In this paper, the distributed estimation of an unknown target parameter through a power-constrained Wireless Sensor Networks (WSN) including clusters is investigated. Data collection algorithm includes routing and scheduling. In WSN, sensor nodes observe the environmental events, perform a local data compression and at the first step, forward the results to a cluster head (here it is called fusion center). In this study, Cluster Head (CH) provides a final estimation of the observed target parameter using collected data. Then CH sends the estimation results to the sink. The main purpose of the proposed algorithm is to extend network lifetime while the target parameter is estimated with the desirable precision. To achieve this goal, following three objects are regarded: 1) proposing estimation model for different target parameter types, 2) proposing intra cluster scheduling/routing algorithm for collecting data inside the clusters according to the estimation model requirements and 3) proposing inter cluster routing in order to transmit the estimation process results to the sink. According to the inherent characteristics of WSN applications, we consider the unknown target parameter as a random variable with a known prior distribution function. Based on the application, different prior functions are admissible. In this paper we investigate the impact of different prior functions on estimation process performance. Simulation results show that the proposed algorithm achieves its goals.

Keywords