Task Execution Latency Minimization-based Joint Computation Offloading and Cell Selection for MEC-Enabled HetNets

Multi-access edge computing (MEC) is a promising platform for cellular heterogeneous networks (HetNets). It brings the computation services of the core cloud into the nearest point of mobile devices (MDs). This paper jointly studies the problem of computation offloading and cell selection strategies in MEC-enabled HetNets under the assumption that MD can select one of the three ways to execute its task, i.e., complete local computing mode, complete MEC computing mode or partial computing mode. The problem of joint computation offloading and cell selection is formulated as a task execution latency minimization problem. This formulated optimization situation is NP-Hard, which is inconvenient to solve. Hence, we deduce the problem into two subproblems which is computation offloading subproblem and cell selection subproblem and propose extensive search method and Khun-Munkres algorithm respectively to obtain the optimal solution. Numerical results show the efficacy of our proposed algorithm.