Abstract: Chaotic particle swarm optimization (C-PSO) was proposed based on chaos theory to optimize the optimization index of Ad Hoc network. In the process of network optimization, C-PSO algorithm took full advantage of chaotic system in the randomness, ergodicity and sensitivity to avoid the “precocious” phenomenon of PSO algorithm, to avoid falling into the local optimal area, and to enhance the global collection capacity. Based on the NS-3 simulation system, the simulation results of C-PSO algorithm and PSO algorithm were tested.Through comparing and analyzing the three network performance indexes of packet loss rate, network life cycle and network throughput, the results show that C-PSO algorithm is superior to PSO algorithm, which verifies the validity and reliability of C-PSO algorithm for Ad Hoc network optimization. The algorithm can be applied to realize optimization of Ad Hoc network.