Three-way decisions based service migration strategy in mobile edge computing

In mobile edge computing (MEC), the mobile device cannot always stably connect with the same edge server due to the movement of the user. Considering the complexity and randomness of users’ movement, most of the existing service migration strategies based on two-way decisions (namely either migrate or not migrate) may make wrong decisions, which will increase the energy consumption and significantly impact users’ experience. To address such an issue, we employ three-way decisions for service migration. Specifically, first, based on users’ movement trajectory, we allocate users into three regions: migration region, non-migration region and delay region. Afterwards, we execute different operations for users in different regions accordingly. In the migration region, considering the energy consumption and service delay time, we propose a new service migration method which utilizes the latest migration radius and the lowest energy consumption for service migration. In the delay region, the edge server will keep collecting users’ movement information to prepare for the further decision making. In the non-migration region, no action is required by the edge server. Comprehensive simulation experiments using parameter settings consistent with real-world edge computing environments are conducted to show its superior performance compared with other strategies.