Generative Adversarial Networks (GANs) have been widely used for generating synthetic data for cases where there is a limited size real-world data set or when data holders are unwilling to share their data samples. Recent works showed that GANs, due to overfitting and memorization, might leak information regarding their training data samples. This makes GANs vulnerable to Membership Inference Attacks (MIAs). Several defense strategies have been proposed in the literature to mitigate this privacy issue. Unfortunately, defense strategies based on differential privacy are proven to reduce extensively the quality of the synthetic data points. On the other hand, more recent frameworks such as PrivGAN and PAR-GAN are not suitable for small-size training data sets. In the present work, the overfitting in GANs is studied in terms of the discriminator, and a more general measure of overfitting based on the Bhattacharyya coefficient is defined. Then, inspired by Fano’s inequality, our first defense mechanism against MIAs is proposed. This framework, which requires only a simple modification in the loss function of GANs, is referred to as the maximum entropy GAN or MEGAN and significantly improves the robustness of GANs to MIAs. As a second defense strategy, a more heuristic model based on minimizing the information leaked from the generated samples about the training data points is presented. This approach is referred to as mutual information minimization GAN (MIMGAN) and uses a variational representation of the mutual information to minimize the information that a synthetic sample might leak about the whole training data set. Applying the proposed frameworks to some commonly used data sets against state-of-the-art MIAs reveals that the proposed methods can reduce the accuracy of the adversaries to the level of random guessing accuracy with a small reduction in the quality of the synthetic data samples.

Cardiotocography (CTG) plays a vital role in fetal well-being monitoring by tracking fetal heart rate (FHR) and uterine contractions (UC). However, CTG interpretation suffers from subjectivity, resulting in low agreement among observers and potentially unnecessary medical interventions. Existing AI-based diagnostic models struggle with generalization and are only effective on distinct CTG samples. This study introduces a novel approach employing deep semi-supervised learning for anomaly detection in CTG signals, marking the first attempt in this direction. A modified GANomaly model is proposed, trained on normal CTG data, and evaluated for abnormality detection. This model employs an encoder-decoder structure with a discriminator, minimizing reconstruction and latent space errors while learning the normal CTG signal distribution. Leveraging the CTU-UHB dataset, our model demonstrates superior performance compared to existing methods during inference.

This paper adopts Arimoto's$\alpha$-Mutual Information as a tunable privacy measure, in a privacy-preserving data release setting that aims to prevent disclosing private data to adversaries. By fine-tuning the privacy metric, we demonstrate that our approach yields superior models that effectively thwart attackers across various performance dimensions. We formulate a general distortion-based mechanism that manipulates the original data to offer privacy protection. The distortion metrics are determined according to the data structure of a specific experiment. We confront the problem expressed in the formulation by employing a general adversarial deep learning framework that consists of a releaser and an adversary, trained with opposite goals. This study conducts empirical experiments on images and time-series data to verify the functionality of$\alpha$-Mutual Information. We evaluate the privacy-utility trade-off of customized models and compare them to mutual information as the baseline measure. Finally, we analyze the consequence of an attacker's access to side information about private data and witness that adapting the privacy measure results in a more refined model than the state-of-the-art in terms of resiliency against side information.

Adversarial robustness has become a topic of growing interest in machine learning since it was observed that neural networks tend to be brittle. We propose an information-geometric formulation of adversarial defense and introduce Fire, a new Fisher-Rao regularization for the categorical cross-entropy loss, which is based on the geodesic distance between the softmax outputs corresponding to natural and perturbed input features. Based on the information-geometric properties of the class of softmax distributions, we derive an explicit characterization of the Fisher-Rao Distance (FRD) for the binary and multiclass cases, and draw some interesting properties as well as connections with standard regularization metrics. Furthermore, we verify on a simple linear and Gaussian model, that all Pareto-optimal points in the accuracy-robustness region can be reached by Fire while other state-of-the-art methods fail. Empirically, we evaluate the performance of various classifiers trained with the proposed loss on standard datasets, showing up to a simultaneous 1% of improvement in terms of clean and robust performances while reducing the training time by 20% over the best-performing methods.

A privacy-preserving adversarial network (PPAN) was recently proposed as an information-theoretical framework to address the issue of privacy in data sharing. The main idea of this model was to use mutual information as the privacy measure and adversarial training of two deep neural networks, one as the mechanism and another as the adversary. The performance of the PPAN model for the discrete synthetic data, MNIST handwritten digits, and continuous Gaussian data was evaluated and compared to the analytically optimal trade-off. In this study, we evaluate the PPAN model for continuous non-Gaussian data, i.e. smart meters data, where lower and upper bounds of the privacy-preserving problem are used. These bounds include the Kraskov (KSG) estimation of entropy and mutual information that is based on the k-th nearest neighbor. In addition to the synthetic data sets, a practical case for hiding the actual electricity consumption from smart meter readings is examined. The results show that for continuous non-Gaussian data, the PPAN model performs within the determined optimal ranges and close to the lower bound.

The rapid development and expansion of the Internet of Things (IoT) paradigm has drastically increased the collection and exchange of data between sensors and systems, a phenomenon that raises serious privacy concerns in some domains. In particular, smart meters (SMs) share fine-grained electricity consumption of households with utility providers that can potentially violate users’ privacy as sensitive information is leaked through the data. In order to enhance privacy, electricity consumers can exploit the availability of physical resources such as a rechargeable battery (RB) to shape their power demand as dictated by a privacy-cost management unit (PCMU). In this article, we present a novel method to learn the PCMU policy using deep reinforcement learning (DRL). We adopt the mutual information (MI) between the user’s demand load and the masked load seen by the power grid as a reliable and general privacy measure. Unlike previous studies, we model the whole temporal correlation in the data to learn the MI in its general form and use a neural network to estimate the MI-based reward signal to guide the PCMU learning process. This approach is combined with a model-free DRL algorithm known as the deep double${Q}$-learning (DDQL) method. The performance of the complete DDQL-MI algorithm is assessed empirically using an actual SMs data set and compared with simpler privacy measures. Our results show significant improvements over state-of-the-art privacy-aware demand shaping methods.

The weighted-type fractional Fourier transform (WFRFT)-based hybrid carrier (HC) system has been proposed to bridge the gap between orthogonal frequency division multiplexing (OFDM) system and single-carrier system with frequency domain equalization (SC-FDE). In this paper, we propose a novel HC system based on efficient fast-convolution (FC) without a cyclic prefix to meet the requirements of complex scenarios towards future communications. The proposed system can achieve flexible optimization of system performance, such as minimizing out-of-band emission (OOBE), peak to average power ratio (PAPR) and bit error rate (BER). This efficient FC based hybrid carrier (FCHC) system is among the most competitive waveforms. The closed-form BER expressions are derived and analyzed over additive white Gaussian noise (AWGN) and frequency-selective fading channels. A modified equalizer over frequency-selective channels is also provided to resist the inter-block interference (IBI). Moreover, the banded minimum mean-square error (MMSE) equalization based on FC can be embedded in the proposed system over doubly-selective channels, and then the interference caused by fading channels can be further suppressed through the combination of FC and HC. The simulation results demonstrate the superiority of the proposed system over the conventional HC system in terms of OOBE, PAPR and BER. The significant flexibility and combined performance advantages make the proposed system a potential scheme for future communications.

We consider the downlink of an unmanned aerial vehicle (UAV) assisted cellular network consisting of multiple cooperative UAVs, whose operations are coordinated by a central ground controller using wireless fronthaul links, to serve multiple ground user equipments (UEs). A problem of jointly designing UAVs’ positions, transmit beamforming, as well as UAV-UE association is formulated in the form of mixed integer nonlinear programming (MINLP) to maximize the sum UEs’ achievable rate subject to limited fronthaul capacity constraints. Solving the considered problem is hard owing to its non-convexity and the unavailability of channel state information (CSI) due to the movement of UAVs. To tackle these effects, we propose a novel algorithm comprising of two distinguishing features: (i) exploiting a deep Q-learning approach to tackle the issue of CSI unavailability for determining UAVs’ positions, (ii) developing a difference of convex algorithm (DCA) to efficiently solve for the UAV’s transmit beamforming and UAV-UE association. The proposed algorithm recursively solves the problem of interest until convergence, where each recursion executes two steps. In the first step, the deep Q-learning (DQL) algorithm allows UAVs to learn the overall network state and account for the joint movement of all UAVs to adapt their locations. In the second step, given the determined UAVs’ positions from the DQL algorithm, the DCA iteratively solves a convex approximate subproblem of the original non-convex MINLP problem with the updated parameters, where the problem’s variables are transmit beamforming and UAV-UE association. Numerical results show that our design outperforms the existing algorithms in terms of algorithmic convergence and network performance with a gain of up to 70%.

In this paper we consider signal detection in multiple-input-multiple-output (MIMO) systems with an impulsive noise channel. The existing, near optimal, sphere decoder (SD) achieves good performance, however, the computational complexity is directly related to the number of nodes visited during the tree search and the signal-to-noise ratio (SNR). Using neural network techniques, a Deep Learning Detector (DLD) is proposed. The DLD method detects signals transmitted in an impulsive noise channel, after an off-line training phase. The detection process of DLD has lower complexity than the average SD complexity, while exhibiting good performance. What is even more interesting is that the computational complexity of DLD is constant across SNR, in contrast to the SD detectors, which have an exponential complexity across the SNR. This constant complexity could be very helpful when implementing a detector in practice because it could allow for better optimization of resources. To evaluate the performance of our proposed method we have used a low level simulator that generates a fairly accurate model of a MIMO system with an impulsive noise channel. The complexity analysis and simulation results validate the arguments presented in this paper.

