Sensor coordination is an important technical means for combat aircraft to achieve mission performance and RF stealth performance balance. Aiming at the problem that the traditional guided search method is difficult to solve the multi-shot information guided search problem of the data link, the probabilistic method is used to establish the relationship between the multi-shot continuous guidance information and the cumulative discovery probability and the cumulative intercept probability. A multi-step joint optimization based on tracking accuracy and tracked positioning accuracy is proposed for the singlestep decision problem of covariance based linear programming collaborative tracking method. The Markov decision process is introduced to model the typical air combat dynamic process of airborne sensor collaborative search and cooperative tracking, and the radar RF stealth performance optimization under task performance constraints is realized. The simulation results of typical air combat scenarios show that the proposed optimal control strategy can reduce the radar radiation time by more than 43% compared with the empirical control strategy.