Enhancement of energy efficiency in 5g heterogeneous cran using hybrid scheduling and adaptive activation kronecker neural network based joint power allocation algorithms

Abstract

In fifth-generation (5G) networks, deploying heterogeneous networks newline(HetNets) with macro-cells layered over tiny cells is a practical solution to newlinehandle the growing demand for mobile traffic but increases energy newlineconsumption. An energy-efficient (EE) 5G heterogeneous cloud radio access newlinenetwork using a hybrid online green algorithm-based sleep scheduling and newlinecost-efficient deadline-aware scheduling algorithm (OGASCDASA) is newlineproposed to optimize energy efficiency in the downlink by reducing micro newlineand pico cell energy use while maintaining QoS. Hybrid OGASCDASA is newlineproposed to minimize remote radio side energy use when maintaining newlinecoverage and quality of service (QoS) of H-CRAN. On the cloud side, newlinebaseband units (BBUs) use hybrid simulated annealing with the Gaussian newlinemutation and distortion equalization algorithm and Battle Royale optimization newlineto reduce the energy consumption of BBUs by decreasing the number of BBU newlineservers. newlineThis EE-HCRAN-Hybrid OGASCDASA-SAGMDEBROA method newlineattains 20.48%, 27.34%, and 32.24% higher throughput and 28.30%, 17.30%, newlineand 32.94% lower delay compared to the existing models such as newlineheterogeneous computational resource allocation for NOMA (HCRA-NOMA newlineGMECS), energy-efficient hierarchical resource sharing in uplink-downlink newlinedecoupled NOMA heterogeneous networks (EE-HRA-NOMA-HetNet), and newlineenergy-aware hierarchical resource management with backhaul traffic newlineoptimization in heterogeneous cellular networks (EA-HRM-BTO-HCN), newlinerespectively. newline