Multiple Sclerosis (MS) lesion segmentation from MR images is important for neuroimaging analysis. MS is diffuse, multifocal, and tend to involve peripheral brain structures such as the white matter, corpus callosum, and brainstem. Recently, U-Net has made great achievements in medical image segmentation area. However, the insufficiently use of context information and feature representation, makes it fail to achieve segmentation of MS lesions accurately. To solve the problem, 3D attention context U-Net (ACU-Net) is proposed for MS lesion segmentation in this paper. The proposed ACU-Net includes 3D spatial attention block, which is used to enrich spatial details and feature representation of lesion in the decoding stage. Furthermore, in the encoding and decoding stage of the network, 3D context guided module is designed for guiding local information and surrounding information. The proposed ACU-Net was evaluated on the ISBI 2015 longitudinal MS lesion segmentation challenge dataset, and it achieved superior performance compared to latest approaches.

In this work, we consider the downlink of an unmanned aerial vehicle (UAV) assisted cellular network consisting of multiple cooperative UAVs, whose operations are coordinated by a central ground controller using the fronthaul communications, to serve multiple ground users. A problem of jointly designing UAV's location, transmit beamforming, as well as UAV-user association is formulated in the form of mixed integer nonlinear programming (MINLP) to maximize the sum user achievable rate while considering the constraints of limited fronthaul capacity. Solving the formulated problem is computationally hard owing to the its non-convex nature and the unavailability of channel state information (CSI) due to the undetermined and flexible movement of UAVs. To tackle these effects, we propose a novel algorithm exploiting the deep Q-learning approach to take the hassles of unavailable CSI for determining UAV's location and invoking the difference of convex (DC) based optimization method to efficiently solve for the UAV's transmit beamforming and UAV-user association given the determined UAV's location. The algorithm recursively solves the formulated problem until convergence. Numerical results show that our design outperforms the existing work in terms of algorithmic convergence and network performance and achieve a gain of up to 70% compared to the existing algorithms.

In systems that aim to mitigate the inter-carrier interference (ICI) caused by doubly selective (DS) channels, the partial Fast Fourier Transform (PFFT) has emerged as an interesting alternative to the conventional FFT. In this letter, we propose a novel PFFT demodulation scheme based on fast convolution (FC) structure. The proposed scheme uses the overlap-save operation of FC to further suppress the residual interference (RI) to improve performance over DS channels, including in terms of bit error rate (BER), while, at the same time, allowing for an improvement in spectral efficiency by avoiding the use of a cyclic prefix (CP). We provide a theoretical analysis of RI and numerical simulations to demonstrate the superiority of the proposed scheme in terms of BER performance.

This paper proposes full-duplex transmissions from legitimate nodes to achieve channel estimation performance deterioration at an eavesdropper as compared to the legitimate receiver. The proposed discriminatory channel estimation (DCE) technique comprises of two stages where, in the first stage, the self-interference channel is estimated by the respective legitimate nodes. Followed by in-band full-duplex transmission from both legitimate nodes for channel estimation at legitimate nodes, while providing equivocation at the eavesdropper due to the superposition of two signals. The discrimination of channel estimation performance provides secrecy against the passive eavesdropper while delivering information to the legitimate receiver. We provide the mean square error (MSE) to indicate the performance achieved by linear minimum mean square error (LMMSE) estimators. We have also provided bit error rate (BER), and secrecy capacity analysis to indicate the performance of secure communication achieved by securing the channel estimates from the eavesdropper. The BER analysis shows that for proposed DCE, BER at the eavesdropper is close to 0.1 while the legitimate node is able to robustly decode the information. Finally, simulation results show that the proposed DCE outperforms existing DCE techniques for the considered scenario.

Spectrum sharing and device-to-device (D2D) transmission are among the key features of modern communication networks. In this work, we are particularly interested in these two techniques from a sharing for secrecy perspective. The considered communication model consists of a multi-user cellular system and an underlying secondary system comprising K D2D pairs. All cellular and D2D transmissions are subject to an eavesdropping attack. Given a predefined secrecy condition, imposed by the primary system to guarantee a desired secrecy throughput, KS D2D pairs are allowed to share the spectrum and send their secret data while the remaining K-KS device transmitters operate as cooperative jammers. First, we characterize the achievable secrecy rates for both systems under a joint secrecy constraint on all transmitted cellular and D2D messages. Then, we propose a device selection scheme to determine the optimal number of devices, KS, that maximizes the secrecy throughput of the secondary system while satisfying the primary's secrecy condition. The obtained results show that both parties can win under the proposed transmission scheme; the cellular system can significantly improve its secrecy throughput, and the D2D pairs get to share the spectrum and achieve secure transmissions.

The random nature of wireless channel can be utilized to achieve secure communications. Recently, there have been some contributions towards finding practical approaches to provide physical layer security, but robust and secure channel estimates are critical in achieving secrecy. In this paper, we propose that in-band full-duplex communication can be utilized by the legitimate transceiver pair to mask the channel estimates from the eavesdropper. The proposed secure channel estimation technique comprises of two stages. In the first stage, the self- interference channel is estimated by respective nodes. In the second stage, in-band full-duplex transmission from both legitimate nodes is utilized to achieve channel estimates at legitimate nodes while providing equivocation at the eavesdropper. The performance analysis shows that the proposed scheme provides robust channel estimation at the legitimate nodes while maintaining equivocation at the eavesdropper.

The long term evolution-advanced (LTE-A) and worldwide interoperability for microwave access (WiMAX) standards are the two main contenders in the 4th generation (4G) wireless systems, which adopt the single carrier-frequency division multiplexing access (SC-FDMA) and orthogonal frequency division multiplexing access (OFDMA) schemes in the uplink, respectively. However, these two schemes have certain advantages and disadvantages. As the weighted fractional Fourier transform (WFRFT)-based system can merge the SC- FDMA scheme with the OFDMA scheme, it inherits these two schemes' characteristics and has the potential to obtain better performance. For enhancing the bit error rate (BER) performance of the future uplink communication, in this paper, we propose a new order selection method for the WFRFT- based system to obtain better BER performance than the existing conventional uplink SC-FDMA or OFDMA scheme concerning the peak-to-average power ratio (PAPR) and the uplink power control (UPC) parameter under frequency selective fading channels.

Non-orthogonal multiple access (NOMA) has been recognized as a promising technique for future radio access. In this paper, different from [3], it is proposed that the cyclic prefix (CP) is used in the asynchronous uplink NOMA transmission for mitigating asynchronous effects. Through analyzing the inter-user interference (IUI) and residual interference, it is shown that the asynchronism can degrade the uplink NOMA users' bit error rate (BER) performance severely for the conventional-successive interference cancellation (Conv-SIC) technique. By applying the SIC-phase compensation (SIC-PC) technique and CP, the asynchronous effects can be mitigated. The BER performance of near and far users can be improved substantially, compared with other existing successive interference cancellation (SIC) techniques.

In this paper, the performance of hybrid carrier (HC) systems based on weighted fractional Fourier transform (WFRFT) is investigated in an uplink non‐orthogonal multiple access (NOMA) scenario. NOMA is a promising technology to improve the system capacity, in which two users (far‐user and near‐user relative to a base station) are allocated to use the same time‐frequency resources, and the successive interference cancellation (SIC) technique is implemented to decode signals at the receiver. Considering the actual error decoding in the SIC process (i.e. imperfect SIC), NOMA cannot avoid the inter‐user interference (IUI) and residual interference (or error propagation). Therefore, firstly IUI and residual interference are analysed, and signal to interference plus noise ratio (SINR) of the far‐user is expressed mathematically considering the residual interference. Then, based on the analysis of IUI, considering different WFRFT orders, a near‐user BER expression over additive white Gaussian noise (AWGN) channels is derived. Furthermore, the optimal WFRFT order selection to minimise the interference influence in the uplink is formulated and solved efficiently. Simulation results have verified the mathematical expression of SINR, the near‐user theoretical BER expression, and the proposed optimal WFRFT order selection to obtain the maximum sum spectral efficiency.

This paper investigates the incremental-redundancy hybrid-automatic repeat request (IR-HARQ) transmission over independent block-fading channels in the presence of an eavesdropper, where the secrecy of the transmission is ensured via introduction of dummy messages. Since the encoder only knows the statistics of the channel state, the secrecy and the reliability are defined in a probabilistic framework. Unlike previous works on this subject, we design a coding strategy tailored to IR-HARQ by splitting the dummy-message rate over several rate parameters. These additional degrees of freedom improve the match between the dummy-message rates and the realizations of the eavesdropper channels. We evaluate the performance in terms of secrecy outage probability, connection outage probability, and throughput and compare it with the benchmark paper by Tang et al. Numerical examples illustrate that, comparing to existing alternatives, splitting of the dummy-message rate provides higher throughput and lower expected duration/average delay.

