Research On Resource Allocation Scheme For Ad Hoc Networks Based On User Utility

With increasing demand for mobile services in different contexts recent years,ad hoc network technology has been rapid developed. For its distributed,self-organization and other characteristics, it can be applied to kinds of flexiblescenarios. However, its communication radius and battery power, bandwidthresources are limited, and the interference problems between links are difficult tohandle. Therefore, the network resource efficiency is not high. In ad hoc networks,how to improve the quality of service is one important issue. The key problem is theestablishment of user-centric service model with limited resources and reasonableallocation of network resources efficient. It has aroused researchers’ attention inrecent years.However, flows in ad hoc network will contend with each other even if they donot share the same link. This makes that mature achievements get in the cablenetwork resource allocation cannot be copied to the ad hoc network. To adapt to thenew features, we should use the research methods which are different from the wirednetwork to derive resource allocation strategies for efficiency of limited wirelessresources and service experience.In this paper, we propose a resource allocation method based on user utility forthe wireless ad hoc network. First, according to the different types of service types,we select user-centric utility functions, and combined with them to construct a newnode energy utility function. Then analyzing network resource model, design thebandwidth allocation scheme, the purpose is to enable network users overall utilitymaximization. The algorithm uses a linear method to approximate utility function,and introduced the concept of the shadow price for resource pricing. Each flowdetermines its "cost" according to the utility functions. Data flow bid for resources,and the more the bid, the higher priority access rights. In this way, network achievesthe purpose of the whole network utility maximization. At last, the simulationresults verify the effectiveness of this bandwidth allocation scheme, and also foundthat the uneven distribution of energy consuming lead to the network a lowaccumulated utility..In order to provide a load balancing mechanism for ad hoc network, wepropose an AODV protocol in the network layer. The random delay function inoriginal AODV is replaced by energy-related delay functions. We set nodes withhigher the energy shorter delay, avoiding excessive low energy node to participaterouting. Finally, we compare our algorithm with original one and the classicalalgorithm MMBCR to verify its validity. Then we combined the bandwidth allocation algorithm which showed in third chapter to maximize the network utility.