Power Allocation Based on Instantaneous Observation Quality for Distributed Detection in Wireless Sensor Networks

Proper Power allocation can improve energy efficiency in wireless sensor networks (WSNs). Previous works considered only the channel statistics and did not take the instantaneous observations into consideration. Applying the instantaneous quality of an observation to the transmission power control of a node can improve the overall energy efficiency. In this work, we proposed nearly optimal power allocation schemes based on the instantaneous observation quality in each node for distributed detection in WSNs. The instantaneous quality of an observation is related to the log-likelihood ratio (LLR) of the signal observed by a node. According to the simulation results, at least 20 % of energy can be saved when compared to the equal power allocation scheme.