In this work, we investigate the ergodic secrecy rate region of a block-fading spectrum-sharing system, where a D2D communication is underlying a cellular channel. We consider that both the primary and the secondary transmissions require their respective transmitted messages to be kept secret from a common eavesdropper under a joint secrecy constraint. The presented results are for three different scenarios, each corresponding to a particular requirement of the cellular system. First, we consider the case of a fair cellular system, and we show that the impact of jointly securing the transmissions can be balanced between the primary and the secondary systems. The second scenario examines the case when the primary network is demanding and requires the secondary transmission to be at a rate that is decodable by the primary receiver, while the last scenario assumes a joint transmission of artificial noise by the primary and the secondary transmitters. For each scenario, we present an achievable ergodic secrecy rate region that can be used as an indicator for the cellular and the D2D systems to agree under which terms the spectrum will be shared.

A weighted fractional Fourier transform (WFRFT) scheme is proposed for the orthogonal frequency division multiplexing (OFDM) system to combat impulsive noise (IN) in substation communications. According to the central limit theorem, the analytic expression of the real-valued IN probability density function (PDF) after WFRFT is derived and for ease of calculation, the closed form expression of the approximate PDF is also put forward. The PDF of IN shows that WFRFT makes IN more concentrated to the small amplitudes in the fractional Fourier transform domain than in the frequency domain at the receiver. As a result, the WFRFT scheme achieves a better bit error rate (BER) performance than the conventional OFDM system. In addition, the proposed scheme is also compatible with other existing IN mitigation techniques, such as dynamic peak-based threshold estimation-lookup table (DPTE-LUT) blanking. Simulation results have verified that the derived analytic and closed form PDF expressions of IN after WFRFT and the advantage of the proposed scheme over the counterpart OFDM system in the IN channels at high SNR with and without the DPTE-LUT blanking.

In order to reduce interference to the primary system and further improve the spectrum efficiency, the Fractional Fourier Transform is introduced as a tool to solve the spectrum sharing problem and a new spectrum sharing protocol is proposed in this paper where sinusoidal signals and chirp signals are utilized for primary users and cognitive users, respectively. In our spectrum sharing protocol, secondary users take the initiative to access the spectrum or act as a DF relay to assist the primary system adaptively according to the channel conditions of the direct transmission link. Moreover, the joint optimization of power allocation and user selection of distributed secondary users is studied to achieve the maximum achievable data rate for the secondary system while guaranteeing the target data rate of primary user. Simulation results show that compared with previous protocols, the outage probability of both primary and secondary systems can be improved, and with an increase in the number of secondary users, a higher achievable data rate can be obtained for the secondary system.

In this paper, we consider the performance analysis of a decode-and-forward (DF) cooperative relaying (CR) scheme over bursty impulsive noise channel. As compared to existing literature, here, we generalize the performance analysis to the multi-relay scenario with and without considering error propagation from the relays. For this scheme, we evaluate the bit error rate (BER) performance in the presence of Rayleigh fading with a maximum a posteriori (MAP) receiver. From the obtained results it is seen that, similar to single relay scheme, the proposed MAP receiver attains the lower bound derived for multi-relay DF CR scheme also, and performs significantly better than the conventional schemes developed for additive white Gaussian noise (AWGN) channel and memoryless impulsive noise channel. Moreover, the performance improvement of optimal MAP receiver over the memoryless receivers is substantial with increasing the number of relays.

Relay technique is introduced in the Long Term Evolution Advanced (LTE-Advanced) system to extend the network coverage, improve the performance of cell-edge user equipments (UEs) and increase the fairness among different UEs. The combination of multiple-input-multiple-output (MIMO) and relay can further increase frequency efficiency and improve system performance. In this paper, based on the architecture of MIMO relay-enhanced TD-LTE-Advanced system, a fairness -aware resource allocation algorithm is proposed, and the resource allocation is divided into 3 phases: a frequency resource partition scheme is firstly proposed to realize the tradeoff between fairness among UEs served by various relay nodes (RNs) and system throughput; secondly, a backhaul resource allocation algorithm is presented to match the eNB to relay transmission with the relay to UE transmission with reduced feedback overhead; finally, a resource allocation algorithm from eNB and relay to UEs is given to maximize the proportional fairness so that user fairness can be guaranteed. Simulation results show that both system throughput and user fairness can be improved with limited feedback overhead, using the proposed algorithms.

Relay selection among available relays is crucial in improving the performance of cooperative relaying. In this paper, we address the relay selection for a multi-relay decode-and-forward (DF) cooperative relaying scheme over channels impaired by bursty impulsive noise in the presence of Rayleigh fading. For this scheme, we propose a novel best relay selection technique which is based on both the channel state information of source-relay and relay-destination links and the state of the impulsive noise that affect those links. Then, we provide a end-to-end bit error rate (BER) expression for the proposed scenario using BPSK modulation. Finally, the accuracy of the provided analytical results and the effectiveness of the proposed best relay selection algorithm under different network topology is discussed. It is shown that, by using the bursty impulsive noise characteristics, the proposed relay selection algorithm performs significantly better than the conventional relay selection algorithm which is optimized for additive white Gaussian noise environment.

Power substations generate a significant “bursty impulse noise” that might interfere with wireless technologies working in the vicinity of power equipment. Existing wireless systems are not designed for such an environment; we propose a Maximum a Posteriori (MAP) receiver designed with Markov-Gaussian models in order to mitigate the impact of impulsive noise in substations. We study and compare different noise models implemented in the receiver and we discuss the performance of the receiver based on the characteristics of the impulsive noise. Our proposed model can be used by a MAP receiver to offer optimum performances from low signal to noise ratio (SNR). When the communication is disturbed by impulsive noise measured in the field, the MAP receiver still offers better performance than when using other models, but mainly at higher SNR.

Wireless sensor network (WSN) technologies and distributed processing are essential to develop ubiquitous sensing in the Internet of Things (IoT) paradigm, wherein sensors pervasively collect data and perform information processing and communication tasks to achieve a common objective. This paper presents the formulation and analysis of distributed estimation algorithms based on the diffusion cooperation scheme in the presence of errors due to the unreliable data transfer among nodes. In particular, we highlight the impact of transmission errors on the least-mean squares (LMS) adaptive networks. We develop the closed-form expressions for the steady-state mean-square deviation (MSD), which is helpful to assess the effects of the imperfect information flow on the behavior of diffusion LMS algorithms in terms of steady-state error. The model is then validated by performing Monte Carlo simulations. It is shown that local and global steady-state MSD values are not necessarily monotonic increasing functions of the error probability. We also assess sufficient conditions that ensure mean and mean-square stability of diffusion LMS strategies in the presence of transmission errors. We examine a practical scenario where errors occur at the medium access control (MAC) level. To overcome the problem of unreliable data exchange, we implement a random pairwise strategy that improves the performance of the estimation algorithm in the presence of high transmission error rates. Moreover, issues such as scalability in the sense of network size and regressor size, convergence behavior during the transient phase, spatially correlated observations, as well as the effect of the distribution of the noise variance are studied.

In this paper, we consider the performance analysis of a decode-and-forward (DF) cooperative relaying (CR) scheme over channels impaired by bursty impulsive noise. Although the Middleton class-A model and the Bernoulli-Gaussian model give good results to generate a sample distribution of impulsive noise, they fail in replicating the bursty behavior of impulsive noise, as encountered for instance within power substations. To deal with that, we adopt a two-state Markov-Gaussian process for the noise distribution. For this channel, we evaluate the bit error rate performance of direct transmission (DT) and a DF relaying scheme using M-ary phase shift keying (M-PSK) modulation in the presence of Rayleigh fading with a maximum a posteriori (MAP) receiver. From the obtained results, it is seen that the DF CR scheme in bursty impulsive noise channel still achieves the space diversity and performs significantly better than DT under the same power consumption. Moreover, the proposed MAP receiver attains the lower bound derived for DF CR scheme and leads to large performance gains compared with the conventional receiving criteria, which were optimized for additive white Gaussian noise channel and memoryless impulsive noise channel.

In this work, various channel coding schemes that can be used in hybrid automatic repeat request (HARQ) transmission protocols are investigated from an energy efficiency point of view. Conventional HARQ, where only one bit is used to inform the transmitter about the decoding success or failure, is compared to adaptive HARQ where the transmitter adapts either the length or the transmit power of the codewords using outdated channel state information (i.e., experienced by the receiver during the past transmissions). Describing the problems within a Markov decision process framework, we find optimal adaptation policies for both persistent (unlimited number of transmission) and truncated HARQ protocols. Numerical examples obtained in a Rayleigh block fading channel show that, in terms of energy efficiency, the adaptation of the codewords length provides notable gains over power adaptation and conventional HARQ.

Wireless technologies and vehicle-mounted or wearable medical sensors are pervasive to support ubiquitous healthcare applications. However, a critical issue of using wireless communications under a healthcare scenario rests at the electromagnetic interference (EMI) caused by radio frequency transmission. A high level of EMI may lead to a critical malfunction of medical sensors, and in such a scenario, a few users who are not transmitting emergency data could be required to reduce their transmit power or even temporarily disconnect from the network in order to guarantee the normal operation of medical sensors as well as the transmission of emergency data. In this paper, we propose a joint power and admission control algorithm to schedule the users' transmission of medical data. The objective of this algorithm is to minimize the number of users who are forced to disconnect from the network while keeping the EMI on medical sensors at an acceptable level. We show that a fixed point of proposed algorithm always exists, and at the fixed point, our proposed algorithm can minimize the number of low-priority users who are required to disconnect from the network. Numerical results illustrate that the proposed algorithm can achieve robust performance against the variations of mobile hospital environments.

