Efficient Resource Allocation in 5G Communication Networks through Mobile Edge Computing: A Deep Learning Approach

– Techniques for computing offloading are challenging to design for the considerations such as the mobile edge networks, or base stations, which are deployed at random, limited capacities of servers, and the constant mobility of end users. In light of the capacity of DRL in dealing with intricate, dynamic challenges, this investigation formulates a procedure for determine the optimal method for compute offloading and resource management. First, the authors provide a comprehensive description of a specific case with a multi-user mobile edge network, which is MBS, SBS, and multiple terminal nodes. They next describe and elaborate on the Communication and Computation Overhead that is generated. In addition, the optimization goal, defined in this paper, involves taking the overall system energy consumption into full compte when harmonised with predictable task delays. The subsequent suggestion on decreasing the power consumption level will involve proposing a learning method, namely Deep Deterministic Policy Gradient (DDPG). Finally, simulation research proves that the introduced DDPG approach by the authors enhances the results of other correlation methods in the aspect of target value identification and the overall energy usage of the system, which is only 15. 6 J. Furthermore, it is possible to observe that the proposed algorithm predicts higher allowances of communication resources as compared to the existing method.