Source and Channel Coding

An error resilient video coding scheme based on unequal error protection over error prone channels is proposed. Due to the sensitivity of VLC to channel errors, a single bit error can cause loss of synchronization at the receiver, leading to catastrophic degradation of image quality. Multiplexed codes are an effective approach to combat losses of synchronization. In this paper, multiplexed codes appropriate to video coding are designed and then integrated into a block based motion compensated video coder. We compared performances of this structure with one of the most used codes: Huffman code. Results show that the proposed scheme achieves high error resilience at almost no cost in compression efficiency.

Scalable multimedia data transmission are subject to specific constraints such as the quality of service (QoS) of sensitivity classes and the transmission rate (yielding a maximum size of each frame to send). Many scalable source decoders are used to discarding data than processing an erroneous stream. This featuring class structure is helpful to define a strategy that determines the maximum number of classes to send and deliver the suitable protection and transmission scheme (coding rates and modulation) to apply in accordance with the transmission constraints. It leads to the possible truncation of frame parts transmitted with an unequal error protection (UEP) scheme for severe channel conditions. In a MPEG-4 speech frames context, we compare our approach to other methods using equal error and existing UEP schemes. It results in a significant improvement of the peak SNR (PSNR) quality in poor channel transmission conditions

The objective of image compression is to reduce irrelevance and redundancy of the image data in order to be able to store or transmit data in an efficient form. Pulse Code Modulation (DPCM). In the first step wavelet transform is used and in the second step the image is processed through the DPCM. The simulation results show a improvement in the performance compared to the wavelet transform technique. The comparison was also made by choosing various combinations of band and measuring the compression ratio and PSNR.

Investigation of the underlying physics or biology from empirical data requires a quantifiable notion of similarity - when do two observed data sets indicate nearly identical generating processes, and when they do not. The discriminating characteristics to look for in data is often determined by heuristics designed by experts, $e.g.$, distinct shapes of "folded" lightcurves may be used as "features" to classify variable stars, while determination of pathological brain states might require a Fourier analysis of brainwave activity. Finding good features is non-trivial. Here, we propose a universal solution to this problem: we delineate a principle for quantifying similarity between sources of arbitrary data streams, without a priori knowledge, features or training. We uncover an algebraic structure on a space of symbolic models for quantized data, and show that such stochastic generators may be added and uniquely inverted; and that a model and its inverse always su...

MESHGRID is a scalable 3D object representation method which is part of MPEG-4 AFX. This paper proposes a new approach for optimized protection of MESHGRID-represented objects against transmission errors occurring over error-prone channels. An unequal error protection approach is followed, to cope with the different error-sensitivity levels characterizing the various resolution and quality layers produced by MESHGRID. The protection level to be employed for each layer is determined by solving a joint source and channel coding problem. The Linfinite distortion metric is used, and its superiority over the classical L-2 norm is proven. In this context, a novel fast algorithm for solving the optimization problem is conceived. The proposed joint source and channel coding approach offers resilience against transmission errors, enables a real-time implementation, and preserves all the scalability features and animation capabilities that characterize MESHGRID.

An enhanced Vertical Bell Laboratories layered space-time (V-BLAST)-MIMO system complemented with joint source and channel coding scheme is proposed. The coding scheme is rate compatible punctured convolutional code (RCPC Code) with unequal error protection (UEP) to provide variable gain diversity with single encoder-decoder structure. The proposed system showed an ability to provide better error protection levels to data streams as