In TD-LTE-A downlink, multi-user beamforming is used to decrease co-channel interference of different users and increase system performance, and Signal-to-Leakage-and-Noise Ratio (SLNR) beamforming algorithm has been proved to have better performance in terms of sum capacity and average BER performance compared with other beamforming algorithms with moderate complexity. In order to further improve the performance, the power allocation algorithm is combined with the framework of SLNR beamforming. Based on the combined algorithm, we can maximize sum capacity and guarantee users' requirements in multiuser downlink scenario with multiple-input-multiple-output (MIMO). The problem is transformed into convex optimization and then solved by geometric programming. Simulation results show that sum capacity of the proposed algorithm is slightly inferior to water-filling algorithm, however, the proposed algorithm can satisfy users' requirement while water-filling algorithm cannot. Then, a sub-optimal algorithm is also proposed to reduce implementation complexity at the cost of 10% sum capacity loss.

The rapidly growing traffic demand on mobile networks and the new services offered by service providers requires effective strategies for handling network resources. One such strategy is the ability of mobile devices to be attached to multiple radio nodes simultaneously. By this strategy the average data rate, that a mobile terminal can transmit/receive, increases and improves the reliability. This new strategy enables the use of dynamic load balancing among radio nodes, which in turn protects mobile networks from congestion caused by sudden load increase or poor channel conditions. Nevertheless load balancing requires techniques for efficiently splitting traffic without increasing the delay or causing packet reordering. In this paper we focus our study to the uplink load balancing case. We propose QBALAN, a new load balancing algorithm that uses two main strategies: long-term strategy splits traffic that remains on the system for long periods of time, and the short-term strategy splits traffic such that packet reordering is minimized. Numerical results show that our algorithm reduces the splitting error while reducing packet reordering.

In this work, a novel diagram called guaranteed power diagram with limited communication range is introduced. The proposed diagram is used to develop distributed deployment algorithms for a network of nonidentical mobile sensors with limited communication ranges, where the coverage priority of different points in the field is specified by a priority function, and the information of sensors' locations is inaccurate due to measurement errors. The proposed algorithms are iterative and in each iteration, the sensors find their new locations and move towards them. Simulation results confirm the effectiveness of the proposed algorithms.

We investigate mobile sensor deployment algorithms when the sensors have a limited communication range and they estimate other sensors location from the messages they receive from them. These pragmatic constraints bring two implications: the former implies that each sensor will only be aware of the presence of those sensors whose communication ranges include that sensor, and the latter indicates that location information can be inaccurate, due to estimations error. Consequently, the conventional Voronoi-based mobile sensor deployment algorithms fail to guarantee a simple, reliable coverage detection; additionally, the sensors are prone to collision. We introduce a new set of Voronoi-based diagrams, named guaranteed Voronoi diagrams with limited communication, to tackle the above problems.

In this paper, we focus on the analysis of the end-to-end outage probability of hybrid automatic repeat request (HARQ) in a multi-hop, block-fading channel. First, we use a Markov chain to model the multi-hop system with truncated HARQ type-I (HARQ-I) and derive closed-form expressions for the outage probability and throughput; we also obtain lower and upper bounds on the energy efficiency of the system. Then, we extend the analysis to cover HARQ with packet combining where we provide a general formulation of the outage probability; in special cases, closed-form expressions for outage probability and throughput are obtained. Finally, for a multi-hop system with opportunistic routing, we present recursive formulas to obtain outage probability and throughput. The expressions we derived allow us to draw conclusions about the outage diversity of different multi-hop protocols. In particular, we show that the diversity order is improved for opportunistic multi-hop system and is linearly increasing with the number of allowed transmissions for each packet. Numerical results illustrate how the analytical expression can be used to compare various multi-hop HARQ transmission strategies.

In this work, we analyze hybrid ARQ (HARQ) transmission over the independent block fading channel. We consider two scenarios with respect to the message sent by the receiver via the feedback channel: i) “conventional”, one-bit feedback used to inform the transmitter about the decoding success/failure (ACK/NACK), and ii) the multi-bit feedback message, where on top of ACK/NACK, the transmitter is provided with additional information about the state of the receiver. To minimize the outage probability under long-term average and peak power constraints, we cast the problems into the dynamic programming (DP) framework and solve them for Nakagami-m fading channels. An approximate, closed-form solution for the high signal-to-noise ratio (SNR) regime is proposed using geometric programming (GP). The obtained results quantify the advantage of the multi-bit feedback over the conventional one-bit approach, and show that the power optimization can provide significant gains over conventional power-constant HARQ transmissions even in the presence of peak-power constraints.

A novel technique for electroencephalogram (EEG) compression is proposed in this paper. This technique models the intrinsic dependence inherent between the different EEG channels. It is based on methods borrowed from dipole fitting that is usually used in order to find a solution to the classic problems in EEG analysis: inverse and forward problems. To compress the EEG signals, the forward model based on approximated source dipoles is first used to provide an approximation of the recorded signals. Then, (based on a smoothness factor) appropriate coding techniques are suggested to compress the residuals of the fitting process. Results show that this technique works well for different recordings and for different patients, and is even able to provide near-lossless compression for certain types of recordings.

End-to-end outage probability and end-to-end throughput are analyzed in a multi-hop channel with truncated automatic repeat request type-I (ARQ-I). The system is modelled as a Markov chain and the closed-form expressions for the outage and throughput are derived for the cases of variable and equal distances between the nodes. The limits on the performance in high SNR and large number of transmissions are derived. We show that the diversity order is linearly increasing with the number of allowed transmission for each packet.

Soft source decoders, in conjunction with error correcting channel codes, can be used to improve the error resilience of digital communication systems based on variable length codes. In this paper, we present a novel approach to reduce the complexity of maximum a posteriori variable length decoders implemented on a bit-symbol trellis. The decoding algorithm is implemented in a narrow corridor along the trellis diagonal to reduce the decoder complexity. Furthermore, by periodically adjusting the corridor boundaries, a significant reduction in complexity is achieved at the price of a small degradation in decoding performance.

In this work, we analyze hybrid ARQ (HARQ) protocols used over independent block fading channels. We assume that the transmitter is unaware of the channel state information (CSI) at the moment of transmission but knows the statistics of the channel. We consider two scenarios with respect to the feedback received by the transmitter: (i) “conventional”, one-bit feedback carrying the message about the decoding success/failure (ACK/NACK), and (ii) the multi-bit feedback scheme when, on top of ACK/NACK, the receiver feeds back information about the state of the decoder, which is equivalent to the CSI experienced by the receiver in the past (unsuccessful) transmissions. The feedback can then be used to allocate power across the HARQ transmission attempts (in the case of one-bit feedback) or adapt the power for each re-transmission as a function of multi-bit feedback. The objective in both cases is the minimization of the outage probability under peak and long term average power constraints. We solve the optimization problems using the dynamic programming framework. The obtained results quantify the advantage of the multi-bit feedback over the conventional approach and indicate that power optimization can provide significant gains over the conventional constant-power HARQ transmission even in the presence of peak-power constraints.

Single-cell-based coverage hole detection algorithms are developed for mobile sensors deployment under inaccurate location information and non-identical sensing ranges. Existing Voronoi-based diagrams require the exact location of sensors to guarantee a simple, single-cell-based coverage detection, and they miss the mark if the location information is inaccurate. The guaranteed Voronoi-based diagrams, proposed in this paper, extend the existing diagrams in a way to guarantee the single-cell based coverage hole detection when upper bounds on location errors are provided. Simulation results show that with inaccurate location information the proposed algorithms can largely increase the network coverage. Even if the location information is exactly known, we suggest assuming some error margins to improve the network coverage based on the proposed algorithms.

We analyze HARQ transmission over a fading channel with the help of a relay node. In the absence of the channel state information (CSI) at the transmitters, we establish the problem of maximizing the overall throughput of HARQ transmission by optimizing the variable transmission rate. We present a closed form of the throughput with respect to the transmission rates and use the well-known Dynamic Programming technique along with the necessary simplifications on the problem to find the optimal set of transmission rates for both the Source node and the Relay node in source-relay-destination scenario. We show that such an approach yields a significant increase in throughput compared to fixed-rate HARQ.

We analyze the average throughput and the outage probability for relay-based incremental redundancy HARQ transmission over block fading channel. We focus on the effects of having an error-free multi-bit feedback channel instead of the single-bit ACK/NACK feedback used in conventional HARQ. This multi-bit feedback message provides the cooperating nodes with outdated channel state information (CSI) so that they can adapt their transmission rate. We discuss a network with M relay nodes and assume adaptive transmission rate for all the cooperating nodes. We describe the adaptation problem as a Markov Decision Process (MDP) and employ the Dynamic Programming (DP) for optimization. The numerical results obtained in the case of one relay transmission, indicate that significant throughput gains can be obtained when compared to non-adaptive, i.e, fixed-rate HARQ. Moreover, we study the performance limits of the system model by finding an upper bound of the throughput for the cooperative HARQ channel which is applicable to both adaptive and non-adaptive models. Finally, we analyze the discretization issues and conclude that only a small number of feedback bits is required by the adaptive system to outperform the conventional single-bit HARQ.

