This paper mainly investigates the coordi-nated anti-jamming channel access problems in multi-user scenarios where there exists a tracking jammer who senses the spectrum and traces the channel with maximal receiving power.To cope with the challenges brought by the tracking jammer,a multi-leader one-follower anti-jamming Stackelberg (MOAS) game is formulated,which is able to model the complex inter-actions between users and the tracking jammer.In the proposed game,users act as leaders,chose their chan-nel access strategies and transmit firstly.The tracking jammer acts as the follower,whose objective is to find the optimal jamming strategy at each time slot.Be-sides,the existence of Stackelberg equilibriums (SEs)is proved,which means users reach Nash Equilibriums(NEs) for each jamming strategy while the jammer finds its best response jamming strategy for the cur-rent network access case.An active attraction based anti-jamming channel access (3ACA) algorithm is de-signed to reach SEs,where jammed users keep their channel access strategies unchanged to create access chances for other users.To enhance the fairness of the system,users will adjust their strategies and re-learn after certain time slots to provide access chances for those users who sacrifice themselves to attract the tracking jammer.