Task Offloading Strategy using Double Q-Learning based Optimization in MEC

Deep learning assisted mobile edge computing (MEC) has recently gained a lot of attention because of its superior ability to decrease the amount of energy and latency of MEC offloading. Energy usage during MEC offloading should be kept to a minimum, a multi-objective optimization problem is formulated, and a low-complexity allocating resources approach is implied. The paper proposes a deep double q-learning (DDQN) based optimization for MEC energy consumption reduction. The result analysis shows that energy consumption with respect to variable tasks and variable users. According to result analysis, it was observed that with increasing number of task and number of users, the energy consumption increases. With DDQN optimization energy consumption is reduced as compared to without DDQN optimization.