In order to jam overall radar net best, to address problem that the radar jammer power is allocated optimally to all radars in radar net, the two-person zero-sum game theory (TPZS) is proposed to distribute jamming power to radars. By expansion of jamming vector, emission vector and other vectors, the vectors are transformed to matrixes. Jammer jams each radar in network by certain powers by calculations with TPZS theory based on mutual information, the power distribution can be calculated in the case of different distances, different suppression coefficient of the radars. Under the condition that the total power of jammer is limited, the jamming power can distributed to the radars in network reasonably. Simulation shows in most cases that the needful jamming power of TPZS is more less than regular proportion method in condition that the radar net can be jammed effectively. For the jammer, radars in network are different more big, the TPZS method can embodies its superiority more. Since the mutual information theory of TPZS can be used to characterize detecting overall effect of radar net comprehensively and systematically, the jamming want to reduce the amount of the mutual information by jamming with TPZS for power allocation. Based on the guideline, the allocation of jamming resource is targeted and the overall jamming effect is good.