Abstract
In recent years, Network Mobility (NEMO), technology, which supports a mobile network moving throughout the wireless network as a Whole unit, has been capturing more and more attention. Although many approaches based on basic network mobility support have been proposed, rare attempts have been made on the topic of Quality of Service (QoS) provisioning for NEMO. This thesis addresses the main challenge of providing QoS for NEMO, for both the short and long-timescale variants, in terms of the unstable and time-varying available resources of the mobile network. The short-timescale and the long-timescale problems are caused by the bandwidth fluctuation of the wireless link and the frequent handoffs of mobile networks, respectively. Aiming to solve the short-timescale QoS problem for NEMO, two novel scheduling algorithms, the Adaptive Rotating Priority Queue (ARPQ) and Capacity-Adaptive Weighted Fair Queuing (CA-WEQ), are proposed with different approaches. The ARPQ tries to give the existing priority queuing algorithms more fairness consideration, whilst, CA-WFQ aims to improve die classic fair queuing algorithms with priority differentiation. Despite of different working principles, both ARPQ and CA-WFQ provide bounded delays for tile traffic classes with higher priorities and maintains reasonable throughputs for the classes with lower priorities. The simulation results prove; that both of the algorithms prevent the problems of the existing scheduling algorithms and show better overall performance in terms of throughput and delay bound. In addition to CA-WFQ, a Timestamp Updating Scheme is specified to prevent the transmission disorder problem caused by the instant weight regulation in WFQ. By updating the time stamps of the packets accordingly, the scheme keeps the correct transmission order in WFQ under the weight-changing circumstance. This achievement can be further applied to all the weight-adaptive fair queuing algorithms. A signalling protocol suite, NEMO QoS Protocols (NEMO-QP), is proposed to resolve the long-timescale QoS problem of NEMO. It consists of three schemes, including the Dynamic SLS Negotiation Scheme that negotiates SLS dynamically On both intra-network and inter-network levels; the QoS Monitoring Scheme, that informs of possible SLS violations caused by the movement of die mobile network and the SLs Aggregation Scheme, which solves the scalability problem of the mobile network. Combining the three schemes together, NEMO-QP can effectively avoid the long-timescale QoS problem for NEMO by reserving the varying network resources with dynamic SLSs.