Distributed Successive Approximation Coding using Broadcast Advantage: The Two-Encoder Case

TL;DR

This paper introduces a novel two-encoder distributed coding scheme leveraging broadcast advantages in wireless sensor networks, achieving successively refinable rate-distortion performance and potential energy savings.

Contribution

It presents a new ping-pong inspired successive approximation coding scheme that exploits broadcast links, extending to multiple encoders and enabling energy-efficient distributed image coding.

Findings

Scheme is successively refinable on the rate-distortion surface.

Potential energy savings over traditional distributed coding.

Extension to multiple encoders and applications in wireless image coding.

Abstract

Traditional distributed source coding rarely considers the possible link between separate encoders. However, the broadcast nature of wireless communication in sensor networks provides a free gossip mechanism which can be used to simplify encoding/decoding and reduce transmission power. Using this broadcast advantage, we present a new two-encoder scheme which imitates the ping-pong game and has a successive approximation structure. For the quadratic Gaussian case, we prove that this scheme is successively refinable on the {sum-rate, distortion pair} surface, which is characterized by the rate-distortion region of the distributed two-encoder source coding. A potential energy saving over conventional distributed coding is also illustrated. This ping-pong distributed coding idea can be extended to the multiple encoder case and provides the theoretical foundation for a new class of distributed image coding method in wireless scenarios.