The unreasonable connectivity of the Wireless Sensor Network(WSN) would cause an error in calculating the distance between the node to be tested and the known node. An improved Artificial Immune Algorithm(AIA) is proposed to optimize DV-Hop unknown node coordinates. Firstly, the original average hop distance is weighted, and then the deviation value generated by the distance between the beacon nodes in the network is utilized to construct the hop distance correction value to obtain the final average network hop distance. Finally, AIA is introduced to the calculation of the coordinates of the nodes to be tested. Because the AIA is easy to fall into local optimum and the convergence speed is too slow, the Gaussian variation method is adopted to improve the AIA in the local search process, and the scope of search is expanded to get optimized node coordinates to be tested. The Matlab simulation proves that compared with the original DV-Hop algorithm, the average positioning error of the improved algorithm in the total number of nodes, the proportion of beacon nodes and the communication radius is reduced by about 15%. The improved algorithm has higher positioning accuracy, better stability and convergence.