Distributed interval estimation for sensor networks with energy-limited attacks and stochastic communication protocols

Abstract

In this paper, the problem of the distributed interval estimation is investigated for wireless sensor networks under the energy-limited attack and the stochastic communication protocol (SCP). A model with randomness and restrictions is constructed to describe the energy-limited attack, which is considered to occur in two positions, one is the transmission process of sensors and local estimators (LEs), and another is the transmission process of LEs and their neighbor nodes. Furthermore, to mitigate the communication burden, the SCP is adopted to distribute network access rights to sensor nodes. All in all, the objective of this paper is to develop a distributed interval estimator (DIE) which can use local and neighbor information to generate an interval that covers all potential estimation states when SCP, bounded noises, and energy-limited attacks are all present at the same time. To achieve this, the positive system theory and the stability analysis theory are used to obtain some sufficient conditions for the error dynamic system to be positive system and stable, and the needed DIE gains are achieved by solving the linear matrix inequality. Finally, an example is presented to demonstrate the effectiveness and superiority of the offered method.