Research On Load Balance Technology Between TD-SCDMA And GSM Systems

TD-SCDMA is the 3rd generation mobile communication standard proposed by China, and the system can support flexible broadband multimedia services. GSM is the 2nd generation mobile communication system which has the largest user population worldwide. Balancing the loads of the two networks in their common coverage area can improve the utilizations and performances of both systems, and thus provide more reliable services and better experiences to their users.For the purpose of load balance, we focus on the expressions of normalized loads of TD-SCDMA and GSM systems, and the methods of load balance through services migration.The unified measurement of the normalized loads in the heterogeneous networks is studied first. Uplink and downlink normalized loads of multi-cell multi-service multi-carrier TD-SCDMA system is investigated by mathematical analysis. The expression is deduced for the number of co-existing users with various types of services, and the normalized load is obtained for TD-SCDMA system employing smart antenna and multi-user detection. Neighboring cell interference is studied, and its statistics are obtained under the conditions of different scrambling code allocation schemes. Capacity is then expressed with respect to outage probability. The load calculation for GSM system is analyzed and discussed based on Campbell Model. An equivalent service is constructed considering all types of services and the channel number, and the total equivalent traffic supportable is computed. Then the normalized load expression of mixed services is derived.The method of handoff-based load balance is analyzed. Users are migrated to a lower-load network to achieve load balance when certain trigger condition is satisfied. A load gap threshold changing adaptively with the current load level is employed, so that handover is more easily triggered when load is higher and difficult triggered when load is lower. Therefore, the risk of rapid performance deterioration can be avoided when load is high, and the unnecessary balance operation can be avoided when load is low. Considering the characteristics of the two systems comprehensively, a constraint expression is derived for the volume of services to be migrated, so that the load gap after the migration is guaranteed small. Users to be migrated are chosen by means of multi-attribute decision, which makes the user selection reasonable. The effects of main parameters of the algorithm are also discussed. Simulation results show that this algorithm can balance the loads of the two systems effectively, and the performance of the higher-load system can be improved obviously when the performance of the lower-load system is guaranteed.Load balance is discussed in detail when asymmetric services are carried in the two systems. The uplink and downlink loads are different because the rates of uplink and downlink asymmetric services are different. The levels of uplink and downlink loads constitute four load relationships. By analyzing each case, the trigger condition of handover operation, the constraint expression of the volume of migration services and the method adopted for choosing the users to be migrated are obtained for the case.The method of redirection-based load balance is analyzed, then a load balance algorithm based on joint handoff and redirection is proposed. The redirection-based load balance is trigged when a new call arrives, the user can be admitted, rejected or redirected to another system according to the pre-set criteria. On the contrary, handoff-based load balance can be triggered only in the beginning of each measurement cycle, and a batch of the user residing in the system are migrated if needed. Using the proposed joint algorithm, users can be migrated through both handoff and redirection operations. The advantages, disadvantages and applicabilities of these three algorithms are summarized. Simulation results show that the performance of load balance algorithm based on joint handoff and redirection is slightly better than that of the handoff-based load balance algorithm, and the these two algorithms are all significantly outperform the redirection-based load balance algorithm.