A topology optimization model based on improved cultural gene is proposed to solve the problem of low network robustness caused by neglecting the convergence characteristics of sink nodes in the current topology optimization of wireless sensor networks.Considering the cascading characteristics of wireless sensor networks,a local search operator based on convergence intermediate entropy is designed.The experiment results show that the proposed local search operator can effectively improve the topology optimization process.Compared with simulated annealing algorithm and mountain climbing algorithm,the proposed topology optimization algorithm based on the improved cultural gene model has better comprehensive performance.The simulation results further verify that the proposed model can be used for reference to improve the cascade robustness of wireless sensor networks.