latency inference. Simulation results demonstrate that the proposed method achieves superior estimation performance, highlighting its suitability for real-time deployment in hardware-impaired environments.

This work introduces a new semi-blind multi-modulus approach for joint source separation and carrier frequency offset (CFO) estimation in Multiple-Input Multiple-Output Orthogonal Frequency Division Multiplexing communications. A hybrid semi-blind cost function based on the Least Squares (LS) and Multi-Modulus (MM) is designed for source recovery and CFO estimation. An alternating gradient descent optimization method is utilized to iteratively minimize the cost function along the search direction. Furthermore, an efficient initialization procedure is introduced to reduce the algorithm's complexity and ensure its convergence. The simulation results show that our method performs very well in terms of CFO estimation and data recovery.

This research presents a comprehensive investigation into subjective video shakiness assessment. A collection of 30 shaky videos was gathered, covering relevant categories such as climbing, driving, large parallax, rotation, running, and walking, with different scenes and levels of shakiness. A pairwise comparison (PWC) was conducted, involving human observers who evaluated the perceived quality of shaky videos, and the results have been converted into quality scores using the Just-Objectionable-Differences (JOD) scaling method. The shakiness assessment framework was proved effective by correlations between objective metrics and subjective judgments, and it can serve as a benchmark for future advancements in the field, fostering improvements in video stabilization technologies and applications. The complete dataset is made publicly available through the following link:Shakiness-QuAD

The goal of this innovative practice paper is to describe the implementation of a project-based learning mechanism in the academic curriculum. One such mechanism called ‘Innovative Projects’ has been implemented at our university to systematically foster an innovation culture among all students of engineering and technology programs at the grassroots level. The primary goal of this initiation is to encourage, inspire, and nurture all freshman students of various engineering and technology programs by assisting them in developing innovative ideas and prototyping them. This initiative provides an excellent opportunity for new undergraduate students to develop innovative ideas using Arduino and Raspberry Pi embedded boards. Innovative projects also provide an opportunity for freshman students to interact with first year students from other departments in order to improve their research skills and encourage interdisciplinary projects. The plan allows students to select their own project group members and a faculty advisor, allowing them to take advantage of intensive course training and laboratory facilities. This strategic plan ensures that students gain knowledge through a project-based learning process to solve many real-world technology-related problems. In this paper, we will describe our experiences of successfully implementing innovative project courses in the freshman year of all engineering and technology programs offered at our university. In the previous and present academic years, two courses on innovative projects were introduced for the first time in the curricula of all engineering and technology programs offered at our university. These two courses are named Innovative Project-Arduino Using Embedded C (CSE 1002) and Innovative Project-Raspberry Pi Using Python (CSE 1003). The CSE 1002 and CSE 1003 courses were offered for the first time in the freshman year during the spring semester of the academic year 2021–22 and during the fall semester of the academic year 2022–23, respectively. The introduction of these two innovative project courses has completed its initial successful pilot run for the first batch of over 5000 freshman students. These two courses are currently being offered for the second batch of over 5000 freshman students. Experiences gained during the pilot run are being infused into the second batch offering of these courses and will be shared in this article. All aspects starting from the design of these two innovative projects courses, the establishment of a dedicated innovative laboratory to support these courses, pedagogical approaches adopted for teaching these courses, assessment, and evaluation of these courses will be discussed. The introduction and successful implementation of such innovative project courses on a massive scale in the freshman year of all engineering and technology programs at the university level is a novel concept. Focus groups conducted with the freshman students of the various engineering and technology programs demonstrate that they valued this opportunity because it provided them with the necessary technical and entrepreneurial skills that may be required to succeed in the capstone projects later in their respective undergraduate programs.

In digital communication, nonlinearity is a frequent source of signal and channel distortion. Such distortions often need equalization techniques or devices to correct them. An equalization technique for recovering nonlinear multichannel signals in convolutive mixture is presented in this article. The proposed work uses the mutually referenced equalization technique to estimate the equalizer and the transmitted signal from the nonlinear convolutive mixture while in the present of quadratic nonlinearities. The proposed semi-blind model builds a cost function that offers an equalization solution by combining data, pilots, and the mutually referenced equalizer approach. The proposed method offers a number of benefits, including ease of implementation, resistance against channel order misspecification, and the ability to provide several equalization delays with a single solution. The simulation findings demonstrate that the proposed approach has fascinating performance characteristics and is resilient to low SNR.

