International audience

In the Wyner-Ziv source coding problem, a source X has to be encoded while the decoder has access to side information Y. This paper investigates the indirect setup, in which a latent source S, unobserved by both the encoder and the decoder, must also be reconstructed at the decoder. This scenario is increasingly relevant in the context of goal-oriented communications, where S can represent semantic information obtained from X. This paper derives the indirect Wyner-Ziv rate-distortion function in asymptotic regime and provides an achievable region in finite block-length. Furthermore, a Blahut-Arimoto algorithm tailored for the indirect Wyner-Ziv setup, is proposed. This algorithm is then used to give a numerical evaluation of the achievable indirect rate-distortion region when S is treated as a classification label.

In the framework of goal-oriented communications, this paper investigates parametric regression over coded data. For this problem, information-theoretic bounds are provided in terms of rate versus regression generalization error, by considering quantize and binning achievability schemes. Alternatively, this paper focuses on practical implementations by proposing a coding scheme that combines a scalar quantizer with a non-binary LDPC code for the binning part. Given that the LDPC decoder requires prior knowledge of the regression parameters, the paper introduces a novel method to estimate these parameters directly over the LDPC-coded syndrome, without the need for prior decoding. This technique allows to both address the regression task and initialize the LDPC decoder for further data reconstruction. Monte-Carlo simulations show the efficiency of the proposed approach in terms of regression generalization error.

For finite blocklength polar codes, the minimum distance and the number of low weight codewords are essential to obtain good performance under successive cancellation list decoding with moderate and high list sizes. In this paper, we propose a code design method to decrease the number of low weight codewords for some information lengths with a very low computational complexity. In the proposed method, some information bits are encoded by several rows of the polar encoding matrix, i.e., each of the dynamic frozen bits is chosen the same as one of the preceding information bits. The dynamic frozen bit index set is determined by using the connection between the binary representation of the row indices and the number of common 1-bit positions of any given rows. The resulting design is shown to perform as well as polarization-adjusted-convolutional codes [9] under successive cancellation list decoding but with significant computational complexity savings. These findings pave the way for the use of polar codes in applications with stringent complexity and with low energy consumption constraints.

Reed Muller (RM) codes are known for their good minimum distance. One can use their structure to construct polar-like codes with good distance properties by choosing the information set as the rows of the polarization matrix with the highest Hamming weight, instead of the most reliable synthetic channels. However, the information length options of RM codes are quite limited due to their specific structure. In this work, we present sufficient conditions to increase the information length by at least one bit for some underlying RM codes and in order to obtain pre-transformed polar-like codes with the same minimum distance than lower rate codes. Moreover, our findings are combined with the method presented in [1] to further reduce the number of minimum weight codewords. Numerical results show that the designed codes perform close to the meta-converse bound at short blocklengths and better than the polarized adjusted convolutional polar codes with the same parameters.

In this work, we give a complete characterization of the $\epsilon$-stable region in dynamic downlink random cellular networks. The $\epsilon$-stable region is the set of arrival rates such that the proportion of unstable queues in the network is not larger than $\epsilon$. We derive upper and lower bounds as well as an approximation of the critical arrival rate, which delimits the $\epsilon-$stable region. The developed model is based on stochastic geometry and queuing theory to handle the interaction between the transmit success probability and the queuing state evolution. Extensive numerical simulations are provided to confirm the tightness of the approximation.

Impulsive noise, where large amplitudes arise with a relatively high probability, arises in many communication systems including interference in Low Power Wide Area Networks. A challenge in coping with impulsive noise, particularly alpha-stable models, is that tractable expressions for the log-likelihood ratio (LLR) are not available, which has a large impact on soft-input decoding schemes, e.g., low-density parity-check (LDPC) packets. On the other hand, constraints on packet length also mean that pilot signals are not available resulting in non-trivial approximation and parameter estimation problems for the LLR. In this paper, a new unsupervised parameter estimation algorithm is proposed for LLR approximation. In terms of the frame error rate (FER), this algorithm is shown to significantly outperform existing unsupervised estimation methods for short LDPC packets (on the order of 500 symbols), with nearly the same performance as when the parameters are perfectly known. The performance is also compared with an upper bound on the information-theoretic limit for the FER, which suggests that in impulsive noise further improvements require the use of an alternative code structure other than LDPC.

