Distributed Interval Type-2 Fuzzy Filtering for Wireless Sensor Networks With Intermittent Measurements

Though wireless sensor networks (WSNs) help to enhance data fusion accuracy, measurement exchanges between sensors are not always reliable due to inherent open communication channels. In this paper, a distributed interval type-2 (IT2) fuzzy filter is presented in the context of WSNs with intermittent measurements. Firstly, IT2 fuzzy models are employed to formulate one type of nonlinear systems with parameter uncertainties. Then, dual random data packet dropouts phenomena are considered, including the measurements transmitting from sensors to data fusion centers, and the measurements transmitting from data fusion centers to distributed filters. Bernoulli variables are adopted to depict the random data packet dropouts. Furthermore, to guarantee the robust mean-square asymptotic stability of the filtering error system, less conservative sufficient conditions are derived to seek for distributed filter gains. Finally, simulation results on Henon mapping systems with parameter uncertainties verify the robustness of the presented distributed fuzzy filter for WSNs with intermittent measurements.

Automated summary

In this paper, a distributed interval type-2 fuzzy filter is proposed for wireless sensor networks with intermittent measurements. IT2 fuzzy models are used to describe nonlinear systems with parameter uncertainties, and dual random data packet dropouts are considered. Bernoulli variables are employed to depict the random data packet dropouts, and less conservative sufficient conditions are derived to seek for distributed filter gains to guarantee the robustness of the filtering error system. The simulation results on Henon mapping systems with parameter uncertainties validate the robustness of the distributed fuzzy filter for WSNs with intermittent measurements.