A novel class of fast Multi-Modulus algorithms (fastMMA) for Blind Source Separation (BSS) and deconvolution are presented in this work. These are obtained through a fast fixed-point optimization rule used to minimize the Multi-Modulus (MM) criterion. Here, two BSS versions are provided to separate the sources either by finding the separation matrix at once or by separating a single source each time using a fast deflation technique. Further, the latter method is extended to cover systems of convolutive nature. Interestingly, these algorithms are implicitly shown to belong to the fixed step-size gradient descent family, henceforth, an algebraic variable step-size is proposed to make these algorithms converge even much faster. Apart from being computationally and performance-wise attractive, the new algorithms are free of any user-defined parameters.

In this paper, we introduce a novel approach for 2D blind multichannel identification using the helix transform in conjunction with the Cross Relation (CR) method. The helix transform is used to convert the 2D convolution of image and channels into 1D convolution. The CR method, known for its simplicity, efficiency and low computational cost, is then adapted and used to estimate the unknown channel coefficients. A main advantage of the proposed approach resides in its ability to help extending the plethora of methods from 1D to 2D blind system identification and ease their implementations.

In this paper, a robust iterative block decision feedback equalization (DFE) algorithm is developed for uplink single-carrier frequency division multiple access (SC-FDMA) systems. Three important problems that can adversely affect the performance of the DFE in SC-FDMA systems are to be addressed here, i.e., the feedback symbols reliability, the feedback correlation metric, and the phase noise due to inaccuracies in the fabrication process of the crystal oscillator. Instead of using all the detected symbols in the feedback loop of the DFE, only the highly reliable symbols are selected to be fed back. This results in the improvement of the error propagation, one of the common problems in a DFE. Also, the feedback correlation is an important metric in the design of the DFE, and hence an elegant method is proposed in this design and found to perform better than the available existing designs in the literature. Finally, the transmitter and receiver phase noise is iteratively compensated using its corresponding time- and frequency-domain properties. Simulation results demonstrate the robustness of our designed iterative block DFE.

In this work, a variable step-size (VSS) blind equalization algorithm is presented for a multimodulus blind equalization scheme. The parameters of the proposed algorithm are selected using a particle swarm optimization strategy. Eventually, the best parameters for the proposed VSS blind equalization algorithm are obtained and better performance is obtained when compared to the trial-and-error method for choosing these parameters. Ultimately, a considerable reduction in computational complexity when compared to the fixed-step algorithm with extra constraints. Simulation results support this new proposed technique.

In this paper, the problem of blind equalization and source separation of convolutive Multi-Input Multi-Output (MIMO) system is solved using Givens/Shear rotations. Targeting the Multi-Modulus (MM) signals and exploiting the second-order decorrelation among the transmitting sources, two efficient Givens Muti-Modulus (G-MMDDA) and Hyperbolic Givens (HG-MMDDA) Deconvolution algorithms are proposed for the first time. These solutions can be seen as extensions of the blind source separation MM-based method by Shah et al (2015) to the more general case of blind deconvolution for memory MIMO channels. The resulting algorithms are quite appealing as they combine both a good speed of convergence with low computational complexity.

In this paper, we consider bidirectional decode-and-forward (DF) buffer-aided relay selection and transmission power allocation schemes for underlay cognitive radio (CR) relay networks. First, a low complexity delay-constrained bidirectional relaying protocol is proposed. The proposed protocol maximizes the single-hop normalized sum of the primary network (PN) and secondary network (SN) rates and controls the maximum packet delay caused by physical layer buffering at relays. Second, optimal transmission power expressions that maximize the single-hop normalized sum rate are derived for each possible transmission mode. Additionally, the impacts of several system parameters including maximum buffer size, interference threshold, maximum packet delay and number of relays on the network performance are also investigated. The results reveal that the proposed bidirectional relaying protocol and antenna transmission power allocation schemes introduce a satisfactory performance with much lower complexity compared to the optimal relay selection and power allocation schemes and provide an application dependent delay-controlling mechanism. It is also found that the network performance degrades as the delay constraint is more restricted until it matches the performance of conventional unbuffered relaying with delay constraints of three. Additionally, findings show that using buffer-aided relaying significantly enhances the SN performance while slightly weakens the performance of the PN.

