The traditional multi-dimensional data optimized storage methods have the problems of poor storage performance and low signal-to-noise ratio. Therefore, a network multidimensional data optimal storage method based on March algorithm is proposed. According to the mapping function, the network multidimensional data is encoded by integer mapping function.The coordinate values of multidimensional array are obtained by mapping to complete the construction of dimensional data organization. Higher storage efficiency of network multidimensional data is obtained according to binary coding; Sequential storage and block storage are used to organize measurement data, and the occupied space of data storage is saved through block compression storage. Use March algorithm to test the network multidimensional data storage method and complete the optimization of the storage method. The experimental results show that the maximum signal-to-noise ratio of this method in the data storage process can reach 96 at different compression ratios. At different sparsity, the signal-to-noise ratio of this method is still high in the data storage process, and has better data storage performance.