In cognitive radio networks, in order to avoid interference form secondary users to the primary license holders of the spectrum, reliable spectrum sensing is necessary. In scenarios where the noise samples are correlated, the spectrum sensing methods optimized considering impairment by independent noise samples will not provide optimum performances. To address this issue, a locally optimum detection method for random signals under a weakly correlated noise model over fading channels is proposed. The probabilities of false alarm and detection of the proposed detector in the low signal to noise ratio regime are analyzed. The average probabilities are calculated over different channel gains. Numerical and simulation results demonstrate the superiority of the proposed method over the known energy detection method with comparable complexities. Furthermore, we consider the scenario where the estimated correlation is different from the real correlation and investigate the effect of this correlation mismatch on the probabilities of false alarm and detection of the proposed method.

Discrete Fourier transform (DFT) codes have been used to provide robustness against errors and erasures in various applications. This paper focuses on improving error localization of the Bose-Chaudhuri-Hocquenghem (BCH) DFT codes. First, we analyze how the subspace-based error localization outperforms the coding-theoretic one. Then, we propose an extension of the subspace-based error localization, based on additional syndrome, that improves the existing one and is naturally suitable for rate-adaptive distributed source coding (DSC). Further, we propose a new generic subspace-based algorithm to decode BCH-DFT codes. The proposed approach generalizes the encoding and decoding of this important class of DFT codes. It introduces many different decoding matrices for a DFT code; this diversity is then used to diminish the effect of the quantization noise and thus to improve the decoding. Finally, the extended and generalized approaches are combined to maximize the decoding gain. Simulation results demonstrate the capability of the proposed algorithms to perform significantly better than the existing subspace-based error localization, in the presence of quantization noise.

A novel technique for real-time electroencephalogram (EEG) compression is proposed in this paper. This technique makes use of the redundancy between the different frequency subbands present in EEG segments of one channel. It uses discrete wavelet transform (DWT) and dynamic reference lists to compute and send the decorrelated subband coefficients. Set partitioning in hierarchical trees (SPIHT) is also used as source coder. Experimental results showed that the proposed method can not only compress EEG channels in one dimension (1- D), but also detect seizure-like activity. A diagnostics-oriented performance assessment was performed to evaluate the performance of both the compression and detection capabilities of the proposed method. In this paper, we show that the algorithm can positively detect seizure sections in the recordings at bitrates down to 2 bits per sample.

The installation of wireless technologies in power substations requires characterizing the impulsive noise produced by the high-voltage equipment. Substation impulsive noise might interfere with classic wireless communications and none of the existing models can reliably represent this noise in wide band. Previous studies have shown that impulsive noise is characterized by series of damped oscillations with the amplitude, the duration and the occurrence times of the impulses that are random. All these characteristics make this noise time-correlated and the partitioned Markov chain remains an efficient model that can ensure the correlation between the samples. In this study, we propose to design a partitioned Markov chain to generate an impulsive noise that is similar to the noise measured in existing substations, in time and frequency domains. We configure our Markov chain to produce the impulses with the damped oscillation effect, then, we determine the probability transition matrix and the distribution of each state of the Markov chain. Finally, we generate noise samples and we study the distribution of the impulsive noise characteristics. Our Markov chain model can replicate the correlation between the measured noise samples; also the distributions of the noise characteristics are similar in the simulations and the measurements.

A novel technique for Intra-cerebral Electroencephalogram (iEEG) compression in real-time is proposed in this article. This technique uses eigendecomposition and dynamic dictionary update to reduce the EEG channels to only one decor related channel or eigenchannel. Experimental results show that this technique is able to provide low distortion values at very low bit rates (BRs). In addition, performance results of this method show to be better and more stable than JPEG2000. Results do not vary a lot both in time and between different patients which proves the stability of the method.

In this paper we analyze the maximum achievable throughput of Hybrid ARQ (HARQ) protocol for transmission over cooperative block-fading channels. It is assumed that the instantaneous channel state information (CSI) is not available at the encoders and only its outdated version is available via the error-free feedback channel. The outdated CSI is then used to adapt the transmission rate of all cooperating nodes. Considering source-relay-destination scenario, we use Dynamic Programming (DP) to find the optimal rate adaptation policy for both the Source node and the Relay node. Such an approach yields a significant increase in throughput compared to non-adaptive fixed-rate HARQ. Moreover, the maximum achievable throughput is shown for the cooperative HARQ channel which is applicable to both adaptive and non-adaptive models.

In this work we consider the front-end receivers for flat fading MIMO transmission, whose essential feature is the calculation of the log-likelihood ratios (LLRs) for the transmitted bits. When the number of transmit antennas and the modulation size grow, the exact calculation of the LLRs is most often unfeasible due to the necessity to enumerate the symbols affecting the output. In this work, to avoid the burden of explicit enumeration, instead of calculating the likelihoods exactly, we explore the possibilities offered by the saddlepoint approximation of the likelihood functions. The resulting fixed-complexity approach significantly outperforms the well known linear-filter based minimum mean square error detection.

A subspace-based approach for rate-adaptive distributed source coding (DSC) based on discrete Fourier transform (DFT) codes is developed. Punctured DFT codes can be used to implement rate-adaptive source coding, however they perform poorly after even moderate puncturing since the performance of the subspace error localization degrades severely. The proposed subspace-based error localization extends and improves the existing one, based on additional syndrome, and is naturally suitable for rate-adaptive distributed source coding architecture.

This paper is dedicated to the study of HARQ protocols under a secrecy constraint. An encoder sends information to a legitimate decoder while keeping it secret from the eavesdropper. Our objective is to provide a coding scheme that satisfies both reliability and confidentiality conditions. This problem has been investigated in the literature using a coding scheme that involves a unique secrecy parameter. The uniqueness of this parameter is sub-optimal for the throughput criteria and we propose a new coding scheme that introduces additional degrees of freedom. Our code involves Secrecy Adaptation and Rate Adaptation and we called it SARA-code. The first contribution is to prove that the SARA-code has small error probability and small information leakage rate. The second contribution is to show, over a numerical example, that the SARA-code improves the secrecy throughput.

State-space representation of positioning problems has enabled the use of particle filters to probabilistically estimate the location from the noisy sensor measurements. However, in particle filtering, the choice of the motion and sensor models, as well as the importance density used, are crucial for a good approximation. In this work, we have used Gaussian Mixtures to model the prior and likelihood densities, as they can be used for a wide range of distributions and can capture the multimodality of the densities. Additionally, with GMM prior and likelihood densities, we were able to evaluate and use the Optimal Importance Density for particle filters, which resolves the degeneracy of particles and sample impoverishment. We have provided simulation results based on field measurements to illustrate the validity of our models and the improvements made by using our proposed importance density.

In this work, we consider the front-end receivers for flat fading MIMO transmission, whose essential feature is the calculation of the log-likelihood ratios (LLRs) for the transmitted bits. When the number of transmit antennas and the modulation size grow, the exact calculation of the LLRs becomes unfeasible due to the necessity to enumerate the symbols affecting the output. In this work, to avoid the burden of explicit enumeration, instead of calculating the likelihoods exactly, we explore the possibilities offered by the saddlepoint approximation of the likelihood functions. The resulting fixed-complexity approach significantly outperforms the well known linear-filter based minimum mean square error detection. A complexity analysis identifies the main computational bottleneck of the proposed detection algorithm.

In this paper, we study joint network coding and distributed source coding of inter-node dependent messages, with the perspective of compressed sensing. Specifically, the theoretical guarantees for robust ℓ1-min recovery of an under-determined set of linear network coded sparse messages are investigated. We discuss the guarantees for ℓ1-min decoding of quantized network coded messages, based on Restricted Isometry Property (RIP) of the resulting measurement matrix. This is done by deriving the relation between tail probability of ℓ2-norms and satisfaction of RIP. The obtained relation is then used to compare our designed measurement matrix, with i.i.d. Gaussian measurement matrix, in terms of RIP satisfaction. Finally, we present our numerical evaluations, which shows that the proposed design of network coding coefficients results in a measurement matrix with an RIP behavior, similar to that of i.i.d. Gaussian matrix.

To improve the performance of hospital emergency department (ED), we propose a scheduling strategy to reduce the total time of medical tests in ED for a given number of patients. We model the schedulig strategy as a constraint logic programming problem, and then solve this problem with an open source software: Minizinc. Also we compare the medical test time with the scheduling strategy and that without scheduling, which is viewed as a benchmark for our proposed scheduling strategy. The results show that the scheduling strategy can save 12%–25% of medical test time.

In this paper, we detail the design of a hypertension monitoring system via telecommunication and computer technologies and propose a method for theoretically analyzing its system accuracy. In the design of our system, we add a novel decision support unit, building on a diagnosis standard in medicine, into our monitoring system. The architecture of our system, including this decision support system, is detailed in this paper. In the analysis of our system accuracy, we theoretically evaluate the accuracy of a decision support system by linking the system accuracy with the distribution of sensors' errors. Additionally, we verify our proposed analysis of system accuracy in this paper by comparing its result to that of Monte Carlo simulation.

In this paper, in the context of Joint Source-Channel Coding (JSCC), we investigate the idea of adaptive downsampling in Oversampled Filter Banks (OFBs) with polyphase matrix degree greater than zero. By adaptively changing the downsampling rate, we can control the injected redundancy in the sub-band signals. Thus, we can modify the level of signal protection based on the importance of different parts of the signal or variations in the condition of the channel. This can be used in protection of Region of Interests (ROI) in image/video coding and can also be considered as some form of puncturing. Here, using time-domain analysis, we will provide sufficient conditions on OFB structure to perfectly reconstruct the output. Finally, using OFB design examples, we provide some simulations which further reveal the potential applications of this method.

