Distributed power allocation for cognitive tracking based on non-cooperative game in decentralized netted radar.

The traditional centralized power allocation method for cognitive tracking in the netted radar can achieve the best resource utilization efficiency and target tracking performance. However, this method has poor robustness, low fault tolerance, and a high bandwidth requirement. In order to address this challenge, we propose a decentralized distributed power allocation method based on the non-cooperative game for multi-target tracking. Firstly, we model the power allocation problem for decentralized netted radar as a non-cooperative game problem and consider each radar node in the radar network as a player in the game. Then, we employ the determinant of Fisher information matrix to evaluate target tracking accuracy and use it to construct the utility function of the non-cooperative game based on the game rule. Finally, we propose an iterative distributed computing method to solve the Nash equilibrium. Meanwhile, we theoretically prove the Nash equilibrium solution's existence and uniqueness. Numerical simulation results show that the tracking performance of the proposed method is superior to the uniform and random power allocation method. Simultaneously, the tracking performance of the proposed method is close to the centralized power allocation method with a much lower computational load. [ABSTRACT FROM AUTHOR]