Correlated Sources Over Wireless Channels: Cooperative Source-Channel Coding

Abstract

Abstract-We consider wireless sensor networks deployed to observe arbitrary random fields. The requirement is to reconstruct an estimate of the random field at a certain collector node. This creates a many-to-one data gathering wireless channel. One of the main challenges in this scenario is that the source/channel separation theorem, proved by Shannon for point-to-point links, does not hold anymore. In this paper, we construct novel cooperative source-channel coding schemes that exploit the wireless channel and the correlation between the sources. In particular, we differentiate between two distinct cases. The first case assumes that the sensor nodes are equipped with receivers and, hence, every node can exploit the wireless link to distribute its information to its neighbors. We then devise an efficient deterministic cooperation strategy where the neighboring nodes act as virtual antennas in a beamforming configuration. The second, and more challenging, scenario restricts the capability of sensor nodes to transmit only. In this case, we argue that statistical cooperative source-channel coding techniques still yield significant performance gains in certain relevant scenarios. Specifically, we propose a low complexity cooperative source-channel coding scheme based on the proper use of low-density generator matrix codes. This scheme is shown to outperform the recently proposed joint source-channel coding scheme (Garcia-Frias et al., 2002) in the case of highly correlated sources. In both the deterministic and statistical cooperation scenarios, we develop analytical results that guide the optimization of the proposed schemes and validate the performance gains observed in simulations.

Type based estimation over multiaccess channels

“…He showed that transmitting uncoded observations in the Gaussian MAC gives the best scaling law. Other relevant work includes source compression for detection or estimation under communication rate constraints (e.g., [4]- [8]) and joint source-channel coding for the MAC (e.g., [9]- [11]). The information theoretic approach considers the asymptote that the amount of source data/channel resources is large.…”

Communication Via Decentralized Processing

“…Many problems are analyzed within an information theoretic setting, such as distributed source coding, CEO [4] and sensor networks. A connection between these rate-distortion problems to network scenarios is reported in [5], [6], [7]. The CEO setting is also addressed by tools of estimation theory (where it is commonly referred to as fusion center), such as [8], [9] and [10].…”

Sending a Bivariate Gaussian Over a Gaussian MAC

“…For this problem he derived sufficient conditions for the achievability of a distortion pair. More recently, another variation where the two source components are binary with Hamming distortion and where the multiple-access channel is Gaussian was considered by Murugan, Gopala and El Gamal [10] who derived sufficient conditions for the achievability of a distortion pair. Gastpar [11] considered a combined source-channel coding analog of the quadratic Gaussian CEO problem.…”