| Mobile Ad Hoc networks are provided with many merits, such as flexible networking, rapid expandability and distribution control. It is a new centerless, self-constructing, self-organizing and self-managing network, and is applied in military, individual communications, emergencies and other circumstances where networks can not be easily established but require rapid networking. In recently years, wireless network grows rapidly. At the same time, as its application foreground in individual communication is getting wider and wider, mobile Ad Hoc network is becoming one of the hotspots of next generation communication network. The popularization of multimedia operation and evolution of business application leads to high demand of QoS assurance. The characteristics of multi-hops communications, unreliable radio medium, limited bandwidth, distributed control and node mobility presents a challenge to provide QoS in mobile Ad Hoc networks.To cater for different conditions of QoS application, this thesis researches on the key technologies for QoS in mobile Ad Hoc networks according to the network characteristics. The followings are specific contributions of this thesis:(1) This thesis proposes adaptive cross-layer QoS model. Based on the traditional TCP/IP protocol stack, the cross-layer information exchange module and the adaptive decision-making module are added. The cross-layer information exchange module breaks the restriction of the original layer structure, which realizes information sharing among the layers. The adaptive decision-making module, grounded on the information exchanging and sharing among layers, excavates the correspondence among protocol stack and endows the network with observing, learning, and self-improving ability. From a comprehensive view point, the proposed model optimizes according to services QoS requirement and the conditions provided by the network. The model regulates parameters on different layers of the protocol stack according to network conditions, and adapts to the changes of the network automatically by selecting routing protocol, changing transmitting power, data transfer rate, packet length, coding and modulating technology. And it can provide different QoS by differentiating packet priority, realizing effective distribution of network resources and improving the integrated performance of the system.(2) This thesis proposes adaptive QoS routing strategy. The strategy correctly reflects the changes of the network in good time, which provides the basis to make adaptive choices between strategies of proactive or reactive mode, single path route or multi path routes, single QoS index or multiple QoS indexes joint optimization, load balance or not, bandwidth reservation or not, etc. And thus the adaptive, self-constructed, self-managed highly efficient distributed QoS routing mechanism can be realized. The specific design realizes the following routing algorithm:QoS-aware Multiple Objective Optimization Routing protocol (QMOR), which can realize access control by node bandwidth, select the route that meets the QoS bandwidth demand by multiple objective optimization algorithm that includes path delay, the length of existing packets in buffer, retry number; QoS routing protocol based on immunity algorithm, which chooses the resource consumption function as objective function; the appetency is expressed by the reciprocal of the resource consumption function; the restriction conditions are bandwidth and delay; and finding the optimal solution using the immune algorithm by synthetically considering both of hopping number and delay based on bandwidth guaranteed; QoS-Aware multipath protocol (QAMR), which considers the node congestion, packet collision and other influences of local information for the QoS routing, reflects node and path performance by introducing node utilization factor and path utilization factor, selects multipath that meet the QoS requirement, realizing load balance in the procedure of communicating and improving routing fault-tolerant capability; QoS routing protocol based on fuzzy algorithm, which takes the present required bandwidth as access control restriction, sets Mathematical model based on the parameters of delay, hops, Path Usable Degree(PUD), and throughput ratio and solve the model by fuzzy algorithm, analyzes and makes decisions on the routes between the source and destination nodes; design bandwidth reservation mechanism for special sevices of high QoS requirement. Simulation shows that the adaptive router strategy can accommodate network conditions and services requirement and realize good QoS performance.(3) This thesis proposes the adaptive QoS MAC protocol based on IEEE802.11protocol. The proposed protocol assign different priority classes for different traffic according to special characteristics and performance types of the different networks. It introduces the concept of transmission license, and only the node which holds transmission license can participate in the channel contention, changing the number of licenses according to the load of the network adaptively, controlling the number of the nodes that participate in the channel contention, and ensuring the nodes with licenses share the channels through contention. It sets different contention parameters for the different priorities services, and guaranteeing these services performances to have advantages in the channel contention. It provides special performances with absolutely end-to-end QoS guarantee, satisfying simultaneously the high efficiency, pertinence, spatial-reuse, etc. to the largest extent. Simulation shows that compared to IEEE802.11protocol, adaptive QoS MAC protocol meets the QoS requirement with low and high priorities in the networks, satisfies the high efficiency, pertinence, spatial-reuse, etc. to the largest extent at the same time in limited channels.The above-mentioned works carry out some elementary researches on QoS model, routing and medium access control technology, which aim to meet the QoS requirement of mobile Ad Hoc networks. The results obtained in this thesis have great significance on improving QoS performance in the networks. |