Rate Allocation In Cooperative Access Over Heterogeneous Wireless Networks

Wireless network in future is a combination of heterogeneous networks with variousaccess methods, which can support different services with various transmission rates.Therefore, cooperative access through heterogeneous networks is an efficient solution tohigh-speed and high-efficient communication in any place at any time. Moreover, multi-modeterminals and their popularization make it possible to access heterogeneous networks andtransmit through them once at a time, where cooperative transmission in high speed cansatisfy the need for multi-media services with high QoS. Therefore, how to realizecooperative access and hence rate allocation, rate control according to network status,network capability and the traffic requirement is one of the key problems of heterogeneousnetwork fusion. Moreover, the win-win between networks and terminal users should beguaranteed in cooperative access with the heterogeneity being taken into account. And therelated research has been attracted a wide spread attention.Holding such background, how to realize rate allocation and rate control in cooperativeaccess is a difficult problem in the management of heterogeneous network resource. Thisthesis will focus on this topic with the mathematic help of game theory, queue theory,optimization theory and so on. From network throughput gain, transmission time and accesscost, a series of rational rate allocation schemes are proposed. The major contributions of thisthesis are as the following:(1) Rate allocation based on network throughput gainIf we consider the networks as rational players, they would like choose collectively anoutcome in the negotiating method to benefit all the individuals, which is a progress ofbargaining. With the heterogeneity of network transmission capability being taken intoaccount, the thesis formulates the rate allocation as a weighted bargaining game. Moreover,with the maximized throughput gain as the optimal objective, the closed form of the NashBargaining Solution is derived with the Lagrange multipliers method. Simulation resultsdemonstrate that the presented framework is more efficient with completely balanced trafficso as to make full use of the heterogeneous network resource and prevent network saturationas much as possible.(2) Rate allocation based on network transmission timeThe static rate allocation over networks is formulated as a weighted bargaining gameframework, where the heterogeneity of the transmission capabilities of different networks istaken into account. With the minimized transmission time as the optimal objective, the utility function is proved to be convex and the closed Nash Bargaining Solution is derived with theLagrange multipliers method. Moreover, an algorithm is presented to realize rate allocation inthe multi-mode device. Simulation results demonstrate that the proposed method is moreefficient with the least transmission time and suitable for the emergency communicationscenario.(3) Rate allocation based on a pricing function in wireless random access networksCollision probability and access delay will increase with the competitive host number inwireless random access networks. With the pricing function based on channel utilization todescribe access delay cost, a method of rate allocation is presented to achieve the maximumtotal network payoff at a minimum access cost. Simulation results show the proposed methodcan serve as an access selection method and distribute the traffic to networks according toboth network status and channel efficiency. Performance analysis indicates it makes atrade-off between access delay and fairness. Access to heavier loaded random access networkscan be avoided so as to increase access efficiency.(4) Rate control based on potential gameTransmission capability and thus transmission delay varies quickly in wireless channel.To satisfy the quality of service, transmission rate should be self-adaptively adjusted. Thethesis formulates the rate control in a noncooperative game framework and proposes apotential game based rate control alogrithom, which can obtain the optimal transmission ratewith the uniqueness of the Nash equilibrium. Experiment results show such a method canimprove the network utilization in different wireless access networks.(5) Cooperative access with intelligent APThe thesis proposes a cooperative access scenario based on intelligent AP. With theinformation of network status being sent to the mobile terminal through a ZigBee link, adistributed rate allocation and rate control can be realized in the mobile terminal. Hereafter,the software and hardware infraustructure of intelligent AP has been introduced. TheStandardized design of network interfaces in cooperative access and the flow chart of rateallocation have also been presented in the end.In summary, this thesis conducts a research deeply into cooperative access and rateallocation over heterogeneous wireless networks. The results of the research work willprovide partial theoretical evidences and technical bases for the development of futurewireless network fusion.