Opportunistic Power Allocation Schemes for Wireless Sensor Networks

In this paper, we propose and analyze an opportunistic power allocation (OPA) scheme suitable for decentralized parameter estimation in WSNs. In this scheme, only sensors experiencing favorable channels conditions (i.e. above a channel gain threshold) participate in the estimation process by adjusting their transmit power on the basis of local channel state information (CSI). In particular, we address two different problems: the minimization of distortion under a sum-power constraint, and the minimization of transmit power for a given target distortion. For each design, we (1) derive a closed-form expression of the channel gain threshold on the basis of statistical CSI only, and (2) define the corresponding power allocation rule. Then we assess the performance of the proposed OPA scheme by means of computer simulations, and we carry out a comparison with other water-filling (i.e. optimal) and uniform power allocation (UPA) solutions. Results are given in terms of attained distortion level, transmit power, ability to exploit multi-user diversity and number of FC-sensor channels required.