In this paper, a real-valued function that approximates the cumulative distribution function (CDF) of a finite sum of real-valued independent and identically distributed random vectors is presented. The approximation error is upper bounded by an expression that is easy to calculate. As a byproduct, an upper bound and a lower bound on the CDF are obtained. Finally, in the case of lattice and absolutely continuous random variables, the proposed approximation is shown to be identical to the saddlepoint approximation of the CDF.

In this paper, we analyze the impact of reducing the Peak to Average Power Ratio (PAPR) of a time domain Orthogonal Frequency Division Multiplexing (OFDM) signal before its compression, as required for the transmission of IQ samples in a Centralized Radio Access Network architecture. The analysis is performed considering a clipping function for PAPR reduction. First, the distribution of the clipped OFDM IQ samples is obtained which in turn allows to derive the threshold and codebook levels of a non-uniform quantizer optimized for clipped IQ samples and the number of representation bits for each quantized levels in the entropy coding stage. Then, we provide a closed-form expression for the Modulation Error Ratio of the compressed and clipped OFDM signal. The simulation results show that the Gaussian quantizer applied to a clipped signal is not quite robust compared to a quantizer that has been optimized w.r.t. the distribution of clipped IQ samples, especially in the low-rate quantization region.

In this work, we proposed a tractable mathematical framework to analyze the coverage probability in dynamic downlink cellular networks taking into account the queue dynamics with finite buffer restriction. In particular, the developed model is based on stochastic geometry and queueing theory to handle the interaction between the coverage probability and the queueing state evolution. We analyze the coverage probability as well as packet loss probability by discrete time Markov chain (DTMC) at stationary regime. We explicitly derive the influence of the packet buffer length to the coverage and packet loss probability. We show in particular that small buffer length leads to a better coverage probability but also to a larger packet loss probability, advocating for a tradeoff between these two metrics.

Inspired by the subject of performance enhancement of users located at the edge of cells in cellular networks, this paper studies a centralized frequency reuse scheme depending on the link quality of users within the Voronoi cell of the serving base station (BS) for the homogeneous Poisson point process (PPP) network. To quantify how much a user experiencing performance degradation, we consider that each cell is partitioned into multiple regions based on the strength of the signal to interference ratio. In our model, a given resource block (RB) is allocated to one user in a cell and cannot be shared by another user in this cell. The RB is divided into multiple subchannels and the user of interest uses only one of the multiple sub-channels, the mapping depending on the region in which the user belongs, letting a part of the RB unused to reduce the interference to the other cells. This scenario implies that the interfering set of BS depends on the coverage probability of the typical user. We tightly approximate the interfering BS set as a thinned version of the original PPP depending on the probability of coverage of each type of user. Results show that the scheme increases the network's global coverage probability and enhances spectral efficiency by comparing the known fractional frequency reuse scheme.

This paper introduces an upper-bound on the absolute difference between: (a) the cumulative distribution function (c.d. f.) of the sum of a finite number of independent and identically distributed (i.i.d) random variables; and (b) a saddlepoint approximation of such c.d.f. This upperbound is general and particularly precise in the regime of large deviations. This result is used to study the dependence testing (DT) bound on the minimum decoding error probability (DEP) in memoryless channels. Within this context, the main results include new lower and upper bounds on the DT bound. As a byproduct, an upper bound on the absolute difference between the exact value of the DT bound and its saddlepoint approximation is obtained. Numerical analysis of these bounds are presented for the case of the binary symmetric channel and the additive white Gaussian noise channel, in which the new bounds are observed to be tight.

