Research on Key Joint Optimization Techniques for Quality of Service Provisioning in WiMAX Mesh Networks |
|
Author | LiYaJun |
Tutor | YangYuHang |
School | Shanghai Jiaotong University |
Course | Communication and Information System |
Keywords | Wireless mesh network QoS joint optimization cross-layer design WiMAX IEEE802.16 MAC convex optimization |
CLC | TN925.93 |
Type | PhD thesis |
Year | 2010 |
Downloads | 83 |
Quotes | 0 |
Wireless Mesh Networks (WMNs) have emerged as a key technology for next-generation wireless networking. Due to its high-rate and wide-range transmission capabili-ties, WiMAX-based mesh networks have attracted considerable attention during last severalyears. Different from conventional Ad-hoc wireless networks, mesh routers in WMNS arerarely mobile and may not have power constraints. Moreover, the service traffic is mainlyrouted by the wireless mesh backbone between the mesh clients and the wired Internet andgoes through the gateway nodes. As such, the two major concerns, namely node mobilityand power consumption in Ad-hoc networks, are not significant in WMN. Instead, Qualityof Service (QoS) provisioning becomes the main topic in mesh networking.Our work focuses on developing joint optimization algorithms for QoS provisioningin WiMAX-based mesh networks. We aim for providing specific joint optimization solu-tions to provide QoS guarantees while taking account of the MAC scheduling mechanismsin the IEEE802.16mesh standard. our solutions consist of the maximize network utilityoptimization under centralized scheduling mode and the aggregated MAC optimization un-der distributed scheduling mode. In particular, there is always an central node (usually thegateway node) which is in charge of the network management and scheduling in the centralscheduling mode, thereby rendering the optimization-based method possible. We thereforeconcentrate on developing joint scheduling optimization using network utility function con-strained by link scheduling. On the other hand, we mainly exploit the distributed schedulingmechanism to form an aggregated MAC layer since the link contention has been solved bythe pseduo-random selection algorithm in the standard. The main contents of this thesisinclude:1, presenting an optimization-based QoS multi-path routing and scheduling algorithm.We view the multi-path routing problem as a utility maximization problem. By employingthe contention matrix to represent the wireless link interference, we formulate our convexoptimization routing problem under the impact of interference in wireless mesh networks and resolve it using the primal-dual method. Since the multi-path routing usually resultsin the non-strict concavity of the primal objective function, we first introduce the ProximalOptimization Algorithm to get around such difficulty. We then propose an algorithm to solvethe routing subproblem and the scheduling subproblem via the dual decomposition. Simula-tions demonstrate that our solution have the capabilities to find least-interference routes andachieve the best revenue. Also, our algorithm outperforms the WSP in terms of the revenueof the whole mesh network.2, presenting a network utility maximization framework for fair rate control andscheduling in wireless mesh networks. Our objective here is to control the data rates of theapplications so that they fall within the capacity region of a wireless mesh network, therebyavoiding the performance degradation by system overloading. We formulate the joint ratecontrol and MAC scheduling as a convex optimization framework with the constraint ofMAC link scheduling using the K-hop interference model. We proof the strong duality ofour framework and also demonstrate that our framework has an attractive decompositionproperty such that the rate control problem and the scheduling problem can be decomposedand solved individually. We then focus on the scheduling subproblem and employ ColumnGeneration algorithm to reduce the computational complexity. Simulation results show thatour proposal is efficient in convergency as well as tractable and is able to reach a desirableclose-to-optimal solution.3, proposing a novel distributed routing algorithm for IEEE802.16/WiMax based meshnetworks. Based on the underlying IEEE802.16standard distributed scheduling mechanism,our routing algorithm is incorporated into the medium access control layer. Each node deter-mines the next-hop nodes for the passing flows according to the scheduling information andattempts to forward packets in the very earliest slots. The simulation results show that ourproposal can considerably reduce the delay of traffic flows and also achieve load balance toa certain degree.4, addressing the problem of heterogeneous Quality of Service (QoS) provisioning inIEEE802.16-based wireless mesh networks. Based on distributed scheduling in the IEEE802.16-based mesh mode, we first devise a novel differentiated service framework, which isimplemented at the medium access control layer. A multiple-class traffic fairness model isalso presented. We then present a QoS-aware fair packet scheduling (QFPS) algorithm to ful-fill the QoS provisioning. QFPS has the capability to allocate proportional bandwidth. Also,with QFPS, traffic flows with different priority level are served using strict priority policywhile fairness are imposed on these flows with the same priority level according to our pro- posed fairness model. The experimental results of the QFPS demonstrate that various trafficflows are provided with a differentiated service that offers quality of service guarantees.Also, our proposal can gain higher network throughput as compared with its counterparts.