This paper investigates the impact of certain side information that are available in the channels and/or signals on the blind system identification through the Cramer-Rao Bound (CRB). More precisely, we considered a Single Input Multiple Output (SIMO) system, and studied, for different scenarios, the performance bounds for channel estimation in both deterministic and Bayesian cases. The latter correspond to the situations where side information is brought by either a pilot sequence (semi-blind case), channel sparsity (specular channel case) or certain signal's statistical properties such as the non-circularity. This analysis allows us to have a better understanding of the behavior of the blind channel estimation when the considered side information is taken into account.

In this work, a novel subspace-based method for blind identification of multichannel finite impulse response systems is presented. Here, we exploit directly the block Toeplitz channel's structure in the signal's linear model to build a quadratic cost function, whose minimization leads to the desired channel estimation up to a scalar factor. This method can be extended to estimate any predefined linear structure, e.g., Hankel, that is usually encountered in linear systems. Simulation findings are provided to highlight the appealing advantages of the new structure-based subspace method over the standard subspace method in certain adverse identification scenarios.

Multimodulus algorithms (MMA) based adaptive blind equalizers mitigate inter-symbol interference in a digital communication system by minimizing dispersion in the quadrature components of the equalized sequence in a decoupled manner, i.e., the in-phase and quadrature components of the equalized sequence are used to minimize dispersion in the respective components of the received signal. These unsupervised equalizers are mostly incorporated in bandwidth-efficient digital receivers (wired, wireless or optical) which rely on quadrature amplitude modulation based signaling. These equalizers are equipped with nonlinear error-functions in their update expressions which makes it a challenging task to evaluate analytically their steady-state performance. However, exploiting variance relation theorem, researchers have recently been able to report approximate expressions for steady-state excess mean square error (EMSE) of such equalizers for noiseless but interfering environment. In this work, in contrast to existing results, we present exact steady-state tracking analysis of two multimodulus equalizers in a non-stationary environment. Specifically, we evaluate expressions for steady-state EMSE of two equalizers, namely the MMA2-2 and the βMMA. The accuracy of the derived analytical results is validated using different set experiments and found in close agreement.

This study addresses the ill‐conditioning problem of the memory polynomial (MP) model with application to the predistortion of highly non‐linear power amplifiers with memory effects. A resource‐efficient lattice‐based MP structure built using the cascade of a MP generator and a lattice predictor is proposed to overcome the ill‐conditioning of the MP's data matrix. The proposed model performances are benchmarked against those of the MP model as well as the orthogonal MP model. The experimental results demonstrate the suitability of the proposed predistorter as it achieves similar performance in the time and the frequency domains compared to the MP counterpart while alleviating its ill‐conditioning problem.

In this paper, we design and analyze a Newton-like blind equalization algorithm for the APSK system. Specifically, we exploit the principle of minimum entropy deconvolution and derive a blind equalization cost function for APSK signals and optimize it using Newton׳s method. We study and evaluate the steady-state excess mean square error performance of the proposed algorithm using the concept of energy conservation. Numerical results depict a significant performance enhancement for the proposed scheme over well established blind equalization algorithms. Further, the analytical excess mean square error of the proposed algorithm is verified with computer simulations and is found to be in good conformation.

In this work, the convergence and tracking behavior of the ε-normalized sign regressor least mean fourth (NSRLMF) algorithm are analyzed in the presence of white and correlated Gaussian data. Furthermore, the stability bound on the step-size of the ε-NSRLMF algorithm to ensure convergence in the mean, which also leads us to the mean convergence of the ε-normalized sign regressor least mean mixed-norm (NSRLMMN) algorithm is derived. Finally, simulation results are conducted to confirm the validity and performance of the proposed adaptive algorithm for both white and correlated Gaussian regressors.

Recently, distributed adaptive algorithms have been proposed to solve the problem of estimation over distributed networks. In a diffusion protocol, each node in the network function as an individual adaptive filter whose aim is to estimate the parameter of interest through local observations. All the estimates obtained from the nodes are then locally fused with their neighboring estimates in the network. The aim of this work is to improve the signal processing capability of the distributed network in a novel way by applying the diffusion normalized least mean squares (ε-NLMS) algorithm. The simulation results will show that the diffusion ε-NLMS algorithm outperforms its counterpart the diffusion least mean squares (LMS) algorithm for slowly changing environment, where data are expected to show high correlation.