Multi-user shared access (MUSA) is one of the most promising non orthogonal multiple access (NOMA) schemes to accommodate the stringent requirements of massive machine type communications (mMTC). MUSA adopts a grant-free access strategy and uses non-orthogonal spreading sequences to accommodate a large number of users in the same radio resources. The inter-user interference is subsequently handled using successive interference cancellation (SIC) algorithm, which allows a low decoding complexity. However, sequential decoding can be affected by error propagation when the power difference between simultaneous transmissions is not significant. In this paper, we propose a power allocation algorithm in order to avoid the error propagation problem by guaranteeing sufficient power difference between concurrent received signals. We formulate an optimization problem which aims at minimizing the system average bit error rate. We show that the optimization problem has an optimal global solution and an algorithm is proposed for its resolution. Simulation results show the impact of performing power allocation for SIC decoding on the bit error rate.

This paper aims at characterizing the energy efficiency (EE)-area spectral efficiency (ASE) tradeoff in random geometry networks with multiple-antenna arrays at base stations (BSs). In particular, ASE and EE are studied with respect to the transmit power when BSs use a signal-to-leakage-plus-noise ratio (SLNR) precoder. When the static power consumption cannot be neglected, EE behaves linearly with respect to ASE before a sharp decreasing of EE due to the interference-limited characteristic of the network. Our contribution relies on the derivation of a closed-form expression for ASE with SLNR precoder in the asymptotic regime, i.e., when the number of antennas and users grows to infinity, using stochastic geometry. We derive EE from a linear power consumption model afterwards. Unlike conventional SLNR precoders, the average signal-to-interference-plus-noise ratio and the leakage to other cells are considered in a geometry-dependent network. Extensive Monte Carlo simulations show that despite the asymptotic nature of the theoretical analysis, the closed-form expressions are tight with respect to simulations even for the moderate number of antennas and users. Hence, the analysis can be used for realistic network performance analysis.

This paper deals with the energy-efficiency spectralefficiency (EE-SE) tradeoff in regular hexagonal interference-limited networks. We aim to characterize the EE-SE region when non negligible static circuitry power consumption is assumed and when several frequency reuse factors as well as shadowing are taken into account. Our contribution is three-fold: a parametric expression of the interference to signal ratio, a.k.a. fparameter, obtained with a curve fitting approach is proposed, the EE-SE tradeoff is then studied when several frequency reuse factors are considered and a new characterization, called the c-EE-SE tradeoff, is proposed when shadowing is considered. This definition relies on the outage capacity achievable and its corresponding EE and remains particularly useful when longterm fading is present.

Time-Reversal (TR) precoding schemes are regarded as promising candidates for robust and energy efficient wireless systems. These schemes are characterized by their low computational complexity at the receiver side. In multipath Rayleigh fading channel, this paper presents analytical expressions of the average bit error rate (BER) of TR technique and the Equal Gain Transmission (EGT) scheme in MISO-OFDM systems. It further provides simple and tight upper bounds on the ergodic capacity of these schemes. Hence, these theoretical analyses might be very useful to avoid lengthy simulations. It also provides an interesting conclusion: for large number of transmit antennas, i.e. in Massive MIMO, TR technique outperforms the EGT scheme and its performance approaches that of Maximum Ratio Transmission (MRT) scheme. Hence, it confirms the applicability of TR scheme in the future Massive MIMO systems. Results with LTE-A system parameters are provided in order to corroborate our analysis.

In this paper, we introduce network coding (NC) at intermediate nodes of a linear network where coding is conducted on address correlated packets. In particular, we propose a novel algorithm for decoding network-coded messages that enhances the bandwidth usage and reduces the resource allocation by allowing nodes in the network to perform distributed decoding. We show that, compared to traditional decoding at end nodes, distributed decoding allows reduction not only in the number of transmitted bytes but also in the number of resources needed to perform NC. Our strategy is compared to decoding at end nodes and experiments show a reduction of 64% in the number of transmitted bytes and a reduction of 93% of buffering time when using distributed decoding to exchange 1000 packets between end nodes of a linear network of 8 nodes.

