Computing and Informatics

The scale of data shows an explosive growth trend, with wide use of cloud storage. However, there are challenges such as network latency and energy consumption. The emergence of edge computing brings data close to the edge of the network, making it a good supplement to cloud computing. The spatiotemporal characteristics of data have been largely ignored in studies of data placement and storage optimization. To this end, a temperature matrix-based data placement method using an improved Hungarian algorithm (TEMPLIH) is proposed in this work. A temperature matrix is used to reflect the influence of data characteristics on its placement. A data replica matrix selection algorithm based on temperature matrix (RSA-TM) is proposed to meet latency requirements. Then, an improved Hungarian algorithm based on replica matrix (IHA-RM) is proposed, which satisfies the balance among the multiple goals of latency, cost, and load balancing. Compared with other data placement strategies, experiments show that the proposed method can effectively reduce the cost of data placement while meeting user access latency requirements and maintaining a reasonable load balance between edge servers. Further improvement is discussed and the idea of regional value is proposed.