On Energy Saving Technologies Of Wireless Network For 5G Systems

With the rapid development of mobile communication technology and the exponential increase of mobile data traffic, the research on new type of network has drawn great attention in industry, such as hyper dense small cell network (SCN) and heterogeneous networks (Hetnets). SCN is characterised by deploying low power nodes (LPNs) to reduce the distance between user equipment (UE) and wireless network, which greatly improves the network capacity and user experience. Hetnets consisting of the co-deployed macrocells and small cells can meet the various application demand. These new network architectures have already been the key technologies of establishing 5G mobile communication systems. However, while the increase of small cell deployment density improves the network capacity, it simultaneously brings increasingly severe network energy consumption problem. Therefore, network energy saving techniques that can guarantee considerable network quality of service have been paid close attention. This thesis focuses on developing energy saving techniques regarding the 5G-oriented SCN and Hetnets.To begin with, we survey the existing energy saving technologies of wireless network thoroughly. On the network architecture and deployment characteristics of the 5G-oriented wireless network, we first intro-duce SCN and Hetnets. Then, we emphatically study the existing long term base station (BS) energy saving techniques with respect to homogeneous networks (Honets) and Hetnets, including BS sleeping and cell range expansion. To save energy in Honets, the low loaded cells or sectors can be switched off, while the remaining active cells or sectors compensate the emerging coverage holes through coverage range expansion. On this basis, we study the energy saving techniques of Hetnets. Generally, the low loaded LPNs can be dynamically switched into sleeping mode, while the macrocells are responsible for network coverage. Finally, we briefly introduce the short term BS energy saving techniques of wireless network, which can facilitate the optimal configuration of network resource on smaller time scale.Targeting on saving energy in hyper dense SCN, we study dynamic micro BS sector sleeping algorithms under variational network traffic. Considering the homogeneous SCN scenario where micro BSs with three sectors are densely deployed, we propose a two-stage dynamic sector sleeping algorithm to periodically con-figure the working mode of all sectors. The first-stage algorithm tests active sectors in a reasonable order according to a sector utility function, aiming at switching off the redundant sectors as more as possible. We propose two schemes to realize this algorithm. The heuristic scheme tests all sectors in network, which incurs considerable complexity. The progressive scheme is proposed to overcome the shortcoming of the heuris-tic one by avoiding the unnecessary testing process. The second-stage sector mode readjusting algorithm lets some sectors which are not suitable for sleeping keep active, and adjust the antenna electrical downtilt of some active sectors, ensuring network coverage. Sequently, for the purpose that the transferred UEs reselect proper serving sectors and frequency resource, we propose an UE transferring algorithm. Simulations demonstrate that these algorithms allow the number of active sectors tracking the network traffic fluctuation effectively, and the two first-stage schemes lead to similar energy saving effect. However, the progressive scheme can significantly reduce the execution time of the sleeping algorithm.On the basis of the research on sector sleeping algorithm in the previous chapter, we continue to study the sleeping algorithm of pico BS in Hetnets. Starting with the Hetnets model, we elaborate the process of UE associating with macro/pico BSs and the corresponding radio resource allocation scheme. Then, with an overall consideration about the characteristics of Hetnets, we design a pico BS utility function to assist the sleeping algorithm in testing active pico BS on a reasonable order. Then we propose two dynamic pico BS sleeping algorithms (i.e., heuristic and progressive) to save energy in Hetnets by periodically configure the working mode of all pico BSs. The heuristic algorithm relies on the utility function to test all pico BSs one by one, which is not suitable for dense deployment of pico BSs in Hetnets due to its high complexity. In order to accelerate the execution of the heuristic algorithm, we propose the progressive pico BS sleeping algorithm. In the purpose of switching off pico BSs as far as possible, we design an UE transferring algorithm for UEs of the tested pico BS, which first transfers these UEs to the nearby macro BSs. When this transferring process fails, the algorithm will try to transfer these UEs to the nearby pico BSs. Simulations demonstrate that the execution time of the progressive algorithm is averagely one third of that of the heuristic algorithm. Besides, the progressive algorithm can switch off more PBSs than the heuristic algorithm while slightly affects the blocking probability, which indicates that the progressive algorithm has a better potential to save energy.