At the present stage wireless sensor network exists some deficiencies such as finite energy, limited communication capability, multi-hop routing and dynamic topology. In addition too many nodes which are distributed intensively influence also performance of wireless sensor network. Besides distributed optimization problem always impact to efficiency of the network. For data processing and routing selection of wireless sensor network, a new routing algorithm based on mobile agent and ant colony optimization is proposed in the paper. The ant colony algorithm is improved and optimized in the paper. The algorithm can balance between routing efficiency and node loading. So on the one part the computing capability of sensor node and efficiency of data processing are improved. On the other part redundant data and communication consumption are reduced so that life expectancy of sensor is extended. Simulation experiments show that the new routing algorithm could solve the problems of wireless sensor network, and it has good practice application effect.