In this work, a leakage‐based variant of the least mean fourth (LMF) algorithm, the leaky least mean fourth (LLMF) algorithm, is proposed. This algorithm will help mitigate the weight drift problem experienced in the conventional LMF algorithm. The main aim of this work is to derive the LLMF adaptive algorithm, analyse its convergence behaviour, and examine its performance in different noise environments. Furthermore, the tracking and transient analysis of the proposed LLMF algorithm are carried out using the energy‐conservation relation framework. Finally, a number of simulation results are carried out to corroborate the theoretical findings, and show improved performance obtained through the use of LLMF over the conventional LMF algorithm in a weight drift scenario.

In this work, the tracking analysis of the Normalized Least Mean Fourth (NLMF) algorithm is investigated for a random walk channel under very weak assumptions. The novelty of this work re sides in the fact that no restrictions are made on the dependence between the input successive regressors, the dependence among input regressor elements, the length of the adaptive filter, the distribution of noise and filter's input. Moreover, in our approach, there is no restriction made on the step size value and therefore the analysis holds for all the values of the step size in the range of stable NLMF algorithm. The analysis is based on a recently proposed performance measure called effective weight deviation vector which is the component of weight deviation vector in the direction of input regressor. In this paper, asymptotic time-averaged convergence for the mean square effective weight deviation, mean absolute excess estimation error, and the mean square excess estimation error for the NLMF algorithm are established. Finally, a number of simulation results are carried out to corroborate the theoretical findings.

The least-mean fourth (LMF) algorithm is best known for its fast convergence and low steady-state error especially in non-Gaussian noise environments. Recently, there has been a surge of interest in the LMF algorithm with different variants being proposed. The fact that different variable step-size least-mean square algorithms have shown to outperform its fixed step-size counterpart, a variable step-size least-mean fourth algorithm of the quotient form (VSSLMFQ) is proposed here. Therefore in this work, the proposed algorithm is analysed for its performance in the steady-state and it is shown to achieve a lower steady-state error then the traditional LMF algorithm. Finally, a number of computer simulations are carried out to substantiate the theoretical findings.

It is a well established fact that the addition of a constraint to an adaptive algorithm improves its performance properties. Consequently, in this work, a noise-constrained least mean fourth (NCLMF) adaptive algorithm is developed. The NCLMF algorithm is based on a constrained minimization problem that includes knowledge of the noise variance. Moreover, this noise constrained LMF algorithm can be seen as a variable-step-size LMF algorithm. The convergence analysis as well the tracking analysis of the NCLMF adaptive algorithm are developed using the concept of energy conservation. Finally, simulation results are presented to demonstrate the superiority of the NCLMF algorithm over the conventional LMF algorithm as well corroborating the theoretical findings.

This paper presents the design of a new diffusion algorithm over adaptive networks. The algorithm assumes knowledge of variance of additive noise. The design is based on the Noise-Constrained Least-Mean Squares (LMS) Algorithm and the new algorithm becomes a type of variable step-size algorithm for which the step-size variation rule results directly from the constraint. The design of the Noise-Constrained Diffusion LMS algorithm has been included. Simulation results show that the new algorithm outperforms the existing Diffusion LMS algorithm as well as its Incremental counterpart.

In this work, the steady-state analysis of the normalized least mean fourth (NLMF) algorithm under very weak assumptions is investigated. No restrictions are made on the dependence between input successive regressors, the dependence among input regressor elements, the length of the adaptive filter, the distribution of noise and the filter input. Moreover, in our approach, there is no restriction made on the step size value and therefore the analysis holds for all the values of the step size in the range where the NLMF algorithm is stable. The analysis is based on the effective weight deviation vector performance measure. This vector is the component of weight deviation vector in the direction of the input regressor. The asymptotic time-averaged convergence for the mean square effective weight deviation, the mean absolute excess estimation error, and the mean square excess estimation error for the NLMF algorithm are derived. Finally, a number of simulation results are carried out to corroborate the theoretical findings.

This work proposes a scalable architecture for implementing a mixed-norm LMS-LMF adaptive algorithm using a 16-bit fixed-point arithmetic representation. The hardware scalability allows flexibility in the choice of selecting the order of the filter without redesigning the hardware. The filter also allows flexibility in using application specific sampling frequencies.