The orbital angular momentum (OAM) is a physical property of an electromagnetic (EM) radiation, distinct from polarization, carrying a mechanical moment. In radio frequency range, OAM can be viewed as a phase gradient on the wave front. In this paper, the problem of the detection of the orbital angular momentum (OAM) order (or topological charge) for electromagnetic (EM) radiation is addressed. The OAM order detection problem is first presented as a multiple hypothesis testing problem. The maximum likelihood test leading to the error probability on the OAM order detection in additive white Gaussian noise (AWGN) channel is derived when the transmitting and receiving arrays are perfectly aligned and misaligned. A simpler receiver based on fast fourier transform (FFT) is then presented. The performances of this receiver are then presented through extensive simulation runs for various degrees of misalignment between the transmitter and the receiver.

In this paper, we analyze the problem of deploying small cellular cells into a large broadcast cell in the framework of the convergence between the Long Term Evolution (LTE) and the Digital Broadcasting Video (DVB) technologies. In this work, a non- cooperative deployment of these technologies is considered in the sense that cellular and broadcast networks interfere on each other. In literature, there is no work dealing with the ergodic capacity of full cross-interfering links in a DVB-LTE convergence scenario. Our contribution lies in two folds: i) the global ergodic capacity is derived in closed-form with Rayleigh fading channel on each link. This derivation allows us to evaluate the global data rate of the DVB and LTE in a spectral overlap scenario. ii) The analytical forms allow the evaluation of the influence of system parameters on the global and broadcast data rate, e. g. mobile and broadcast power, distance between the broadcast receiver and the LTE base station.

In this paper, we consider the performance of coded wireless communication systems experiencing non-frequency selective fading channels in shadowed environments. The quality of service (QoS) in a wireless network is dependent on the packet error outage (PEO). We address the problem of finding a tractable expression for the coded PEO over Nakagami-m channels with shadowing, considering multilevel modulations, various block and convolutional channel coding schemes and hard decision decoding. In order to obtain the coded PEO, an inversion of the coded packet error probability (PEP) w.r.t. the signal to noise ratio (SNR) is needed. To this end, we propose an invertible approximation for the coded PEP w.r.t. the uncoded bit error probability (BEP) in Nakagami-m fading channels which is accurate for all BEPs of interest. The BEP itself depends on the average SNR and we hence make use of previous results on the inversion of the uncoded BEP w.r.t. the SNR in Nakagami-m fading channels, holding for M-PSK and M-QAM signals. We were thus able to obtain a reliable closed form expression for the coded PEO in flat fading and shadowing channels. The theoretical findings are verified by means of simulations.

This paper studies the energy efficiency (bits/joule) in relay assisted networks using Chase combining hybrid-ARQ (HARQ-CC) retransmission technique. As energy efficiency depends on the average delay and the packet error probability (PER), we start by deriving an approximation of the PER for HARQ-CC retransmission scheme. This approximation is really simple compared to the existing expressions in the literature and is proved to be tight enough to address the energy efficiency study. In particular, the energy efficiency is investigated in terms of destination position and in terms of transmission power of the source and the relay. The results give some insights on the optimal transmission power according to the maximum number of retransmissions maximizing the energy efficiency.

In this paper, we are interested in resource allocation strategies for wireless time-hopping ultra-wide band (TH-UWB) communications with multiple rate capabilities between users. Multiple rates are achieved by assigning different processing gains, i.e. Nf, to users. For this purpose, the multiple-access interference (MAI) variance accounting for multi-rate is needed. It is a challenging task due to the lack of a suitable closed-form expression for the MAI variance in a multi-rate context. We further study the multi-rate resource allocation problem in uplink TH-UWB systems for which an optimal search cannot be envisaged due to the exponential complexity induced. Our contribution lies in three-fold: i) A new intercode correlation expression accounting for multi-rate communications is derived, and the variance of the MAI averaging over the codes is obtained. ii) The multi-rate resource allocation problem is tackled by relaxing the integer constraint on the processing gains and modeled via a signomial programming problem. iii) Based on this, a branch and bound (BB) algorithm is derived for the allocation of the processing gains in TH-UWB systems. We also propose a really simple heuristic with linear complexity for the Nfallocation. We show that the algorithm proposed outperforms the BB algorithm in average throughput and average starvation rate.

