Distributed estimation scheme based on cooperative communication in wireless sensor networks

In the distributed estimation of an unknown source in wireless sensor networks, the observations are transmitted through fading channels. Thus channel fading influences the estimation distortion significantly. Recently, cooperative communication has been proposed to combat channel fading by providing multiple independently fading paths. In this letter, the authors use cooperative communication to decrease estimation distortion. First, they propose a distributed estimation scheme with amplify-and-forward cooperation strategy. To minimise distortion, the optimal power allocation of the individual node transmission power is studied, in which the maximum number of active sensor nodes is limited so as to restrict slot overhead. The optimal solution and a sub-optimal solution with less computation overhead are given. Numerical results show that, the proposed scheme could achieve a lower estimation distortion even in the case of using fewer slots, when compared with the conventional estimation scheme with non-cooperative transmission. They also show that our scheme could achieve full estimation diversity.