Sensor data forwarding strategies for state estimation in multi-hop wireless networks

State estimation is of great importance in various applications based on wireless networked control systems. Since the wireless sensor nodes are typically equipped with power-limited battery, they are only able to emit relatively low-power signal which constrains the reliable communication distance between the transmitter and the receiver. One effective way to tackle such situation is to forward the transmitted packet through relay nodes. In this paper, we investigate the data forwarding strategy design for accurate remote state estimation in multi-hop wireless networks. Based on the computational capability of relay nodes, we first propose two relay strategies, namely, Direct Forwarding Strategy (DFS) and Local Processing and Forwarding Strategy (LFS). Necessary and sufficient conditions of estimation stability are derived for these two strategies respectively. We also prove that LFS is always better than DFS in terms of estimation accuracy at the expense of more energy consumption. We further propose an Event-triggered Forwarding Strategy (EFS) which is able to balance the estimation accuracy and relay energy consumption. Numerical examples are employed to demonstrate the effectiveness of our design.