Research On The Key Technologies Of Self-Optimization In The LTE Network

With the continuous development of wireless communication technology and the growing demand for user service, the number of network base stations increase rapidly,2G/3G standards coexist, and the LTE make the network become more complex. At this time, if the conventional optimizing method is still used, it will increase labor costs, and can not optimize network performance quickly. The technique of Self-optimization is able to enable the network to automatically sense the changes in the environment, and automatically adjust the control parameters of the system, and attracts the special attention of researchers, in order to improve network performance.Based on the analysis of the characteristics of LTE self-optimizing network and self-optimizing control theory, this paper studied the key self-optimizing technlogies of LTE network in-depth, proposed the novel optimizing models and improved self-optimizing algorithms, and verified their correctness through theoretical analysis and computer simulation. This article mainly focuses on handover parameters self-optimization, mobility load balancing self-optimization and coordination between them. The main innovative contributions are listed as follows.1. A novel handover state model based on call dropping (HOSMCD) is proposed. This model uses the Hamilton optimal funtion to figure out the optimal control conditions:when the situation of the call dropping rates of each cell varying with the cell load variation is the same, the call dropping number of the entire networ can be reduced to minimum, at the same time, the network will reach to a stable state. Then, the impact of handover parameters on the handover performance is studied based on the theoretical analysis and simulation, which derives the adjustment rules of handover parameter varying with the change of handover performance and reducing the system complexity by cutting down the control parameter number in need.2. With the analysis of the relationship between the CQI information and user's speed, cell load, SINR, a handover parameter improved self-optimizing algorithm based on the CQI information is proposed which simulation result shows that the algorithm can ruduce the call dropping rate effectively. Finally, with the analysis of rapid handover issue, this paper proposed a novel self-optimizing mechanism based on the event and periodic report, which can reduce the signal load.3. According to the characters of spatial distribution and time distribution of network load, the kalman filter load pridicting model is establised, which can draw the judge condition. Base on the conclusion above, a novel ant colony self-optimizing algorithm of load balancing is proposed, which is based on stimulation intensity of all users in the cell including cell load, the received signal quality and user location. This algorithm can make sure the judge range of handover parameters and direct some appropriate users to move to neighboring cells in order to achieve load balancing. The simulation results show that the proposed algorithm can reduce the number of unsatisfied users significantly compared with iterative algorithm and drop the HO failure rate.4. With the analysis of network's coverage types in the Inter-RAT scenario, a new load balancing self-optimizing algorithm based on coverage type is proposed. The simulation results tell that compared to traditonal method, this algorithm converge faster and can effectively implement network load balancing.5. Based on the analysis of the PingPang effect of load transfer in mobility load balancing, this paper proposed an improved self-optimizing mechanism, and designed a new message with life cycle format which can reduce the signaling burden.6. By analyzing the relationship between handover parameter self-optimization (HOPSO) and mobility load balancing self-optimization (MLBSO) using handover trigger event, the conclusion is as follows: although these two achieve their objectives by adjusting HO parameters, but their actions are in opposite direction. Based on the above-mentioned, the handover parameter self-optimization and Load balancing self-optimization cooperation model (HLCM) is established. The objectives of this model are reducing the entire network user blocking rate and the entire network radio link failure rate. With the help of Lagrange multiplier method, the optimal control condition is detailed: when the situation of the call dropping rates of two neighboring cell varying with the cell load variation is the same, HLCM model gets its goals. Finally, the simulation results of HLCM algorithm are shown as following:(1) the particle swarm algorithm can effectively locate the optimal solution in the Pareto front surface; (2) The analysis of time complexity of the HLCM algorithm shows that the convergence rate fast through game theory; (3) HLCM algorithm can achieve coordination between HOPSO and MLB SO, while optimizing the user blocking rate and failure rate of radio link.