Motivated by a host of recent applications requiring some amount of redundancy, frames are becoming a standard tool in the signal processing toolbox. In this paper, we study a specific class of frames, known as discrete Fourier transform (DFT) codes, and introduce the notion of systematic frames for this class. This is encouraged by application of systematic DFT codes in distributed source coding using DFT codes, a new application for frames. Studying their extreme eigenvalues, we show that, unlike DFT frames, systematic DFT frames are not necessarily tight. Then, we come up with conditions for which these frames can be tight. In either case, the best and worst systematic frames are established from reconstruction error point of view. Eigenvalues of DFT frames, and their subframes, play a pivotal role in this work.

In order to reduce the Peak-to-Average Power Ratio (PAPR) of multicarrier signals, a DFT precoding technique can be particularly useful. In this paper we propose to combine such a DFT precoding with the OFDM/OQAM modulation to provide a new frequency access scheme, that we call in short DFT-OQAMA. This DFT-OQAMA technique is compared to the Single-Carrier Frequency Division Multiplex Access (SC-FDMA) recently proposed for the UpLink (UL) of the 3GPP LTE system.

In this paper, we propose a novel method to estimate the accuracy of a decision support system for hypertension monitoring. The decision support system is designed building on a diagnosis standard in medicine, and one decision made by this system depends on both the blood pressure data gathered by medical sensors and some contexts that are manually entered by clinicians. When analyzing the system accuracy, we take into account both the potential sensors' errors and the errors of entering context. In addition, we propose a novel method for the estimation of system accuracy; this method is motivated by the fact that the traditional method to estimate system accuracy would overestimate the system accuracy (the details would be presented in Section III.C.). Finally, we compare the system accuracy estimated by the proposed method and that estimated by the traditional method. Our study shows that our proposed method can well estimate the system accuracy.

We show how real-number codes can be used to compress correlated sources, and establish a new framework for distributed lossy source coding, in which we quantize compressed sources instead of compressing quantized sources. This change in the order of binning and quantization blocks makes it possible to model correlation between continuous-valued sources more realistically and correct quantization error when the sources are completely correlated. The encoding and decoding procedures are described in detail, for discrete Fourier transform (DFT) codes. Reconstructed signal, in the mean-squared error sense, is seen to be better than or close to quantization error level in the conventional approach.

To enhance the capacity of patients supported by wireless in-hospital monitoring systems, a bandwidth allocation scheme for the transmission of medical data in the wireless local area network (WLAN) is proposed. The problem of bandwidth allocation, subject to limited wireless bandwidth, quality of service (QoS) requirements of medical data transmission, as well as electromagnetic interference, is modeled as a non-polynomial (NP) optimization problem. To save the computation time of this NP problem, we propose an accelerated genetic algorithm by dynamically adjusting both the inheritance probability and the mutation probability, and then compare it with other off-the-shelf genetic algorithms. Our study shows that our proposed algorithm can save computation time and attain the same result of bandwidth allocation in comparison with other algorithms.

We study the capacity of discrete memoryless channels with synchronization errors and additive noise. We first show that with very large alphabets, their capacity can be achieved by independent and identically distributed input sources, and establish proven tight lower and upper capacity bounds. We also derive tight numerical capacity bounds for channels where the synchronization between the input and output is partly preserved, for instance using incorruptible synchronization markers. Such channels include channels with duplication errors, channels that only insert or delete zeros, and channels with bitshift errors studied in magnetic recording. Channels with small alphabets and corrupted by synchronization errors have an infinite memory. Revisiting the theoretical work of Dobrushin and adapting techniques used to compute capacity bounds for finite-state source/channel models, we compute improved numerical capacity lower bounds for discrete memoryless channels with small alphabets, synchronization errors, and memoryless noise. An interesting and some- what surprising result is that as long as the input sequences are not completely deleted, the capacity of channels corrupted by discrete timing errors is always nonzero even if all the symbols are corrupted.

In video transmission systems, compressed video is delivered over channels that are not necessarily error free. Channel errors can lead to a mismatch in the encoder/decoder prediction loop which will propagate the errors to the succeeding frames. As a result, the quality of the received video at the decoder side may drop significantly. In this paper, we aim to reduce the introduced mismatch by modifying the reference frame. Our technique makes use of previously Intra coded blocks and a new leaky prediction structure in order to improve the robustness. This technique was applied on temporal and spatial scalability of scalable extension of H.264/AVC. Simulation results show the effectiveness of our scheme especially for medium and high motion sequences.

In this paper, a method is proposed to compress multichannel electroencephalographic (EEG) signals in a scalable fashion. Correlation between EEG channels is exploited through clustering using a k-means method. Representative channels for each of the clusters are encoded individually while other channels are encoded differentially, i.e., with respect to their respective cluster representatives. The compression is performed using the embedded zero-tree wavelet encoding adapted to 1-D signals. Simulations show that the scalable features of the scheme lead to a flexible quality/rate tradeoff, without requiring detailed EEG signal modeling.

Oversampled Filter Banks (OFB) can be used for erasure correction due to the redundancy that they insert into the signal. People have studied erasures using a frame-theoretic approach in which erasure is defined as the deletion of the frame expansion coefficients corresponding to a frame element. This approach applies to OFBs and results in an erasure being interpreted as losing all samples in a sub-band. However, a more realistic and flexible scenario is when an erasure is considered as the loss of sub-band samples only for particular time instances. Alternatively in an erasure-free channel, analogous to the concept of puncturing, the sub-band samples might be erased deliberately to change the code rate. In both of these scenarios, the erasure pattern of sub-bands can differ for different time instances. In this paper, we investigate this situation and derive a set of sufficient conditions that should be satisfied by the OFB in order to reconstruct the output. Simulation results will be presented for substantiating the proposed OFB structure which show the efficiency of this method in increasing the resilience of OFB against erasures.

In this paper, we propose a method for reconstructing the output of an Oversampled Filter Bank (OFB) when instantaneous erasures happen in the sub-band domain. Instantaneous erasure is defined as a situation where the erasure pattern changes in each time instance. This definition differs from the type of erasure usually defined in literature, where e erasures means that e channels of the OFB are off and do not work at all. This new definition is more realistic and increases the flexibility and resilience of the OFB in combating the erasures. Additionally, similar to puncturing, the same idea can be used in an erasure-free channel to reconstruct the output, when sub-band samples are discarded intentionally in order to change the code rate. In this paper we also derive the sufficient conditions that should be met by the OFB in order for the proposed reconstruction method to work. Based on that, eventually we suggest a general form for the OFBs which are robust to this type of erasure.

With the rapid development of multimedia technology, video transmission over error prone channels is widely used. Using predictive video coding can lead to temporal and spatial propagation of channel errors, which consequently results in high degradation in the quality of the received video. In order to address this problem different error resilient methods have been proposed. In this paper, a number of the error resilient methods based on reference frame modification are overviewed briefly and examined with scalable extension of H.264/AVC (SVC). We propose a new method based on hierarchical structure used in temporal scalable coding. Average gains of 0.76 dB over the improved generalized source channel prediction (IGSCP) method and 2.26 dB over normal coding are achieved.

Transmission of compressed video over unreliable networks is vulnerable to errors and error propagation. Multi-hypothesis motion compensated prediction (MHMCP) which was originally developed to improve compression efficiency has been shown to have a good error resilience property. In this paper we improve the overall performance of MHMCP in packet loss scenarios by performing optimal mode switching within a rate distortion framework. The approach builds on the recursive optimal per-pixel estimate (ROPE), which is extended by re-deriving recursion formulas for the more complex MHMCP setting, so as to achieve an accurate estimation of the end-to-end distortion. Simulation results show significant performance gains over the standard MHMCP scheme and the importance of performing effective mode decisions. We also show results with comparison to conventional ROPE.

In this paper, we present a channel state information (CSI) feedback reduction algorithm to be used in an adaptive orthogonal frequency division multiplexing (OFDM) system. Adaptive OFDM transceivers, which offer improved data rates and power efficiency over conventional OFDM systems, require a duplex connection between the transmitter and receiver such that the CSI necessary for the transmitter to recalculate the optimal bit and power distributions can be fed back. When a large number of subcarriers are employed, the amount of CSI to be transmitted in the reverse link can consume a prohibitively large amount of bandwidth. In order for adaptive OFDM systems to be implemented in practice, this feedback data has to be reduced. Existing methods of reducing CSI feedback entail compressing the bit and power distributions separately, resulting in the transmission of redundant information due to the correlation between the two distributions. The proposed algorithm removes this redundancy and improves upon an existing feedback compression algorithm that both reduces the number of feedback transmissions necessary to maintain some desired bit error rate (BER) and compresses the CSI by feeding back subchannel gains rather than the bit and power distributions themselves. After applying the proposed algorithm to an adaptive OFDM system with parameters characteristic of wireless local area networks (WLANs), the results show that the number of feedback transmissions can be reduced by up to 90% for slowly varying channels and the CSI to be fed back can be compressed by more than a factor of three.

