Abstract: As a new network architecture, Software Defined Network (SDN) provides flexibility and centralized control for network traffic management. However, with the development and popularity of SDN, various high-traffic service software emerge endlessly, bringing an increasing amount of data, followed by the demand for efficient and reliable path calculation algorithm, which brings great challenges to the network load. Therefore, how to effectively improve the efficiency of data transmission has become one of the important research directions of SDN. This paper studies the problem of optimal path planning in SDN, and proposes an improved path planning method based on bidirectional search-enhanced Dijkstra algorithm. The algorithm improves the efficiency of path search by starting the search from the source point and the destination point at the same time and meeting somewhere in the middle. This algorithm is implemented in a simulated SDN environment and compared with the current common algorithms. The results show that the improved bidirectional search Dijkstra algorithm is more efficient in calculating the shortest path, especially in large networks. In addition, the network link weight is dynamically updated during operation, which verifies the robustness and adaptability of the algorithm.