Video monitoring applications for underground rely on wireless train-to-wayside communication systems which require high data rate as well as high Quality of Service (QoS) level. In order to satisfy both constraints we propose a combined source and channel coding approach in the context of MIMO (Multiple Input Multiple Output) video transmission. In the present case, MIMO transmission is based on the PHY layer of IEEE 802.11n Wi-Fi standard currently deployed in a railway tunnels. Two different strategies are studied: first, the association between Multiple Description Coding (MDC) and a STBC (Space Time Block Code) MIMO scheme is considered when no channel information is available at transmitter side. In the case when perfect channel information is available at transmitter side (CSIT), a Singular Value Decomposition of the MIMO channel is possible. This transmission scheme is then associated with scalable video coding, which consists here in the separation of the scene into different Regions Of Interest (ROI). The creation of the regions of interest is based on the Flexible Macroblock Ordering (FMO) technique introduced in the new H.264/AVC compression standard. The stream associated to the area with the maximal perceptual relevance is transmitted on the eigen-channel with the higher gain. Consequently, this strategy which provides unequal protection against channel errors, allows guaranteeing better robustness and acceptable reconstructed video quality at the control-centre. The two different strategies of transmission have been evaluated thanks to realistic simulations. Two antenna configurations representative of real cases encountered in railway tunnels are considered. The channel model is generated by using the correlation based Kronecker model obtained by computing the channel matrix with a 3D ray tracing tool. Simulation results show that the two proposed solutions allow enhancing the reconstructed video quality

Bit error rate (BER) performance of convolutional coded quaternary DPSK (QDPSK) with Viterbi decoding is theoretically investigated in Rayleigh fading environments. The probability density functions of the path and branch metric values of Viterbi decoding are derived. The BER's after decoding due to additive white Gaussian noise (AWGN) and cochan-ne1 interference are theoretically analyzed. Rate 1/2 codes and their symbol punctured high-rate codes are considered, and the optimal symbol positions for deletion to minimize the BER after decoding are presented for the codes with constraint length K = 3-7. It is shown that Viterbi decoding considerably reduces the desired signal-to-interference power ratio (SIR), as well as the signal energy per information bit-to-noise power spectrum density ration ( Eb /No), necessary to achieve a certain BER. The spectrum efficiency of the cellular mobile radio system, achievable by the use of the symbol punctured codes, is also evaluated.

The codeword error rate (WER) performance of noncoherent FSK with soft decision decoding of block codes using Chase's second algorithm is investigated in a Rayleigh fading channel. The received signal envelope is sampled and used as channel measurement information. The theoretical upper and lower bounds of the WER are derived assuming independent Rayleigh envelope samples in a received block. When the Golay (23, 12, 7) code is used, soft decision decoding with 6-bit error correction capability (3-bit error and 3-bit erasure) requires an average signal-to-noise power ratio about 5 dB lower than that for minimum distance decoding with 3-bit error correction to obtain a WER = lo-'. The effects of bit interleaving on the WER performance when fading envelope variations is slow compared to the bit rate are investigated through computer simulations. When 23 x M , bit interleaving is used for transmitting M , Golay codewords, the simulation results show that M, = 0.2 x (bit rate/fading maximum Doppler frequency) is sufficient to obtain the full advantages of soft decision decoding. The theoretical analysis was supported by laboratory experiments.

This report shows how one can find a solution to the K-SAT equations with the use of purely local computations. Such a local network, inspired by the Survey Propagation equations driven by an external input vector, potentially has an exponential number of attractors. This gives the network powerful classification properties, and permits to reconstruct either noisy or incomplete inputs. It finds applications from bayesian inference to error-correcting codes and gene-regulatory networks, and its local structure is ideal for an implementaion on FPGA. Here we write its algorithm, characterize its main properties and simulate the corresponding VHDL code. One shows that the time of convergence towards a solution optimally scales with the size of the network.

Scalable multimedia data transmission are subject to specific constraints such as the Quality of Service (QoS) of sensitivity classes and the transmission rate (yielding a maximum size of each frame to send). Many scalable source decoders are used to discarding data than processing an erroneous stream. This featuring class structure is helpful to define a strategy that determines the maximum number of classes to send and delivers the suitable protection and transmission scheme (coding rates and modulation) to apply in accordance with the transmission constraints. It leads to the possible truncation of frame parts transmitted with an Unequal Error Protection (UEP) scheme for severe channel conditions. In a MPEG-4 speech frames context, we compare our approach to other methods using equal error and existing UEP schemes. It results in a significant improvement of the Peak SNR (PSNR) quality in poor channel transmission conditions.

In this paper, we consider the source-channel rate allocation for different transmission schemes. We propose a new trellis structure and a new algorithm that are able to deal with both variable length packet and fixed length packet problems. The trellis description allows to handle any kind of transmission schemes and can therefore be applied to Bit Interleaved Coded Modulations (BICM), parallel channels or Hybrid-Automatic Repeat reQuest (HARQ).

