Optimal data compression and forwarding in wireless sensor networks

In this letter, we present a Linear Programming framework for modeling dynamic data compression and decompression in conjunction with flow balancing in wireless sensor networks. Using the developed framework, we investigated the sensor network lifetimes for different network sizes with various data compression and flow balancing strategies. Our results show that neither compressing all data nor avoiding data compression completely can achieve the longest possible network lifetime. Dynamic data transformation is shown to achieve significantly longer network lifetimes than the lifetimes obtained with the two pure strategies above.

I Introduction

In sensor networks, a collective energy optimization strategy results in a longer network lifetime than the lifetime obtained with a strategy of minimizing the energy dissipation of each node individually [1]. It was shown in [2] that if the nodes optimize their data flows to achieve a network-wide energy balancing, then the achievable lifetime will be prolonged.