ADP-Based Optimal Adaptive Waveform Selection in Cognitive Radar

The problem of adaptive waveform selection can be viewed as a problem of stochastic dynamic programming. However, the computational cost of solution of optimality equations is very high as a result of large state variable space. So far few suitable scheduling algorithms have been proposed. To account for the aboved shortcoming, we can use approximate dynamic programming (ADP) to solve adaptive waveform selection problem. The scheduling algorithm we proposed can minimize the time taken to detect new targets, detect these targets in accordance with importance, and minimize the use of radar resources. Meanwhile we have a lower computational cost compared with the traditional algorithms. Finally, the whole paper is summarized.