Distributed set-membership filtering with error-based dynamic transmission and power allocation-dependent packet loss

Abstract

Network resource constraint is significant challenge in wireless sensor networks, making it difficult to obtain accurate state information. Currently, two main transmission protocols, Round-Robin and Try-Once-Discard, are used to address this issue. However, these protocols only allow one sensor to transmit at a time, which limits their ability to fully capture relevant system information and often results in conservative performance. To overcome these limitations, an error-based dynamic transmission strategy is proposed for distributed sensor networks. This strategy uses a double selection principle to more accurately identify signals worth transmitting by comparing error differences. The selected sensor data is then packaged and transmitted over the wireless network to other sensor nodes. A power allocation-dependent packet loss model is developed, which takes into account the varying packet lengths resulting from the scheduling protocol. Based on this, a sufficient condition is established to ensure that the filtering error remains within ellipsoidal constraints. Simulation results demonstrate that the proposed transmission strategy offers superior filtering performance compared to traditional method. Additionally, this strategy reduces power consumption by an average of 47% over 20 transmission steps compared to the full transmission approach.