Research On Load Balancing In Relay Networks

The relay network effectively mitigates the problem that traditional cellular networks can no longer meet the people's requirement of higher data transmission rate. Relay nodes in the networks divide a longer transmission path into some shorter paths with better transmission quality and it can increase the system capacity. Placing a relay node in the location where a base station can't be set conveniently can extend the coverage area of the system. Although relay networks have higher system capacity and larger coverage area, due to the uncertainty of people's movement, there are still some hot cells being overloaded and many new calls are blocked. But in the surrounding cells, there may be much unused resource because of lacking users. The problem of load imbalance is still serious in relay networks. How to utilize load balancing strategies to solve this problem has been a hot point of research. We also research the load balancing strategy in relay networks.We firstly study load balancing in relay networks with multiple traffic types in this paper. We give a model of load balancing, and formulate it as an optimization problem with the goal of maximizing the percentage of successfully accessed users. We give a five-step load balancing strategy to solve it. Because the data rate of non-real-time traffic is variable, we also dynamically control the relationships of users, relays and base stations to transfer traffic from hot cells to cooler ones and decrease the allocated sub-channels of users with non-real-time traffic to allow more users to access the network. It reduces the blocking probability of users. In order to ensure the performance of real-time users, we give priority to them..We also study load balancing that considers the performance of users moving from cell to cell. In dealing with users'call requests, we fully consider the difference of new calls and handover calls, and give a load balancing strategy that handover calls have higher priority. In this strategy, we set different access threshold for new calls and handover calls, and the new calls have higher threshold. It means that we set some guard channels for handover calls, and it ensures handover calls have more available channels. At the same time, the direct and indirect load balancing approach is adopted to transfer more users from a hot cell to the surrounding cooler ones. It utilizes the system resource effectively and reduces the blocking probability of users'calls. Then we analyze the performance of the proposed strategy using the three-dimensional Markov chain model. Theoretical analysis and Simulation results show that this strategy well ensures the higher priority of handover calls, and the handover calls have a lower blocking probability.