In this work, a new lattice based algorithm for blind equalization is developed. The proposed algorithm is a generalized version of an existing lattice-based soft constraint satisfaction multi-modulus algorithm (L-SCS-MMA) and is named as lattice-based soft constraint satisfaction multi-modulus algorithm of order p (L-SCS-MMA-p). Moreover, a slicer output has been incorparated in the blind estimate of the SCS-MMA algorithm to enhance more the convergence speed of the the algorithm, and the resulting algorithm is named as (LDA-SCS-MMA-p). The performance of the proposed algorithm is investigated under different scenarios.

Abstract In this work, we introduce a new chip‐level linear modified‐SIC multi‐user structure that is asymptotically equivalent to successive over‐relaxation (SOR) iteration, which is known to outperform the conventional Gauss–Seidel iteration by an order of magnitude in terms of convergence speed. The main advantage of this scheme is that it uses directly the spreading codes and not the cross‐correlation coefficients and thus reduces significantly the overall computational complexity. This is critical for the design of low‐complexity multiuser detectors for long‐code CDMA systems such as IS95 and UMTS. We use a matrix algebraic approach to show the equivalence of the proposed scheme to linear matrix filtering. This allows obtaining an analytical expression for both the bit‐error rate (BER) and the asymptotic multiuser efficiency (AME). Moreover, we study the convergence behavior of the proposed scheme and prove that it converges if the relaxation factor is within the interval ]0, 2[. Simulation results are in excellent agreement with theory. Copyright © 2008 John Wiley & Sons, Ltd.

Abstract In this work, we present a new multiuser detection (MUD) structure which results from incorporating the tabu search (TS) heuristic algorithm with the local search (LS) heuristic algorithm. The new proposed structure brings much improvement when compared to both the conventional (matched filter) detector and the decorrelating detector. Moreover, the new TS‐LS‐detector proposed here approximates well the performance of the optimal MUD detector but with a very low computational complexity. Copyright © 2008 John Wiley & Sons, Ltd.

In this work, a statistical noise-constrained least mean fourth (SN CLMF) adaptive algorithm is proposed. Based on the fact that in many practical applications an accurate estimate of the fourth- order moment of the noise is available, or can be easily estimated, the learning speed of the LMF algorithm can be then increased considerably by adding a constraint to it. This noise constrained LMF algorithm can be seen as a variable step-size LMF algorithm. Moreover, the concept of energy conservation is used to carry out the rigorous steady-state analysis. Finally, a number of simulations are carried out to corroborate the theoretical findings, and as expected, improved performance is obtained through the use of this technique over the traditional LMF algorithm.

In digital communication, adaptive channel equalization techniques underpin the successful provision of high speed and reliable data transmission over severely-dispersive channels, e.g. wireless and mobile ones. In a real world that is largely dominated by non- Gaussian interference signals, adaptive equalizers relying heavily on the LMS are bound to yield suboptimal performances. This work addresses this sub-optimality issue by proposing a new adaptive equalizer which judiciously combines the power of the least- mean fourth (IMF) algorithm to better tackle non-Gaussian environments, and the capability of the power-of-two quantizer (PTQ) to greatly reduce the IMF's high computational load. This combination endows the proposed algorithm with a capability of tracking fast-changing channels. A performance analysis of the proposed adaptive channel equalizer, based on a new linear approximation of the PTQ, is also presented. Extensive simulation testing of the proposed adaptive equalizer corroborates very well the theoretical findings predicted by the analysis of the linearized LMF- PTQ algorithm.

In this paper, the recently proposed linear group-wise parallel interference cancellation (LGPIC) detector is identified as a preconditioned block Richardson iterative method. This facilitates considerably the analysis of the convergence behavior of different group-detection schemes. Moreover, by using this approach, the effect of grouping is easily studied as well. Finally, simulation results conducted are in excellent agreement with the theory.

Abstract In this paper, we consider a linear group polynomial expansion successive interference cancellation (GPE‐SIC) detector in a synchronous CDMA system. It is a hybrid detector, which combines parallel and successive cancellation techniques in order to extract the advantages of both schemes. We use the fact that even if the cross‐correlation matrix of the system is not diagonal dominant, the sub‐matrices corresponding to different groups can be forced to be diagonal dominant by suitable grouping of users to approximate the decorrelator/MMSE detector by a low‐complexity polynomial expansion detector. The latter is computationally very efficient if the cross‐correlation matrix of users within the same group is diagonal dominant. Simulation results showed that the (GPE‐SIC) detector has the same performance as the linear group decorrelator successive interference cancellation (GDEC‐SIC) detector but with low‐computational complexity. Copyright © 2006 John Wiley & Sons, Ltd.

