Abstract:

In response to the need for precision-preserving target tracking in the context of continuous detection, a joint power allocation and node selection algorithm is proposed based on an active-passive radar network. Firstly, employs a parallel extended Kalman filter to estimate and predict the target state. And then, derives a predictive Bayesian Cramér-Rao lower bound （a tracking performance metric） as a function of transmission power and receiving nodes based on the prediction information. Finally, the optimization objective is to minimize the combined power usage and node usage of the active-passive radar, under the constraints of limited total transmission power and preset tracking accuracy. The resource allocation model is formulated for active-passive radar as a mixed-integer programming problem. Exploiting the partial monotonicity and variable separability of the model, a fast decoupling solution algorithm is designed. Simulation results show that, compared to the genetic algorithm, the proposed method achieves an approximately optimal solution with lower computational complexity. Additionally, compared to the uniform distribution scheme, the proposed method can ultimately save 80% of system resources while maintaining accuracy and stable tracking conditions, which demonstrates the effectiveness of the proposed method.

Key words: active-passive radar, target tracking, power allocation, sensor selection