Research On Mixed Service Capacity In Multi-User Wireless Network

Nowadays, wireless and mobile communication is undergoing explosive development, which revolutionizes the design of wireless network and system. Traditional service requirement for voice service is being replaced by various kinds of application based on data services. The rapid development of multi-media service makes the wireless network evolve into a fresh stage with various kinds of service types and multi-user. The problem on capacity is always a hot issue both in academy and industry. However, concering recent researches and designing methods only focuse on voice service, and there is no definition and theoretical analysis on what capacity is and how a novel capacity designing method works based on different features of various service types. In order to solve these problems, in this thesis, as described below, analysis and research are conducted on mixed service capacity in multi-user wireless network.First of all, this thesis summarizes what kinds of services exists, how service and traffic model is categorized, and the exsiting definition about capacity in current wireless network.Second, exsting capacity designing methods is elaborated, and advangtages and disadvantaged are concluded by comparing these methods.Third, according to features of data service and the situation that data and voice services co-exsit, a novel capacity designing method is built up based on multi-dimensional markov chain. Through this mathematical tool, the relationship among features of data and voice service, number of users, setting paratemeters in wireless network, and performances in system level has been defined by the analytical equations. This model is made come true by MFC (a programming tool in Visual Studio). The proposition of this model and method provides an effective tool for engineers so that wireless network can be designed and configured much more accurately than before. Last but not the least, a definition and theoretica method on service load and service capacity, which are suitable for all types of services, have been proposed. Service load represents the load influenc introuduced into system by the access of this service. And service capacity represents the theoretical maximum number of this service which can be afforded by a specific wireless network. In this thesis, a research about the influenc on the service capacity, which is introduced by service features (especially self-similarity in data service), and system settings, has been conducted. And a fresh resource allocation scheme, based on service features, has been proposed and compared with traditional scheme based on users in aspects of service capacity and user satisfaction.Above all, the work in this thesis can provide theoretical goals and guidances for engineers in design of wireless network, which has significant influence on both future practical use and academic research.