Self-Optimization Method Researchs In LTE Networks

In order to reduce the human costs in parameter configuration and management of next-generation wireless network, and further improve the performance of network operation and maintenance (OAM), wireless networks must become adept at self-organization (SON)—allowing devices to reconnoiter their surroundings, cooperate to form topologies, and monitor and adapt to environmental changes, all without human intervention. The three commonly distinguished domains of self-organisation are self-configuration, self-optimisation, and self-healing, which are applied in nerwork deployment, network operation and network anomaly, respectively. Among this three components, self-optimisation is the core which can be applied to re-optimize the configuration parameters in self-configuration or to heal the fault region detected by self-healing. Hence, in this dissertation, the self-optimisation component is mainly explored.In fact, the need for the introduction of SON has already been articulated by the Next Generation Mobile Networks (NGMN) alliance, and3rd Generation Partnership Project (3GPP) has also published the related standard where some use cases have been defined. Based on the self-organization frame provided by3GPP, three self-optimization use cases are investigated in long term evolution network (LTE) in this dissertation, self-optimizing load balance, self-optimizing Inter-cell interference coordination and self-optimizing energy saving, which are used in different network load status, respectively:1. Self-optimizing load balance (LB):LB schemes are always adopted to transfer some users in high-loaded cell to light-loaded cell, and then improve the efficiency of radio resource use. The dissertation provides the LB solutions in two ways. Firstly, A high-efficient LB algorithm based on handover is proposed. By leveraging the handover trigger condition in LTE network, the algorithm adaptively adjusts the handover trigger condition upon the load status of cells and differ the probability of users handovering from different cells, then makes the load balanced over cells finally. Furthermore, the relation between cell load and users association with base stations (BSs) is studied, and a distributed admission and handover scheme (DAHS) is proposed. Mobile users with rate requirement access or handover to a base station based on the load utility in a distributed way. To prove the superiority of our algorithms, the dissertation also evaluate the algorithms, and numerical results demonstrate that our proposed schemes outperform the conventional counterpart.2. Self-optimizing Inter-cell interference coordination (ICIC):Soft fractional frequency reuse (SFR) is adopted as a main inter-cell interference coordination technique in the orthogonal frequency division multiplexing access(OFDMA) networks. Therefore, this dissertation focuses on the design of self-optimizing SFR frame. Unlike the fixed power ratio configuration published in current papers, an adaptive power ratio adjustment scheme is explored to maximize the sum-rate of each cell individually under the varied status of user and interference. Besides, the diversity gain of adaptively allocating the cell-edge bandwidth to each cell is also investigated and it’s found that such allocation is meaningless in the large network size or long allocation period. Consequently, allocating the resource blocks to users is discussed and an allocation solution is proposed to either minimize the cell load when all users can be served or maximize the number of served users. Such a solution reduces the processing complexity of BSs and improves the control flexible of quality of service (QoS).3. Self-optimizing energy saving:It’s estimated80%power is wasted by the inefficiency of power amplifier (PA) and the static circuit consumption. Hence, this dissertation discusses the energy saving method of turning part of BSs into sleep mode and transferring the associated users to neighboring cells, which is termed as BSs switching on/off approach. The problem "how to select the smallest set of active BSs that can preserve the quality of service (the minimum data rate) required by the users?" is studied and NP-hard characteristic of problem is proved. Therefore, two low complexity approximation algorithms that yield performance guarantee are proposed in the dissertation. Besides, the implementation and computational complexity of proposed algorithms are also discussed. The research can be a prospective inspiration for operators to desigh energy saving solutions.