Research On Resource Management In Heterogeneous Network

With the 5G in-depth study of the international, the requirements for mo-bile business are getting higher and higher. In addition to the speed as a basic requirement, users pay more attention to the wireless system's delay and sta-bility. The emergence of heterogeneous networks provides a possible solution to these requirements. Heterogeneous network is a traditional cellular network with many low-power nodes inserted, aiming at greatly reducing the distance between the users and the base stations, which can improve the performance of the system.However, the introduction to low-power nodes in the heterogeneous net-work brings some new technical challenges to the system. Firstly, the number of base stations is large and the channel state information between the users and the base stations is huge. So the complexity of control is very high and the network management optimization becomes more difficult. Secondly, the low-power nodes introduced will greatly improve the whole energy consumption.Especially when the system is in low load, it will cause a great energy waste.Therefore, new management methods must be proposed to the heterogeneous network, so as to reduce the adverse effects of the newly introduced low-power nodes.Under such circumstances, this paper firstly analyzes the characteristics of heterogeneous network, and discusses the resource management problem. Sec-ondly based on the strategy of the base station sleeping, the relationship between the energy efficiency of small cell networks and the density of low-power nodes is analyzed. Then an algorithm with QoS constraint is proposed to and reduce the system energy consumption. By the idea that firstly reduces the complexity,then gradually excludes the proceeding and from the perspective of comput-ing complexity, system performance and so on, we find this part of low-power nodes which can be transferred to sleep mode base station, thereby reducing the overall system energy consumption. The performance, energy consumption and complexity of the system achieve a good compromise.Simulation results show that the theoretical analysis can catch the simula-tion results, and the proposed two-stage strategy can effectively deal with the resource management problems of heterogeneous networks. At the same time,the computational complexity of the strategy is low and it can effectively reduce the energy consumption of the system.The main innovations and contributions of this paper are as follows:1) Analysis of heterogeneous network characteristics. In this paper, a mathematical model of stochastic geometry is used to model the low-power nodes and users in the heterogeneous network. A simple user accesses control scheme is proposed based on the characteristics of the heterogeneous network,which makes full use of the heterogeneous network. At the same time, the en-ergy consumption model of base station which is close to reality is given.2) Analyze the energy efficiency in the densely deployed small-cell net-work. In this paper, we first derive the average reachable rate of users with stochastic geometric tools, and then deduce the relationship between system en-ergy efficiency and active base station density from the point of view of energy efficiency of the system with the energy model of low-power nodes. According to this relation, the active low-power node density is optimized theoretically to maximize the energy efficiency of the system.3) A two-stage based station sleep strategy with low complexity of hetero-geneous networks. Aiming at solving the complexity of resource management problem of heterogeneous networks, this paper proposes a feasible way to allo-cate resources which are determined by the needs of users to achieve the user's QoS. Then, a low-complexity two-stage base station sleep strategy is proposed.Some base stations can be turned to sleep mode to reduce the system energy consumption.4) The way of using genetic algorithm as contrast scheme. Because the complexity of the exhaustive search algorithm is too high to find the optimal solution to finite time, this paper uses genetic algorithm to obtain a subopti-mal solution. In the genetic algorithm, combined with the characteristics of heterogeneous networks, we establish a solution model, an evolutionary way and a suitable selecting method. Finally this paper compares the results of the proposed algorithm with the genetic algorithm.