FIR Oversampled Filter banks (OFB) with degree zero polyphase matrices can be used as block codes over the field of real numbers. Since they can be regarded as real-valued codes, one can also use their potential advantage of joint source-channel coding and graceful degradation over noisy channels. Due to the variable state of noisy channels, increase in bit error probability is unavoidable hence increasing the injected redundancy of the input signal or decreasing the code rate is necessary. To decrease the code rate, the downsampling rate should be decreased as well as filter lengths to maintain the degree zero of the polyphase matrix. However, decreasing the filter lengths will deteriorate the performance of the OFB both in compression rate of the input signal and the amount of injected redundancy. This paper proposes the idea of adaptively changing the downsampling rate to modify the code rate without changing the filter length. This could be considered as some form of puncturing. We show that adaptive downsampling results in just changing the way input is fed into the OFB. Simulation results for quantized DFT codes with erasures show the efficiency of this method specially when the number of erasures is not very large.

Leaky prediction schemes have been proposed to reduce transmission error propagation by trading off with video coding efficiency. We propose using leaky prediction to improve the robustness of generalized source channel prediction scheme for H.264/AVC in error prone scenarios. We apply leaky prediction selectively at a macroblock level based on a simple mean square error metric in order to reduce the bit-rate penalty. Substantial performance gains were observed in simulations for medium and fast moving video sequences.

Intra coding is one of the most effective ways of reducing the impact of error propagation caused by predictive coding. However, intra coding requires a higher bitrate when compared to inter coding. In order to use Inter coding and reduce error propagation it is important that inter macroblocks predict from ¿safe¿ areas that have a decreased chance of spreading errors. To this end we propose a low complexity method of biasing the prediction mechanism towards recently intra updated macroblocks. We devise a method of adjusting the distortion used in rate distortion optimization to take into account the temporal distance of the last Intra macroblock. Our simulations show that our intra-distance derived weighting (IDW) method improves video coding performance in a lossy environment by up to 1.4 dB for a modest increase in bitrate.

Conventional rate-distortion (RD) optimized video coding is formulated for an error-free environment. Special considerations have to be made when transmitting video in lossy networks. In this paper we demonstrate a novel method of weighting the distortion used for mode decisions and motion estimation in H.264. By determining an appropriate weighting factor, motion vectors are biased towards macroblocks (MB) that have less influence on the motion propagation path and mode decisions are made considering a MB's future impact. We therefore propose tracking the influence that each macroblock has along the motion propagation path to determine the weights. Considerable performance improvements are unveiled in comparison with state of the art robust video schemes.

Transmissions over hybrid wireline-wireless networks suffer both packet losses and bit errors. To facilitate delay-sensitive audio/video communications over hybrid networks, low-delay error recovery techniques, such as forward error correction (FEC) and multiple description coding (MDC), are utilized to provide protection against bit errors and packet losses, respectively. As a means of joint source-channel coding, the robust multiple description scalar quantizer (RMDSQ) was introduced to combat both packet losses and bit errors. In this paper, a novel RMDSQ system is proposed by utilizing both MDC and FEC-based techniques. In the sense of rate distortion, end-to-end performance of the RMDSQ system against packet losses and bit errors is compared with that of individual FEC and MDC-based techniques. Instead of the traditional two-state Gilbert channel model, a three-state Markov chain is proposed to model hybrid networks and work as the testbed. Experimental results show that the proposed RMDSQ system achieves higher robustness against increasing packet losses and bit errors. In contrast, FEC-based techniques achieve better performance against bit errors; however, their performance deteriorates significantly due to packet losses.

Based on multiple description coding, the robust multiple description scalar quantizer (RMDSQ) was introduced to combat both packet losses and noisy channels over heterogeneous networks. In the RMDSQ, the description with bit errors is utilized to achieve graceful performance degradation. However, because of high computational complexity of the index assignment (IA) by using exhaustive search and the genetic algorithm, it is not practical in applications where the training time is of primary concern. In this paper, we propose a novel IA algorithm with low computational complexity to achieve a balanced RMDSQ, in which parity bits are added to realize robustness against bit errors. The algorithm proposed can be implemented "on-the-fly" by using logic circuits, and easily applied in bitplane-based image compression. Experimental results show that the proposed algorithm outperforms existing algorithms in the sense of both spreads and computational complexity.

We propose two error resilient techniques for transmission of compressed video over error prone scenarios. The first technique is an improvement on the generalized source channel prediction (GSCP) scheme proposed by Yang and Rose in 2006. The second technique is a modification of weighted prediction (WP) present in the H.264 standard. Both techniques provide greater emphasis on Intra MBs during the formation of a reference frame thereby achieving better error resilience than GSCP with a very modest increase in the bit-rate.

Abstract In this paper we present a new method to simplify trellis‐based decoding of linear block codes. Our method is based on removing (pruning) some edges and states from the trellis representation of the code; the trellis is pruned through some hard‐decisions that are made on received bits whose likelihood exceeds a predefined threshold. We show in this paper how this simplification can be accounted for by a new generator matrix (and sometimes a coset leader) that totally parametrize the new trellis, or, equivalently, the set of allowed codewords after simplification. Extensive simulations show that significant computational savings can be achieved, at a very small loss in coding performance, as long as the operating point and threshold are carefully chosen. Moreover, the application of this technique to iterative decoding of product codes is outlined, and our results show that the simplifications do not hinder the convergence, again, as long as proper parameters are chosen. Copyright © 2007 John Wiley & Sons, Ltd.

In this paper, we show how to compute the parameters of the lattice implementation of linear-phase perfect reconstruction filter banks (OLPPRFBs) (L. Gan and K.-K. Ma, 2003). This lattice implementation is based on a parametrization of the filterbank by a series of left-invertible matrices. It is generic enough to cover orthogonal (para-unitary) and bi-orthogonal filterbanks, as well as any oversampling factor. This lattice has been used mainly for design purposes, where the parameters defining the left-invertible matrices are varied, and the corresponding filterbank response computed, until a desired frequency response is achieved. In this paper, we aim at recovering the parameter matrices from the impulse response of the analysis (and, for high oversampling ratios, synthesis) filters of the filterbank.

Error resilient multiple description coding (ERMDC) consists of a robust encoder and an enhanced decoder. It was developed to achieve higher error tolerance than the classical multiple description coding (MDC) for error-prone channels when bit errors of one description exceeded the error correction capability of the applied forward error correction (FEC) code. In this paper, ERMDC is extended to Rayleigh fading channels with additive white Gaussian noise (AWGN) by utilizing soft channel outputs. By using soft channel outputs as receiver inputs, the accuracy of estimates of detectable transmission errors is improved so that the reconstruction distortion is reduced further. Experimental results show that soft input ERMDC outperforms significantly the existing works without extra redundancy.

In this paper we investigate the impact of transmission errors in H.264. Transmission errors propagate into subsequent frames due to motion prediction and result in degraded video quality. Our simulations show that H.264 exhibits non-fading behaviour. We propose a method that introduces a fading characteristic and can eliminate the error propagation after a few frames. We provide a detailed analysis of our results based on a comparison with MPEG-4 and the residual energy per frame.

In this paper, a novel loading algorithm consisting of four variants is proposed for a multicarrier transceiver employing multiple antennas configured for spatial diversity. The primary objective of the proposed algorithm is to increase the overall throughput while ensuring the mean bit error rate (BER) is below a specified limit. To achieve this, spatial diversity is employed to improve the subcarrier signal-to-noise ratio (SNR) values. Simultaneously, (uniform or non-uniform) bit allocation, which is a function of subcarrier SNR, is performed to increase throughput. To reduce power consumption, spatial processing complexity, and hardware costs, antenna subset selection is also performed by the proposed algorithm to choose a set of active transmit/receive antennas. The results show that combining bit allocation with spatial diversity (employing antenna subset selection) can yield substantial throughput increases.

Abstract In this paper, we present a proof‐of‐concept, fixed‐point, DSP hardware implementation of an adaptive bit loading algorithm that is designed for wireless multicarrier transceivers. Adaptive bit loading is used to enhance the performance of multicarrier transceivers by tailoring the subcarrier signal constellations to the channel conditions, which can vary across the subcarriers. Since most bit loading algorithms possess a high computational cost and are unable to cope with rapid variations of wireless channels, they are seldom used in present wireless standards. To prove that adaptive bit loading is feasible for wireless transceivers, our work focuses on the implementation of a known bit loading algorithm that can quickly search for the final bit allocation in an iterative manner. The goal of this algorithm is to yield the largest‐possible throughput while satisfying a mean BER constraint. The performance of the hardware implementation operating in time‐varying channel conditions is studied in terms of the overall throughput. Furthermore, the robustness of the hardware implementation is evaluated, relative to sudden changes in the channel that interrupts the run of the algorithm. Real‐time operations and fixed‐point representation issues are included in the discussion. Additionally, we propose a modified algorithm implementation that is more robust to channel variations. Copyright © 2007 John Wiley & Sons, Ltd.

Abstract In this paper, we explore computationally efficient implementations of the soft output viterbi algorithm (SOVA) applied to Soft‐Input Soft‐Output (SISO) decoding of linear block codes. In order to simplify the trellis‐based decoding of binary block codes with SOVA, we use the technique of sectionalization of the trellis, which has been successfully applied to the simplification of the MAP and Max‐Log‐MAP algorithms. Due to the branch complexity of the sectionalized trellis, we define a generalization of a non‐binary version of SOVA. However, the computational complexity of directly applying this approach remains too high for efficient implementation; we thus introduce the concept of non‐binary SOVA (NSOVA) with propagation of bit‐level reliabilities (BLR). This new algorithm is analyzed from a computational complexity viewpoint. Both serial and parallel implementations are explored. Finally, optimal sectionalizations are derived for selected codes; since the normal SOVA decoding is a particular case of NSOVA with BLR, we show that our approach is more efficient than a bit‐level trellis by showing that, for all the codes tested, the optimal trellis is a sectionalized one. Copyright © 2007 John Wiley & Sons, Ltd.

