| It’s a basic requiremen to take full advantage of the network resources in heterogeneous mobile network to provide QoS transmission services. Concurrent multipath transfer (CMT) method can match network resource with the transmission requirement and also can improve the ability of the network communication more effectively by using on-demand combination and integration of network resources based on network virtualization. It also provides a new method to ensure the QoS. Aiming to the problem of inefficient transmission of network traffic in heterogeneous maneuver network, this paper researches on the service oriented network architecture and the key technologies of Concurrent Multipath Transfer.(1) Aiming to the difficulty of dispatching the network communication resources of maneuver communication systems, this dissertation analyses the architecture of maneuver communication networks, proposes the service-oriented network architecture, and then designs the architecture model based on the fusion of the heterogeneous network resources in maneuver communication. In this dissertation, it puts emphasis upon the following aspects of the proposed architecture, the function entities and interoperability interface, the operative mechanism and applied allocation modes, that is capable to effectively to support the unified scheduling and on-demand assemblage of the heterogeneous communication resources, and also can provide the functions of resources dynamic and flexible restructuring and communication service ability encapsulation. Based on the network architecture, this dissertation proposes the model of CMT services, which is made up of the universal "thin client" architecture of CMT in network sides and the function structure and the work-flow of CMT services.(2) To improve the throughput performance of CMT, this dissertation studies the influence of path metric disparity on the throughput performance of concurrent multipath transferring system based on Ex-padhye throughput analysis model at first. Secondly, with qualitative analysis, this dissertation simulates and validates the influences on the system throughput which caused by the path differences under equal and unequal path parameters, and then presents some basic conclusions on how to determine the path measurements, which can provide some help for the selection of active multipath in theory. Finally, it presents an active multipath selection algorithm based on the above conclusions and makes a comparison with those existing algorithms whitch shows that the proposed algorithm can improve the throughput much more.(3) As the disparity of multipath, it is facing a great challenge to solve the disorder of the received data packets. To lighten the reordering degree and then to improve the throughput performance, we need to estimate the real-time paths’bandwidth and Round Trip Time (RTT) as exactly as possible. In this dissertation, we use the extended vector Kalman filter to predict the available bandwidth and RTT of each path simultaneously. Based on this, we propose a predictive path selection algorithm for CMT in SCTP which also takes the outstanding chunks of the transmitter into account. The simulation results show that the path selection algorithm can lessen the data packets disordering by correctly predicting each path’s bandwidth and RTT in real time. To bandwidth sensitive scene, the algorithm can converge more quickly than Kalman-CMT and can improve the system total throughput in a certain extent. To time and bandwidth sensitive scene, the algorithm can greatly improve the convergence speed and total throughput than Kalman-CMT.(4) In order to solve the problem of the inefficiency of multicast transmission in maneuver communication network, this dissertation proposes CMT oriented group multicast-tree generation (GMG) algorithm. First, it sets up the mathematic model for the unicast and multicast routing under multiple network constrints, and establishes the multiple constraints and cost function related to the QoS evaluation metrics, and based on the above works, it proposes a glowworm swarm optimization (GSO) based multi-constraint multicast tree generation algorithm (GSO-MCM). The GSO-MCM algorithm can be fit for the properties of low computation power, high topology dynamics, and limited channel resources, and improve the emerging speed of the multicast tree generation and decrease the computing complexity. Then, it proposes the GMG algorithm based on the GSO-MCM. It refers to the CMT and transfers the group multicast problem to the P2P CMT problem. Considering the transmission performance of each multicast tree, in order to maximize the throughput of the multicast-tree group, it reaches the goal of enhancement of the multicast-tree group performance. Finally, it demonstrates the efficiency of the proposed GMG algorithm and GSO-MCM algorithm through comparative experiments.
|
PDF Full Text Request