This work presents a novel technique t o analyze the computational efforts of an orderingfree variant of the generalized Dijkstras's algorithm (GDA) and the maximum-likelihood sequential decoding algorithm (MLSDA) based on the Berry-Esseen theorem. Different from the theoretical bounds determined by the conventional central limit theorem argument, which often holds only for sufficiently large blocklength, the new bounds are valid for any blocklength.

Variable length codes (VLCs) are widely adopted in many compression standards due to their good coding efficiency on average codeword length. However, an inherent problem with a VLC is that an error of even one bit can cause serious error propagation and thus loss of synchronization at the receiver, which would lead to a series of non-correctly decoded symbols. Reversible variable length codes (RVLCs) were introduced to significantly mitigate this phenomenon. In this work, a method to find an optimal RVLC in terms of the minimum average codeword length is first formulated as a tree-searching problem, and then, instead of performing an exhaustive search, an 𝐴 * -based construction algorithm is proposed to find an optimal RVLC. The proposed algorithm has been applied to several benchmarks for sources and has found respective optimal symmetric and asymmetric RVLCs.

Equivalence of Codes are constructed classification on finite rings according to odd codes, even codes and recurrent codes. It is determined that the codes are odd codes or even codes and the relationship with the Hadamard codes was studied. In this study, some matrices is written with the elements on the finite rings. Codes are generated with these matrices. The relationship between the generated codes and the perfect codes is revealed.

We present a dynamic joint source and channel coding adaptation algorithm for the AMR speech codec based on the ITU-T Emodel. This model takes both delay and packet loss into consideration. We address the problem of finding the optimal choice of source and channel bit rates given QoS information about the wired and wireless IP network and subject to constraints on maximum packet loss, maximum delay and maximum allowed transmission rate. Our results show that an adaptation is necessary to preserve acceptable levels of quality while making optimal use of the allowed bandwidth. Our technique requires a small number of computations that allows real time operation in parallel to voice streams.

Claude Shannon's pioneering work quantified the performance limits of communications systems operating over classic wireline Gaussian channels. However, his source and channel coding theorems were derived for a range of idealistic conditions, which may not hold in low-delay, interactive wireless multimedia communications. Firstly, Shannon's ideal lossless source encoder, namely the entropy encoder may have an excessive codeword length, hence exhibiting a high delay and a high error sensitivity. However, in practice most multimedia source signals are capable of tolerating lossy, rather than lossless delivery to the human eye, ear and other human sensors. The corresponding lossy and preferably low-delay multimedia source codecs however exhibit unequal error sensitivity, which is not the case for Shannon's ideal entropy codec. There are further numerous differences between the Shannonian lessons originally outlined for Gaussian channels and their ramifications for routinely encountered dispersive wireless channels, where typically bursty, rather than random errors are encountered. This paper elaborates on these intriguiging lessons in the context of a few turbo-transceiver design examples, using a jointly optimised turbo transceiver capable of providing unequal error protection in the context of MPEG-4 aided wireless video telephony. The transceiver investigated consists of Space-Time Trellis Coding (STTC) invoked for the sake of mitigating the effects of fading, Trellis Coded Modulation (TCM) or Bit-Interleaved Coded Modulation (BICM) as well as two different-rate Non-Systematic Convolutional codes (NSCs) or Recursive Systematic Convolutional codes (RSCs). A single-class protection based benchmarker scheme combining STTC and NSC is used for comparison with the unequal-protection scheme advocated. The video performance of the various schemes is evaluated when communicating over uncorrelated Rayleigh fading channels. It was found that the achievable performance of the proposed scheme is within 0.99 dB of the corresponding capacity of the Rayleigh fading channel.

In this paper, we consider the source-channel rate allocation for different transmission schemes. We propose a new trellis structure and a new algorithm that are able to deal with both variable length packet and fixed length packet problems. The trellis description allows to handle any kind of transmission schemes and can therefore be applied to Bit Interleaved Coded Modulations (BICM), parallel channels or Hybrid-Automatic Repeat reQuest (HARQ).

