Load Balancing Optimization Algorithms In Relay-Assisted Lte-A System

Mobility load balancing optimization is an important part of the self-optimizing function in self-organized network, which is a hot issue among researchers. Handover process and load balancing scenarios become more complicated in relay-assisted LTE-A system. This article mainly focuses on the load balancing algorithms in LTE-A relay system.Firstly, a load balancing based relay selection algorithm is proposed. In order to select an optimal relay for a new user, the channel condition and load state of the relay should be taken into account. The approach is efficient in preventing the relay to be overloaded.Secondly, this paper presents a novel multi-objective optimization modeling strategy which integrates mixed users, i.e., the GBR and the BE users. It is a more general model for the practical load balancing situation, which is proved to be NP-hard. A hybrid multi-objective particle swarm optimization algorithm is proposed to solve the original problem. Then a new centralized load balancing scheme is achieved in our work.However, in order to make up for the lack of instantaneity in the centralized algorithm, this paper further tries to apply analytical methods. In consideration of QoS priorities of different users, a decomposition is proposed to solve the original model. Through appropriate mathematical derivations, such as applying the Lagrange multiplier method, sub-optimal solutions for mixed users are deduced. Based on the derived solutions, including resource allocation schemes, a distributed multi-objective load balancing algorithm jointly dealing with mixed users is proposed.Finally, the simulation results of the proposed relay selection algorithm and the two load balancing algorithms are shown in this paper. The prefered load balancing algorithm can be selected according to different users’ needs, since the centralized one and the distributed one both have advantages and disadvantages in system overall performance and instantaneity.