Relaxed Greedy-Based Approach for Enhancing of Resource Allocation for Future Cellular Network

Abstract

The study considers the resource allocation (R_Al) for Orthogonal Frequency Division Multiple Access (OFDMA) future cellular network (i.e., Cloud-RAN), where multiple mobile operators can distribute the Cloud-RAN infrastructure as well as network resources possessed by infrastructure providers. We have designed the resource allocation system by solving the dual-coupled problems at two distinct levels (i.e., Upper Level and Lower Level). The first level problem responsible for slicing the front haul capacity (F_cap) and computation of cloud resources for all operators (O_p’s). This would indeed tend to increase the overall profits for each O_p as well as infrastructure provider by accounting the numerical constraints on F_cap and computational resources. The study introduces a dual-level algorithmic approach to solve this two level R_Al problem. At first-level, system considers both U_level and L_level problems by relaxing discrete values with continuous ones. While in the second-level, we introduce two rounding methods to solve the optimal relaxed problems and attain a practical solution for the proposed problem. Finally, simulation results show that the designed algorithms efficiently perform the greedy approach to resource allocation and attain the discrete value very near to the total rate of upper bound acquired by solving resource allocation relaxed problems.