In this paper we propose an adaptive Joint Source and Channel Coding (JSCC) associated with adequate modulation algorithm to provide a near channel capacity video transmission rate to achieve better system performance. In the proposed algorithm we apply MPEG-2 video coding scheme, and turbo channel coding technique as well as BPSK/QPSK for modulation method. Rates assigned to MPEG-2 source coding and turbo channel coding schemes are based on the feedback information from Performance Control Unit (PCU) based on channel capacity limitation. Therefore, an improved performance for the suggest system design is promised over Rayleigh fading channel conditions.

Contents 5.4 Visual results. [a] Frame no. 435 of the Foreman sequence -original; [b] Frame no. 435 of the Foreman sequence -MPEG-4 -no RoHC, no JSCC; [c] Frame no. 435 of the Foreman sequence -MPEG-4 -RoHC, JSCC. . . . 81 Contents ix 5.5 Examples of frames obtained in: Scenario 2 -classical (a) and adapted approach (b); Scenario 4 -classical (c) and adapted approach (d); Scenario

Standardized wireless transmissions include fixed-size fragments per packet. Relying on this structure, two schemes for distributing redundancy across fragments are considered in this paper to enable blind identification without higher-order statistics of received signals. We prove that both schemes guarantee blind identifiability as well as symbol detectability and propose a subspace based blind identification method. We test their relative merits, and compare them with existing alternatives using simulations.

This report compares a variety of computable information measures for finite strings. These include Shannon's n-block entropy, the three best known versions of the Lempel-Ziv production complexity (LZ-76, LZ-77, and LZ-78), and the lesser known T-entropy. We apply these measures to strings of known entropy, each derived from the logistic map. Pesin's identity allows us to deduce corresponding Shannon entropies (Kolmogorov-Sinai entropies) for the sample strings, without resorting to probabilistic methods.

In this paper, we applied Continuous Phase Frequency Shift Keying (CPFSK) to Trellis Coded Quantization/Modulation (TCQ/TCM) and thus we called Trellis Coded Quantization/ Continuous Phase Modulation (TCQ/TCCPM) for this new system. In this new scheme we use continuous phase frequency shift keying signal set instead of phase shift keying signal set. As an example, an eight state TCQ/TCCPM system is designed and its error performance is evaluated for different Rician fading parameters K and signal to noise ratios.

The entropy of an ergodic source is the limit of properly rescaled 1block entropies of sources obtained applying successive non-sequential recursive pairs substitutions [7], . In this paper we prove that the cross entropy and the Kullback-Leibler divergence can be obtained in a similar way.

The entropy of an ergodic source is the limit of properly rescaled 1block entropies of sources obtained applying successive non-sequential recursive pairs substitutions [7], . In this paper we prove that the cross entropy and the Kullback-Leibler divergence can be obtained in a similar way.

The energy-distortion function (E(D)) for a network is defined as the minimum total energy required to achieve a target distortion D at the receiver without putting any restrictions on the number of channel uses per source sample. E(D) is studied for a sensor network in which multiple sensors transmit their noisy observations of a Gaussian source to the destination over a Gaussian multiple access channel with perfect channel output feedback. While the optimality of separate source and channel coding is proved for the case of a single sensor, this optimality is shown to fail when there are multiple sensors in the network. A network with two sensors is studied in detail. First a lower bound on E(D) is given. Then, two achievability schemes are proposed: a separation based digital scheme and a Schalkwijk-Kailath (SK) type uncoded scheme. The gap between the lower bound and the upper bound based on separation is shown to be a constant even as the total energy requirement goes to infinity in the low distortion regime. On the other hand, as the distortion requirement is relaxed, the SK based scheme is shown to outperform separation in certain cases, proving that the optimality of source-channel separation does not hold in the multi-sensor setting.

We present a method for separating collided signals from multiple users in the presence of strong and wideband interference/jamming signal. More specifically, we consider a massive connectivity setup where few, out of a large number of users, equipped with spreading codes, synchronously transmit symbols. The received signal is a noisy mixture of symbols transmitted through users' flat fading channels, impaired by fast frequency hopping jamming signal of relatively large power. In the absence of any conventional technique suitable for the considered setup, we propose a "model-driven" deep learning method, based on convolution neural network, to suppress jamming signal from the received signal, and detect active users together with their transmitted symbols. A numerical study of the proposed method confirms its effectiveness in scenarios where classical techniques fail. As such, in a two user scenario with wideband jamming signal of power 20 dB above the power any active user, the proposed algorithm achieves error rates 10 −2 for a wide range of AWGN variances.

