A LOCALIZED RECURSIVE ESTIMATION ALGORITHM FOR VECTOR PARAMETER ESTIMATION IN AD HOC WIRELESS SENSOR NETWORKS

Amir Rastegarnia – Mohammad Ali Tinati – Behzad Mozaffari – Azam Khalili

   In this paper a localized recursive estimation scheme for vector parameter estimation in ad hoc wireless sensor networks (WSNs) is proposed. In our setup, each sensor has a noisy measurement from a fixed, but unknown vector parameter. The classical best linear unbiased estimator (BLUE) cannot be implemented in a practical ad hoc sensor network due to its requirement to transmit all real-valued messages to a fusion center (FC). To address this problem, in this paper we propose a recursive estimation algorithm which is based on progressive cooperation between sensors. In our proposed algorithm, each sensor computes an estimate of the unknown parameter using its own local data and some information from prior sensor. The computed estimate is delivered to the next sensor and so on. The last sensor delivers the computed estimate of the unknown parameter to the destination. Our mathematical analysis shows that the proposed recursive algorithm is an unbiased estimator of the unknown parameter and asymptotically approaches to the estimate that would be obtained if each sensor had access to the information across the entire network. Also, the simulation examples are used to show the performance of the proposed scheme.

Keywords: ad hoc network, distributed estimation, wireless sensor network, BLUE, estimation

[full-paper]