Exploring distributed and adaptive compression in cluster-based sensor routing

Distributed compression is appealing to sensor networks due to its ability to compress out the redundancy within individual sources as well as across geographically-separated but content-correlated sources without explicit inter-source communication. This paper investigates how adaptive distributed compression can be efficiently integrated with the practical cluster-based sensor architecture in general and the low-energy adaptive clustering hierarchy (LEACH) protocol in particular. The goal is to exploit the power of distributed compression while keeping the operational overhead minimal. Using an adaptive scheme based on rate-compatible punctured convolutional codes, we show that an additional energy reduction of 20%-41% is achievable compared to conventional compression without modifying the routing protocol or complicating the system architecture.