In this paper we propose application of heuristic algorithms in multiuser detection (MUD). The proposed algorithm combines a tabu search heuristic algorithm with a local search heuristic algorithm. The new proposed structure brings several improvements when compared to both the conventional (matched filter) detector and the decorrelating detector. Additionally, the algorithm proposed here fairly approximates the performance of the optimal MUD detector with much reduced computational complexity.

Multiple access interference and near-far effect cause performance limitation in the conventional single-user detector used in direct sequence/code division multiple access (DS-CDMA)-systems. In this paper, we present a novel multiuser detector (MUD) based on the new heuristic algorithm known as particle swarm algorithm (MUDPSO). We evaluate the BER performance of the proposed algorithm, by means of Monte Carlo simulation technique, and compare it to the BER performances of both the matched filter detector and the decorrelator multiuser detectors. We show that the new algorithm outperforms the other two. Moreover, the performance under near-far scenario, the system capacity and the computational complexity of the proposed detector are also investigated

In this letter, a new multimodulus algorithm for blind equalization of complex communication channels is derived by solving a constrained optimization problem with relaxation. The intersymbol interference (ISI) optimization and phase-recovery capabilities of the proposed algorithm are analyzed. It is shown from computer simulations that superior performance for the derived algorithm over Lin's algorithm is obtained.

In this work, a new algorithm, based on the minimum-disturbance principle with relaxation, is presented for the blind equalization of complex signals. This algorithm combines the benefits of the well-known reduced constellation algorithm (RCA) and constant modulus algorithm (CMA). The convergence characteristics of the proposed algorithm are demonstrated by way of simulations. In addition, closed form expressions are obtained for the statistical (dispersion) constants used in these algorithms.

A new linear hybrid multiuser detection scheme is proposed. The structure cascades a linear successive interference cancellation detector and a linear parallel interference cancellation detector in order to extract the advantages of both. The convergence behavior of the hybrid scheme is investigated and moreover, the condition of convergence is determined. The proposed multiuser detector is compared to the other multiuser detectors in terms of complexity, delay, and other factors. Simulation results illustrate the superiority of the proposed.

The least mean fourth (LMF) algorithm is known for its fast convergence and lower steady state error, especially under sub-Gaussian noise conditions. Meanwhile, the recent work on the normalised versions of LMF algorithm has further enhanced its stability and performance in both Gaussian and sub-Gaussian noise. For example, the normalised LMF (XE-NLMF) algorithm, recently developed, is normalised by the mixed signal power and error power, and weighted by a fixed mixed-power parameter. Unfortunately, this algorithm depends on the selection of this mixing parameter. To overcome this obstacle, in this work, a time-varying mixed-power parameter technique is introduced to optimise its selection. An enhancement in performance is obtained through the use of this procedure in both the convergence rate and steady-state error.

Tracking analysis of normalized adaptive algorithms is carried out in the presence of two sources of nonstationarities: carrier frequency offset between transmitter and receiver; random variations in the environment. A unified approach is carried out using a mixed-norm-type error nonlinearity. Close agreement between analytical analysis and simulation results is obtained for the case of the NLMS algorithm. The results show that, unlike the stationary case, the steady-state excess-mean-square error is not a monotonically increasing function of the step-size, while the ability of the adaptive algorithm to track the variations in the environment degrades by increasing the frequency offset.

In this work, we consider a linear group polynomial expansion successive interference cancellation (GPE-SIC) detector in a synchronous CDMA system. It is a hybrid detector, which combines parallel and successive cancellation techniques in order to extract the advantages of both of the schemes. Benefiting from the fact that even if the cross-correlation matrix of the system is not diagonal-dominant, we can force the cross-correlation matrix of users within the same group to be diagonal-dominant by suitable grouping and approximate the decorrelator/MMSE detector by a low-complexity polynomial expansion detector. This approximation is very accurate if the cross-correlation matrix of users within the same group is diagonal-dominant. Simulation results showed that the (GPE-SIC) detector has the same performance as the linear group decorrelator successive interference cancellation (GDEC-SIC) detector but with lower computational complexity.