Vector Quantization (VQ) is a lossy data compression technique that is often applied in the field of speech communications. When applying VQ to speech signals, usually combined with some for of Linear Prediction Coding (LPC), the input signal is divided into frames of a given length. Each frame is then processed and quantized using VQ. A typical assumption is the fact that the frame size is fixed. In this study, we propose a modification to this technique that allows for variable size frames, providing an additional degree of freedom for the optimization of the encoding process. The results show that this technique effectively improves the quality of the encoded signal at a given bit rate, even if the improvement is not dramatic.

Multiple description coding (MDC) is applied to achieve acceptable performance when only one single description is received correctly, and better resolution of the source when all descriptions are received correctly. In previous work, the description with bit errors is either discarded completely, or estimated by using complicated algorithms. A new central receiver is proposed in this paper so as to achieve better performance by utilizing the residual information between descriptions received correctly or with noise. After theoretical analysis based on MDSQ, optimal and suboptimal solutions are provided. Analytical and experimental results show that the proposed algorithms achieve better performance than the side distortion within a large range of BERs. In particular, the suboptimal algorithm achieves similar performance to that of the optimal with low computational complexity and without accurate channel knowledge.

This paper proposes a low complexity image quality assessment method based on frequency domain transforms. The proposed scheme is easy to implement and reliably estimates image quality across various distortions types including images compressed using frequency domain transforms (JPEG and JPEG2000). The performance of the two proposed metrics (Q-DCT and Q-DWT) is compared with two full-reference metrics, the PSNR and another recent state-of-the-art measure. The proposed metrics provide a global numerical image quality measure but can also be used to obtain local image quality statistics. Preliminary results indicate that the proposed scheme can be adapted to perform well in reduced reference frameworks making it well-suited for real-time applications.

Multiple description coding (MDC) can achieve acceptable performance with only one single correct description, and lower distortion of the source when more descriptions are received correctly. In our previous work, we assumed that because of noisy channel, bit errors were introduced into one description, but the other one was received correctly. Based on this assumption, an enhanced central decoder has been proposed to utilize the residual information between the corrupted and the correct description to reduce the distortion of the reconstructed signals. In this paper, a novel index assignment algorithm for multiple description scalar quantizer (MDSQ) encoder is developed to improve the error detection capability of the central decoder. Genetic algorithm (GA) is used to search for a suboptimal solution. The experimental results show that within a range of bit error rates (BERs), the proposed algorithm provides lower reconstruction distortion than the conventional MDSQ

DFT encoding over the real (or complex) field has been proposed as a means to reconstruct samples lost in wireless network transmissions. The quantization associated with practical implementations results in reconstruction errors that are particularly large when lost samples occur in bursts (bursty erasures). Here we present a setup that effectively creates disjoint DFT codes along the rows (temporal codevectors) and columns (subband codevectors) of a frame under analysis. An erasure burst along a particular codevector can then be broken up by reconstructing some lost samples along the remaining orientation; these samples can then be used as received samples in reconstructing the original codevector, a technique that we refer to as pivoting. Expressions for the temporal code reconstruction error and for temporal-to-subband pivoting operations are presented.

Adaptive bit loading, a technique which is essential for taking full advantage of multicarrier modulation (MCM), has been shown to yield significant performance improvement, especially in the case of wireless MCM. However, because of its high computational cost and the rapid variations of wireless channels, it is seldom used in present wireless standards. In this paper, we present a proof of concept implementation of an adaptive bit loading algorithm, intended for a wireless MCM system, onto a fixed-point DSP. The algorithm considered is iterative in nature and aims to maximize the throughput under a BER constraint. The performance of the implementation is studied in terms of throughput given time varying conditions. Further, the implementation’s robustness is evaluated, relative to sudden changes in the channel that interrupts the algorithm’s run. Issues such as real-time operations and fixed-point representation are included in the discussion. Additionally, a modified implementation, more robust to channel variations is proposed.

In this letter, we study the impact of the design freedom brought about by oversampling in perfect reconstruction oversampled filter banks, on the choice of synthesis filters that maximize the coding gain. In particular, based on the expression of the coding gain for the oversampled case, we show the dependency of the coding gain on both analysis and synthesis filters. We explore how the choice of synthesis filters can affect the coding gain, given a fixed set of analysis filters. We show that the para-pseudo-inverse is a good choice for the maximization of the coding gain when the corresponding analysis filters have been optimized for coding gain under critical sampling.

We present discrete adaptive bit loading algorithms for multicarrier systems with uniform (nonadaptive) power allocation operating in a frequency selective fading environment. The algorithms try to maximize the overall throughput of the system while guaranteeing that the mean bit error rate (BER) remains below a prescribed threshold. We also study the impact of imperfect subcarrier signal-to-noise ratio information on throughput performance. Results show that the proposed algorithms have approximately the same throughput and mean BER as the optimal allocation while having a significantly lower computational complexity relative to other algorithms with near-optimal allocations. Moreover, when compared with algorithms that employ approximations to water filling, the computational complexity is comparable while the overall throughput is closer to the optimum.

In this paper, we evaluate and compare the robustness of several adaptive bit loading algorithms for multicarrier transmission systems, when imperfect subcarrier signal-to-noise ratio (SNR) information is used. In particular, we investigate the impact of the uncertainty of data-aided channel estimation techniques on system performance. We also examine an implementation issue associated with adaptive bit loading algorithms that use metrics related to the SNR. Although such metrics can be derived via closed form expressions, look-up tables are used instead to reduce system complexity, resulting in the SNR values being quantized. Thus, we examine the effects of SNR quantization on system performance. Finally, we present a technique for choosing SNR values in a fixed length look-up table in order to minimize quantization error.

The advent of turbo decoding techniques has revived the interest in soft-input soft-output (SISO) decoders that are capable of outputting reliability information (soft decisions). A classical way of implementing a SISO decoder for a linear block code is to use a trellis-based decoding algorithm, such as MAP, Max-Log-MAP or SOVA [S. Lin et al. (1998)]. Recently, the idea of sectionalizing the trellis diagram representing a code has been proposed as a means to decrease the decoding complexity associated with the SISO decoders [S. Lin et al. (1998), Y. Liu et al. (2000)].This paper investigates the application of a nonbinary SOVA decoding algorithm L. Cong et al. (1999) to the sectionalized trellis of a binary linear block code and generalizes the nonbinary SOVA to a code with time-varying symbol set. Based on the above study to derive a new simplified algorithm, called the nonbinary SOVA with bit-level reliabilities (NSOVA with BLR) that we also investigate in terms of complexity for a serial and a parallel implementation. Based on the computational complexity figures for both algorithms and shows that optimal sectionalizations can be found, that minimize the computational complexity entailed when using these algorithms for decoding.

We present an efficient bit allocation algorithm for multicarrier systems operating in frequency-selective environment. The proposed algorithm strives to maximize the overall throughput while guaranteeing that the mean bit error rate (BER) remains below a prescribed threshold. The algorithm is compared with several other algorithms found in literature in terms of the overall throughput, mean BER, and relative computational complexity. Furthermore, the algorithms are compared with an exhaustive search routine to determine the optimal bit allocation in terms of maximizing throughput given the constraint on error performance. No power allocation is performed by the algorithms. Results show that the proposed algorithm has approximately the same throughput and mean BER as the optimal solution while possessing a significantly lower computational complexity relative to the other algorithms with similar performance. When compared to algorithms which employ approximations to waterfilling, the computational complexity is comparable while the overall throughput is closer to the optimum.

In this paper, we investigate the performances of Gaussian modeling and linear prediction tools for error detection and concealment in the transmission of still images. We consider the transmission of subband encoded images through two types of channels. We model the residual correlation between subband coefficients by considering them as jointly Gaussian variables. The first transmission medium considered is a packet-oriented channel, where some packets are lost during transmission. The problem is to estimate the values of missing coefficients. In this case, particular care must be taken while evaluating correlation matrices from incomplete data. The other system considered is based on a discrete memoryless noisy channel affecting the data being transmitted. The challenge is here first to determine the locations of the errors--which is done through hypotheses tests--and then to replace them by estimates based on their neighbors. The reconstruction via linear prediction is shown to give better results than median filtering based reconstruction. Error detection through this Gaussian model also shows promising results, in particular when channel statistics are taken into account in a joint source-channel decoding framework.

In this paper we propose an original study of the reconstruction of subband compressed images impaired by channel transmission errors. The method proceeds in two steps:first a detection scheme is applied to determine which coefficients of the subband decomposition have been affected by transmission, and then an estimation step tries to evaluate the erroneous coefficients. In our model, subband coefficients are considered to be drawn from jointly Gaussian random processes. Based on this assumption, conditional statistics can be computed which enable to test the likelihood of a given set of received coefficients with respect to the rest of the image. The detection and estimation processes are derived from these statistics. The method is validated through simulation and visual results are provided. The drawbacks of the method are outlined and explained through the discrepancies between the Gaussian assumption and real world images, namely around image edges.

In this paper, we extend structures and factorizations for paraunitary filter banks yielding linear phase filters to the oversampled case. We derive minimal factorizations for a subset of these banks, and propose some design examples.