Research On Traffic Splitting Method In Heterogeneous Integrated Networks

Currently, with the rapid development of various wireless communication networks, such as2G/2.5G/3G, LTE,4G, WLAN, WPAN and WSN, the integration of heterogeneous wireless networks has become an important trend in the development and evolution of future wireless communication networks. The development of multi-mode terminal technology make it possible to access to multiple heterogeneous networks simultaneously and transfer data parallelly. Concurrent multipath transfer technology among heterogeneous wireless networks can make full use of network resources to meet the future needs of communication services, but there are also many problems such as large reordering delay, serious packet disorder, unreasonable diversion scheduling easily leading to throughput suppression and different structure of heterogeneous network easily leading to inequitable distribution of resources. In addition, the joint optimization of transfer efficiency and tariff are also the common concern of users and networks operators. The research on concurrent multipath transfer technology in heterogeneous wireless networks requires to be expanded and further study. This thesis studied the concurrent multipath transfer technology in heterogeneous wireless networks aiming at reducing the reordering delay and joint optimization of QoS and network tariffFirstly, this thesis mainly considers concurrent multipath traffic splitting method in heterogeneous wireless networks to reducing the reordering delay. Based on the network characteristics of heterogeneous wireless networks and traditional traffic splitting algorithms, an optimization model for traffic splitting is formulated based on the optimization goal of minimizing the maximum transmission delay difference between different connections, making the transmission delay of packets through different connections close to reduce the reordering delay. GASA is introduced into the problem solving of concurrent multipath traffic splitting, thus the optimum traffic splitting ratio among the connections is effectively calculated. Simulation results by Matlab simulation software show that, compared with the existing typical mixed methods for traffic splitting, the proposed method can effectively improve the system performance in terms of average packet error rate and throughput while maintaining a lower reordering delay.In addition, because of the contradictory relationship between service performance and network tariff, this thesis studies traffic splitting method on both service performance and network tariff for the first time and the joint optimization of traffic splitting algorithm is proposed for the first time. This method introduced and analyzed the network operator’s billing models, deriving the calculation method of different network tariff and reordering delay and setting it to dual-objective optimization problem. In order to solve the multi-objective problem, the Pareto genetic annealing algorithm (NSGASA) is introduced to get optimum traffic splitting ratio effectively. This thesis analyzed the feasible solutions of traffic splitting algorithm based on the joint optimization of service performance and network tariff under different weight. Simulation results show that, with the increasing of the wightss, the proposed method can effectively improve the tariff performance at the cost of service performance。The curve of tariff and service performance has a good reference value to future integration billing system of heterogeneous wireless networks.In the end, this thesis concludes the full text and the work of next steps is discussed.