HAL is a multidisciplinary open access archive for the deposit and dissemination of scientific research documents, whether they are published or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. L'archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d'enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

We study the MIMO broadcast channel (BC) with perfect channel state information (CSI) at the mobiles and different types of CSI at the base station, i.e. without CSI, with long-term CSI and with CQI. The antennas at the base station are used to apply an orthogonal space-time block code (OSTBC). At the receiver a linear MMSE detector is employed. We first study the optimal transmit and scheduling strategies for these three cases using the weighted sum minimum mean-square error (WSMSE) as the performance criterion. Afterwards, we analyze the impact of the user locations on the achievable minimum WSMSE. Depending on the scheduling strategy (minimum or maximum fairness), having the users rather more spread out within the cell is either helpful or harmful in order to minimize the WSMSE.

The overall energy required to digitize a given physical source can be comparable to the energy required for communication of the produced information bits, especially in cyber-physical sensing systems where radio links are short. When energy is at a premium, this fact calls for energy management solutions that are able to properly allocate the available energy over time between source and channel coding tasks. Energy management is particularly challenging for devices that operate via energy-harvesting, since the controller has to operate without full knowledge of the energy availability in the future. This work addresses the problem of energy allocation over source digitization and communication for a single energy-harvesting sensor. First, optimal policies that minimize the average distortion under constraints on the stability of the data queue connecting source and channel encoders are derived. It is shown that such policies perform independent resource optimizations for the source and channel encoders. The drawback of these policies is that they require an arbitrarily large battery to counteract the variability of the harvesting process and an infinite data queue to mitigate temporal variations in source and channel qualities. Suboptimal policies that do not have such drawbacks are then investigated as well, along with the optimal trade-off distortion vs. delay, which is addressed via dynamic programming tools.

HMM applications show that they are an effective and powerful tool for modelling especially stochastic signals. For this reason, we use HMM for Timit phoneme recognition. The main goal is to study the performance of an HMM phoneme recognizer to fix on an optimal signal parameters. So, we apply different techniques of speech parameterization such as MFCC, LPCC and PLP. Then, we compare the recognition rates obtained to check optimal features. We varied coefficient number of each sample from 12 to 39 for all features. Experimental results show that 39 PLP is the most appropriate parameters for our recognizer. Keywords— HMM, HTK, LPCC, MFCC, PLP, TIMIT

Lossy multicasting of a set of independent, discrete-time, continuous-amplitude source components under the Mean Square Error distortion criterion over binary symmetric broadcast channels is investigated. We considered the practically appealing concatenation of successive refinement source coding with broadcast coding and, specifically, time-sharing of linear binary codes. Three different system optimization criteria are formulated for the lossy multicasting problem. The resulting system optimization is fairly general and applies to a variety of combinations of successive refinement source codes and channel codes. In particular, we investigate in depth the system optimization for a class of channel optimized quantization with successive refinement, obtained by using standard embedded scalar quantizers and linear mapping of the (redundant) quantizer bit-planes onto channel codewords by using a systematic Raptor encoder. This scheme is referred to as Quantization with Linear Index Coding (QLIC). Unlike existing literature on progressive transmission with unequal error protection or channel optimized quantization, we focus on the regime of moderate-to-large code block length and leverage the power of modern sparse-graph codes with iterative Belief Propagation decoding. In this regime, the system optimization takes on the form of simple convex programming that reduces to linear programming for QLIC. The performance of QLIC compares favorably with respect to the state of the art channel optimized quantization in the conventional setting of a single Gaussian source over a binary symmetric channel. For the multicast scenario, we quantified the performance gap incurred by the practical QLIC design with respect to ideal source and channel codes.