In this paper, an efficient rate allocation scheme to maximize the throughput in uplink multi-rate time-hopping ultra-wideband (TH-UWB) communications is proposed. This is a challenging task due to the lack of a suitable closed form expression of the multiuser interference in a multi-rate context. Moreover, an exhaustive search for the optimal solution would be too prohibitive due to the high complexity induced. Our contribution can be summarized as follows: i) A new expression of the multiuser interference variance accounting for multi-rate communications is proposed. ii) Thanks to this, a computational efficient multi-rate allocation scheme is proposed. The algorithm allows to achieve high throughput while keeping a relatively low starvation rate and it outperforms a Max-Min algorithm with a fix rate for all users. Moreover, the simulations based on the theoretical analysis give insights on the influence of the different UWB parameters on the global throughput.

We deal with the analytical performance study of orthogonal space-time block codes (OSTBC) applied to multiple-input multiple-output (MIMO) systems experiencing non-frequency selective fast fading and shadowing. We address the problem of finding a tractable closed-form expression for the symbol error outage (SEO) over Rice fading channels in shadowing environments. In order to obtain the SEO, a symbol error probability (SEP) inversion w.r.t. the average signal to noise ratio (ASNR) is needed since no invertible SEP approximations are available in literature. We propose an accurate approximation for the SEP in Rice fading channels. This new approximation is proved to be invertible w.r.t. the ASNR and this expression thus facilitates the derivation of a tight approximation of the SEO in lognormal shadowing environments.

This letter addresses the problem of finding a tractable expression for the symbol error outage (SEO) in flat Nakagami-m fading and shadowing channels. We deal with M-ary phase shift keying (M-PSK) and quadrature amplitude modulation (M-QAM) which extends our previous results on BPSK signaling. We propose a new tight approximation of the symbol error probability (SEP) holding for M-PSK and M-QAM signals which is accurate over all signal to noise ratios (SNRs) of interest. We derive a new generic expression for the inverse SEP which facilitates the derivation of a tight approximation of the SEO in a lognormal shadowing environment.

The roll-out density of wireless local area networks (WLANs) has recently witnessed a dramatic increase and is currently reaching saturation levels. The frequency bands designated to WLANs does not suffice any more to provide non-overlapping, and hence interference-free, communication bands. A large body of research has been dedicated to a wide variety of optimum maximum likelihood sequence estimation (MLSE) and sub-optimum in-band interference mitigation techniques. Our contribution lies in a reduction of the state-space of a MLSE detector in the case of a desired WLAN receiver experiencing delayed interference from some other transmitters operating in partially overlapping spectral bands and over independent frequency-selecting block-fading channels. Based on the formulation of the optimum receiver, we derive a sub-optimum receiver of reduced complexity and demonstrate its satisfactory performance in the context of strong interference.

The deployment of wireless local area networks (WLANs) has recently gained in popularity with roll-out densities often reaching saturation levels. The frequency bands designated to WLANs do thus not suffice anymore to provide non-overlapping, and hence interference-free, communication. Prior research has therefore been dedicated to a wide variety of mainly cochannel interference mitigation techniques. Our contribution lies in the performance analysis of the more realistic case of spatially close WLAN systems interfering asynchronously in DSSS mode and on adjacent channels. The authors expose the performance of optimum and sub-optimum multiuser detection maximum likelihood sequence estimators (MUD-MLSEs) under varying conditions, e.g. level of interference, degree of spectral overlap, etc. This facilitates a better understanding and design of future interference-limited WLAN systems. The performance results are also important to spectrum policy makers, allowing to knowledgeably estimate the degree of tolerated interference in license exempt bands.