Practical Distributed Source Coding Research Articles

Distributed source coding for utilization of inter/intra source correlation

In many practical distributed source coding (DSC) applications, correlation information plays several important roles at decoding, which provides the prior knowledge of the source statistics for the decoding algorithms to start work. Therefore, the accurate utilization of such correlation information has a significant impact on decoding performance. In this work, a hybrid DSC (HYDSC) scheme is proposed to generate the “soft” information about the clean source symbols, which allows the channel code-based DSC scheme to effectively utilize the inter/intra source correlation without requiring prior knowledge of its statistics, so as to further improve the decoding performance. Further performance improvement, at the cost of moderate increasing in computational complexity, can be achieved by performing additional iterations on the basis of the HYDSC scheme. We mainly focus on, in particular, applications where Slepian–Wolf coding and usual source coding are jointly used to exploit both the correlation among the adjoining source symbols and the correlation between the sources. Experiment results indicate that the presented scheme not only show superior performance from an error correction viewpoint, but also achieve better compression ratio in the case of strong internal correlation of the source sequence, as compared with the previous DSC techniques.

A Sampling-based Distributed Source Coding Scheme

In many practical distributed source coding (DSC) schemes that apply channel coding and iterative decoding, correlation estimation provides the likelihood information about the source bits to initialize the decoding algorithms. Therefore, the decoding depends strongly on the accuracy of this correlation estimation information. In this study, a sampling-based scheme for DSC is proposed that takes advantage of the source redundancy without requiring prior knowledge of its statistics. At the sender, the source sequence is sampled to obtain the sampled and unsampled sub-sequences. Then, the un-sampled sub-sequence is compressed by an arithmetic coder. Meanwhile the syndromes of the sampled sub-sequence are also calculated. At the receiver, the correlation between the side information and the un-sampled sub-sequence is used to estimate the conditional marginal distribution of the sampled sub-sequence. Then, the estimated information is used to initialize the likelihood information of the decoding algorithms. Finally, the likelihood information is combined with the received syndromes to perform iterative decoding to recover the original sampled sub-sequence. Experiment results show that the proposed scheme achieves better rate-distortion performance compared with the existing DSC schemes.

Enhanced Correlation Estimators for Distributed Source Coding in Large Wireless Sensor Networks

In this paper, we propose two estimators based on correlation parameters for the two key steps of a practical distributed source coding (DSC) scheme, namely: 1) the computation of the side-information at the receiver side and 2) the estimation of the required number of bits to compress the readings in order to guarantee a certain symbol error probability. We show that using the proposed estimators, the DSC algorithm performs better in terms of the compression rate and the symbol error rate. In particular, this improvement is especially significant when the number of snapshots used in the training phase is only slightly larger than the observation vector. However, when the number of snapshots is much higher than the observation dimension, our proposed estimators perform similarly to the classical estimators.

Sampling-Based Correlation Estimation for Distributed Source Coding Under Rate and Complexity Constraints

In many practical distributed source coding (DSC) applications, correlation information has to be estimated at the encoder in order to determine the encoding rate. Coding efficiency depends strongly on the accuracy of this correlation estimation. While error in estimation is inevitable, the impact of estimation error on compression efficiency has not been sufficiently studied for the DSC problem. In this paper,we study correlation estimation subject to rate and complexity constraints, and its impact on coding efficiency in a DSC framework for practical distributed image and video applications. We focus on, in particular, applications where binary correlation models are exploited for Slepian-Wolf coding and sampling techniques are used to estimate the correlation, while extensions to other correlation models would also be briefly discussed. In the first part of this paper, we investigate the compression of binary data. We first propose a model to characterize the relationship between the number of samples used in estimation and the coding rate penalty, in the case of encoding of a single binary source. The model is then extended to scenarios where multiple binary sources are compressed, and based on the model we propose an algorithm to determine the number of samples allocated to different sources so that the overall rate penalty can be minimized, subject to a constraint on the total number of samples. The second part of this paper studies compression of continuous valued data. We propose a model-based estimation for the particular but important situations where binary bit-planes are extracted from a continuous-valued input source, and each bit-plane is compressed using DSC. The proposed model-based method first estimates the source and correlation noise models using continuous valued samples, and then uses the models to derive the bit-plane statistics analytically. We also extend the model-based estimation to the cases when bit-planes are extracted based on the significance of the data, similar to those commonly used in wavelet-based applications. Experimental results, including some based on hyperspectral image compression, demonstrate the effectiveness of the proposed algorithms.