Traffic-Driven Energy Efficient Operational Mechanisms in Cellular Access Networks

Abstract

Recent explosive growth in mobile data traffic is increasing energy consumption in cellular networks at an incredible rate. Moreover, as a direct result of the conventional static network provisioning approach, a significant amount of electrical energy is being wasted in the existing networks. Therefore, in recent time, the issue of designing energy efficient cellular networks has drawn significant attention, which is also the foremost motivation behind this research. The proposed research is particularly focused on the design of self-organizing type traffic-sensitive dynamic network reconfiguring mechanisms for energy efficiency in cellular systems. Under the proposed techniques, radio access networks (RANs) are adaptively reconfigured using less equipment leading to reduced energy utilization. Several energy efficient cellular network frameworks by employing inter-base station (BS) cooperation in RANs are proposed. Under these frameworks, based on the instantaneous traffic demand, BSs are dynamically switched between active and sleep modes by redistributing traffic among them and thus, energy savings is achieved. The focus is then extended to exploiting the availability of multiple cellular networks for extracting energy savings through inter-RAN cooperation. Mathematical models for both of these single-RAN and multi-RAN cooperation mechanisms are also formulated. An alternative energy saving technique using dynamic sectorization (DS) under which some of the sectors in the underutilized BSs are turned into sleep mode is also proposed. Algorithms for both the distributed and the centralized implementations are developed. Finally, a two-dimensional energy efficient network provisioning mechanism is proposed by jointly applying both the DS and the dynamic BS switching. Extensive simulations are carried out, which demonstrate the capability of the proposed mechanisms in substantially enhancing the energy efficiency of cellular networks.