Due to the nature of mobile systems in which devices change their locations continuously, there is a need for efficient techniques to integrate responses from different agents. Such techniques implement cross-layering service migration models to handle integration issues. This paper proposes a method that aims to model the migration process in an integrated architecture to resolve cross-layering issues, develop a dynamic policy, and optimize the cost of migration in terms of power consumption and communication. The proposed technique extends the Markovian model to adopt the spatiotemporal aspects of this problem, allowing modern 5G vertical applications to communicate seamlessly in highly connected IoT networks. Because of its dynamicity, we propose to project the environmental aspects into the observation part of HMM. We performed experiments to test the performance of the proposed method in terms of efficiency, reliability, and complexity. The results showed a significant performance in the mobile edge computing environment.