Tracking analysis of the normalized least mean square (NLMS) algorithm is carried out in the presence of two sources of nonstationarities: 1) carrier frequency offset between transmitter and receiver; 2) random variations in the environment. A novel approach to this analysis is carried out using the concept of energy conservation. Close agreement between analytical analysis and simulation results is obtained. The results show that, unlike in the stationary case, the steady-state excess MSE is not a monotonically increasing function of the step size. Moreover, the ability of the adaptive algorithm to track the variations in the environment is shown to degrade with increasing frequency offset.

The convergence analysis of the variable weight mixed-norm LMS-LMF adaptive algorithm is derived. A novel approach is used to study the convergence behavior of three algorithms: mixed-norm- LMS- and LMF-based ones. As a by-product of this novel approach, more general and new necessary and sufficient conditions and excess steady-state error for the LMF have been derived.

The idea of multi-split adaptive filtering is applied to the DFE. The feedforward and feedback sections in the DFE are divided into parallel sub-filters by imposing separately symmetry and antisymmetry conditions on the impulse responses of the filters by using appropriate and distinct sets of linear constraints. The purpose of this is to obtain a new structure that gives better performance than the conventional structure. The performance of multi-split DFE is tested with different algorithms (LMS, RLS, DCT-LMS) and compared to that of a multi-split linear transversal equalizer.

Stop-and-go decision-directed (S-and-G-DD) equalization is the most primitive blind equalization (BE) method for the cancelling of intersymbol-interference in data communication systems. Recently, this scheme has been applied to complex-valued multilayer feedforward neural network, giving robust results with a lower mean-square error at the expense of slow convergence. To overcome this problem, in this work, a fast converging recursive least squares (RLS)-based complex-valued backpropagation learning algorithm is derived for S-and-G-DD blind equalization. Simulation results show the effectiveness of the proposed algorithm in terms of initial convergence.

The severely distorting channels limit the use of linear equalizers and the use of the nonlinear equalizers then becomes justifiable. Neural-network-based equalizers, especially the multilayer perceptron (MLP)-based equalizers, are computationally efficient alternative to currently used nonlinear filter realizations, e.g., the Volterra type. The drawback of the MLP-based equalizers is, however, their slow rate of convergence, which limit their use in practical systems. In this work, the effect of whitening the input data in a multilayer perceptron-based decision feedback equalizer (DFE) is evaluated. It is shown from computer simulations that whitening the received data employing adaptive lattice channel equalization algorithms improves the convergence rate and bit error rate performances of multilayer perceptron-based DFE. The adaptive lattice algorithm is a modification to the one developed by Ling and Proakis (1985). The consistency in performance is observed in both time-invariant and time-varying channels. Finally, it is found in this work that, for time-invariant channels, the MLP DFE outperforms the least mean squares (LMS)-based DFE. However, for time-varying channels comparable performance is obtained for the two configurations.

The demand for high speed wireless systems dictates the use of bandwidth efficient modulation schemes. Trellis coded I-Q 16-QAM is shown here to be suitable for such applications. The performance of this scheme with decision feedback equalization is evaluated over a two-ray Rayleigh channel. Comparisons between the I-Q scheme with a conventional TCM scheme based on 8-PSK show coding gains of about 4 dB at a BER of 10/sup -3/.

This paper presents a unifying view of various error nonlinearities that are used in least mean square (LMS) adaptation such as the least mean fourth (LMF) algorithm and its family and the least-mean mixed-norm algorithm. Specifically, it is shown that the LMS algorithm and its error-modified variants are approximations of two previously developed optimum nonlinearities which are expressed in terms of the additive noise probability density function (PDF). This is demonstrated through an approximation of the optimum nonlinearities by expanding the noise PDF in a Gram-Charlier series. Thus a link is established between intuitively proposed and theoretically justified variants of the LMS algorithm. The approximation has also a practical advantage in that it provides a trade-off between simplicity and more accurate realization of the optimum nonlinearities.

The coefficients of an echo canceller with a near-end section and a far-end section are usually updated with the same updating scheme, such as the LMS algorithm. In this paper we propose a novel scheme for echo cancellation that is based on the minimization of two different cost functions, i.e., one for the near-end section and a different one for the far-end section. Two approaches are addressed and only one of them leads to a substantial improvement in performance over the LMS algorithm when it is applied to both sections of the echo canceller. The proposed scheme is also shown to be robust to noise variations, which is not the case for the LMS algorithm.