Abstract—We consider the transmission of a real iid Gaussian source over a quasi-static MIMO fading channel. The relevant performance criterion is end-toend average quadratic distortion D versus channel SNR, for given spectral efficiency η, defined as the ratio of the source bandwidth over the channel bandwidth. In the limit of high-resolution (vanishing distortion) for high-SNR, we define the distortion SNR exponent a (η) for a family of source-channel coding schemes as the a=− log D/log snr. In our recent paper [1], we have ...

In this paper, delay-limited transmission of Quasi-stationary sources over block fading channels is considered. Considering distortion outage probability as the performance measure, two source and channel coding schemes with power adaptive transmission are presented. The delay limited transmission of a quasi-stationary source over a block fading channel is from the perspective of source and channel coding design and it utilizes performance scaling laws. Because In a Quasi-stationary sources time utilization takes a major role. For that the first one is optimized for fixed rate transmission, and hence enjoys simplicity of implementation. The second one is a high performance scheme, which also benefits from optimized rate adaptation with respect to source and channel states. In high SNR regime, the performance scaling laws in terms of outage distortion exponent and asymptotic outage distortion gain are derived, where two schemes with fixed transmission power and adaptive or optimized fixed rates are considered as benchmarks for comparisons. Various analytical and numerical results are provided which demonstrate a superior performance for source and channel optimized rate and power adaptive scheme.

Serially concatenated and iteratively decoded Irregular Variable Length Coding (IrVLC) combined with precoded Fast Frequency Hopping (FFH) M-ary Frequency Shift Keying (MFSK) is considered. We employ EXtrinsic Information Transfer (EXIT) charts to investigate the 3-stage concatenation of the FFH-MFSK demodulator, the rate-1 decoder and the outer IrVLC decoder. The proposed joint source and channel coding scheme is capable of operating at low Signal-to-Noise Ratio (SNR) in Rayleigh fading channels contaminated by Partial Band Noise Jamming (PBNJ). The IrVLC scheme is comprised of a number of component Variable Length Coding (VLC) codebooks employing different coding rates for encoding particular fractions of the input source symbol stream. These fractions may be chosen with the aid of EXIT charts in order to shape the inverted EXIT curve of the IrVLC codec so that it can be matched with the EXIT curve of the inner decoder. We demonstrate that using the proposed scheme an infinitesimally low bit error ratio may be achieved at low SNR values.

This is a master thesis from a collaboration between Oslo University College and Uninett Research. Uninett have a passive monitoring device on a 2.5 Gbps backbone link between Trondheim and Narvik. They uses measurement with optical splitters and specialized measuring interfaces to trace traffic with Gigabit speed. We would like to investigate the structure and patterns in these data. It is of special interest to classify the traffic belonging to different services and protocols.

A novel adaptive source-channel coding with feedback for progressive transmission of medical images is proposed here. In the source coding part, the transmission starts from the region of interest (RoI). The parity length in the channel code varies with respect to both the proximity of the image subblock to the RoI and the channel noise, which is iteratively estimated in the receiver. The overall transmitted data can be controlled by the user (clinician). In the case of medical data transmission, it is vital to keep the distortion level under control as in most of the cases certain clinically important regions have to be transmitted without any visible error. The proposed system significantly reduces the transmission time and error. Moreover, the system is very user friendly since the selection of the RoI, its size, overall code rate, and a number of test features such as noise level can be set by the users in both ends. A MATLAB-based TCP/IP connection has been established to demon...

This paper describes a technique for enhancing the Mel-filtered log spectra of noisy speech, with application to noise robust speech recognition. We first compute an SNR-based soft-decision mask in the Mel-spectral domain as an indicator of speech presence. Then, we exploit the known time-frequency correlation of speech by treating this mask as an image, and performing median filtering and blurring to remove the outliers and to smooth the decision regions. This mask constitutes a set of multiplicative coefficients (ranging in [0,1]) that are used to discard the unreliable parts of the Mel-filtered log-spectrum of noisy speech. Finally, we apply Log-Spectral Flooring [1] on the liftered spectra of both clean and noisy speech so as to match their respective dynamic ranges and to emphasize the information in the spectral peaks. The noisy MFCCs computed on these modified log-spectra show an increased similarity with their corresponding clean MFCCs. Evaluation on the Aurora-2 corpus shows that the proposed approach competes with state-of-the-art front-ends, like ETSI-AFE, MVA or PNCC.