Energy-efficient Access Control In Cellular Heterogeneous Networks

In LTE-Advanced wireless communication systems based on OFDM and MIMO technologies, the spectral efficiency of systems need to be improved owing to higher transmittability requirements while the large amount of energy consumption of ICT attract widespread concerns. The next generation of mobile communication networks should optimize the network resource and reduce the energy consumption per traffic volume to achieve green communication. Therefore, how to reduce the energy consumption and improve the energy efficiency of communication systems has become a hot issue.Deploying low power nodes such as Relay 、 Pico and Femto overlapped in traditional cellular network which is called heterogeneous network can increase the coverage of network and improve the system capacity. Compared with Macro BS which is cost-effective and consumes large power, small nodes consume less power thus reduce the energy consumption of system. The small nodes in heterogeneous networks deployed in hotspots or blind spots which are close to users, leading to lower path loss and better users performance. Therefore, offloading traffic from Macro cell to small cells in heterogeneous network is an effective solution to improve the system energy efficiency(EE).However, cell splitting and new cells created in heterogeneous networks may result in strong interferences. In heterogeneous networks, the access control of users is a basic challenge. On the one hand, in order to effectively exploit the advantages of small nodes and improve network throughput, the service providers hope that more users can select small nodes as service nodes. However, users far away from center of small nodes suffer from severe interference owing to the low transmit power of small nodes. Thus an effective access control scheme is essential.In this thesis, we investigate the impact of the user access policies on system performance in joint Macro-Pico networks and propose access schemes to optimize system performance. In these schemes, we design optimized algorithms such as dynamic programming and genetic algorithm.Deploying small cell to offload Macro cells’ traffic is called traffic offloading. First, we propose a low-complexity traffic offloading algorithm called TO(Traffic Offloading) algorithm. In TO algorithm, users in Macro cells are offloaded to Pico cells which meet the requirement of RSRP(Reference Signal Received Power) which is the measurement value of users and load requirements. However, these nodes deployed in same geographic location share the same spectrum resource, leading to strong interferences and thus deteriorating the QoS(Quality of Service). Based on it, we combine TO algorithm and FFR(Fractional Frequency Reuse) technique and propose another traffic offloading algorithm called TOFFR(Traffic Offloading based on Fractional Frequency) algorithm. Numerical results demonstrate that both algorithms can effectively improve the energy efficiency of system and TOFFR is better.Next, we address dynamic traffic offloading issue. According to the timeliness of data offloading,there are two types of offloading: on-the-spot offloading and delayed offloading. We consider the user offloading model and model the dynamic traffic offloading as a dynamic programming problem which divides the problem into multiple sub-optimize problems. Then we solve the DP problem with backward induction method. In this scheme, we consider different delay value and exploit users’ rate constraints to reduce the algorithm complexity. Numerical results show that compared with RSRP based traffic offloading algorithm, DP based traffic offloading algorithm can achieve better energy efficiency and offloading efficiency.SE(Spectral Efficiency) is an important performance metric. We next investigate the tradeoff optimization of both SE and EE. We propose an SE-EE joint optimization access scheme. First, we exploit Analytic Hierarchy Process(AHP) to determine the normalized weight coefficient of multiple objectives optimization. Hence, the problem is converted into a single-objective optimization problem. Then, we solve the problem by genetic algorithm(GA). Simulation results show that our proposed joint SE-EE optimization scheme can effectively achieve a desirable tradeoff between SE and EE.