Energy harvesting in 5G networks

Abstract

With exponentially increasing number of users and their demand regarding data rate, current 4G cellular networks need to be evolved to next Fifth Generation (5G) networks. Heterogeneous Cloud Radio Access Network (H-CRAN) is the capable architecture for future high data rate enabled, energy efficient networks. H-CRAN differs from today’s cellular system by addition of extra number of Remote Radio Heads (RRHs) within the vicinity of one Macro Base Station (MBS). This provides high data rates to users with minimized interference by centrally controlling the resource allocation. On the other hand, increased density of hardware in the area, H-CRAN also consumes more grid power of the system. To mitigate the greater power requirements for this type of dense network, Energy Harvesting (EH) techniques are used to minimize the grid energy consumption. In EH, energy is harvested from natural sources like solar, wind etc. By maximizing the harvested energy usage instead of grid power, the system’s Energy Efficiency (EE) can be improved significantly. In this thesis, EE of an H-CRAN consisting of several Green RRHs (G-RRHs), powered by EH modules are explored. A Mixed Integer Non-Linear Programming (MINLP) problem is formulated which maximizes the EE of the system. Mesh Adaptive Direct Search (MADS) algorithm is used to optimize the problem. As a result of this optimization, efficient power and resource allocation is done and higher EE is achieved with low complexity and lower consumption of grid power.