Optimal sensor network configuration based on control theory

Introduction: Recently wireless sensor networks with the function distribution of memory units, communications, calculations etc. have attracted much attention, and there has been increasing research [1, 2]. In these researches, sensor nodes are connected wirelessly and some local estimates are merged into the common estimate by the wireless communication. It is well known that sensor networks are superior to a observation by a system with single sensor in a fault tolerance, load reduction of operator, collection and application of information etc. and its application to not only a sensing system but also the guidance control system via a sensor network has received attention. For example, traffic control systems, nano-medicines and disaster countermeasures. Meanwhile, each sensor node requires electric power more than a case of only sensing because of communications and calculations, but sensor nodes are generally powered and driven by batteries. Moreover it is difficult to change batteries frequently or charge by power cable because of increasing of costs. Therefore, it is important to utilize the energy efficiently to achieve the energy-saving and prolong sensor nodes life. For this requirement, the sensor scheduling, the optimization of the communication rate and communications traffic and decreasing communication distances by the multi-hop communication are discussed [4,3, 5]. Consequently, in this paper, we discuss a sensor scheduling problem selecting available sensor nodes considering the estimation error variance and communication energy in a feedback control system via a sensor network.