Distributed Source Coding PDF Download

Distributed source coding (DSC) refers to the problem of compressing two or more physically-separated information sources, where the sources (e.g. sensors) send the (compressed) information to a central point (e.g. a monitoring station) without communicating to each other. Cooperative communication, also known as user cooperation or the relay channel problem, allows single-antenna users in a multi-user scenario to share antennas with each other to form a virtual multiple-antenna system. These two technologies, originated in the seventies, have caused a resurgence of interest in recent years. In this dissertation, we investigate these two technologies and try to apply distributed source coding into user cooperation. Three topics are studied: The first is to develop a new scheme to attack the noisy-channel Slepian-Wolf coding problem using serially concatenated codes (SCC). The advantage is two-fold: (i) separate refining of compression rate and error protection capability is made possible and (ii) many useful results about SCC can be used to achieve joint optimization. The second topic considers the design of two-user cooperation schemes by exploiting distributed source coding technologies. Two user cooperation schemes, Slepian-Wolf cooperation and Wyner-Ziv cooperation, are proposed, which exploit Slepian-Wolf coding and Wyner-Ziv coding, respectively. Possibly the first practical schemes that implement the idea of compress-and-forward, these schemes offer substantial gains over existing cooperative strategies especially when the inter-user channel is at outage. The third topic is to develop an extension for cooperative communication to High-Speed Packet Downlink Access (HSDPA) cellular system which increases both the overall cell capacity as well as coverage near the cell edge at a relatively low cost.