Distributed State Estimation in Rotational Shiftwork Sensor Networks with Communication Constraint

In this paper, a distributed state estimation problem is studied for a new class of rotational shiftwork sensor network with communication constraint. The main feature of such novel sensor networks is that only incomplete information is available for each sensor: (i) part of states of target plant can be detected; (ii) information exchanged among sensors suffered from randomly part loss. In this situation, the addressed distributed estimation problem would become more difficult, since less estimation information of the target system can be used. In practice, the batteries equipped for sensors are limited and non-replaceable. To reduce the power consumption, the sensors investigated here have two working modes: active and hibernated modes. In active mode, sensors can behave in a normal way, while in hibernated mode, a sensor has to cut off all the communication with environment to save energy. Two working modes alternate via a random variable. By resorting to Lyapunov functional method, a condition is derived for designing a distributed state estimator and ensure that the target plant can be estimated in mean square under certain criteria. In addition, a performance analysis of the shiftwork sensor networks is investigated.