A Distributed Hybrid Filter for target tracking in sensor networks

In this paper, we propose the Distributed Hybrid Filter (DHF) to overcome the performance degradation in Sigma Point Filters caused by bad linearization. A quantifiable criterion to judge the quality of linearization will be given. If the linearization quality is unsatisfactory, a sensor will use some other local processing method which does not rely on linearization. The DHF can accommodate multiple sensors that are using different local processing methods and is shown to have superior performance to the Distributed Sigma Point Information Filter (DSPIF).