An Efficient Updating Scheme for Popular Edge Data in User-Centric Internet of Things Systems

As an emerging computing paradigm, the data access delay can be reduced by caching popular content on edge servers in user-centric Internet of Things (IoT) systems. However, most existing research overlooks the spatiotemporal dynamics of user requests in the IoT system, where different users may request the same data at different times. At this point, different edge servers may repeatedly cache duplicate data from the cloud, increasing the load on the backbone network and causing transmission congestion, thereby reducing user experience. Therefore, there is an urgent need for edge data update solutions that meet user requests in the IoT. This article addresses the popular edge data update (PEDU) problem in user-centric IoT systems environments from the service provider’s perspective, aiming to maximize the comprehensive benefit of data update schemes. Specifically, we model the PEDU problem as a constrained optimization problem (COP). For small-scale PEDU scenarios, we propose the PEDU-IP method using integer programming techniques. For large-scale PEDU scenarios, we develop the PEDU-A method to compute approximate solutions and establish its theoretical approximation ratio. Finally, we evaluate and analyze the PEDU-IP and PEDU-A methods against four baseline approaches on real-world datasets. Experimental results demonstrate that our proposed methods can effectively address the PEDU problem with high efficiency.