Wireless Local Area Networks (WLANs) have become a major success in telecommunications during the last few years, due to advantages such as mobility, flexibility, and easier maintenance. A device called an access point (AP) acts as a base station in WLAN for connecting a group of users to the network via radio signal. During the planning of such a network an important problem is to determine the optimal number of these devices and their placement/distribution so that coverage, capacity, and physical security are maximised at minimum cost. In this thesis we are using continuous optimisation techniques to optimise the number of APs and their distribution while cost of deployment is reduced and physical security of the network is enhanced. To find the number and placement of APs, we developed a multi-objective functions model based on path losses and power for free space environments. The two functions in the models are combined by using a balancing parameter. Since it is recognised that some of the objectives can be handled one at a time, in another approach, we followed a step-by-step procedure. We start with a novel optimisation model based on path losses for indoor environments including obstacles. Cost of deployment is saved by finding the minimum number of APs ensuring that the path loss at each test point/receiver is below the given maximum path loss. Next, the physical security of the network is enhanced by placing the APs far from places accessible to unauthorised users to reduce the risk of intrusion into the network. This is achieved in the framework of the model by introducing potential unauthorised users in unauthorised areas for whom coverage is minimised. Due to the presence of obstacles in indoor buildings, the path loss function is discontinuous. Therefore, the objective functions are very complicated and most of the existing optimisation algorithms cannot be applied to solve the problem. We use a global optimisation algorithm that is not used by other researchers to solve the same problem. To validate the accuracy of the optimisation model and performance of the numerical methods, we run tests on several indoor buildings and use wide range of WLAN parameters. The results demonstrate the quality of our model and algorithm. Based on the proposed model and algorithm, we developed a software to assist the network designers in planning wireless LANs. / Doctor of Philosophy
| Identifer | oai:union.ndltd.org:ADTP/266109 |
| Date | January 2007 |
| Creators | Kouhbor, Shahnaz |
| Publisher | University of Ballarat |
| Source Sets | Australiasian Digital Theses Program |
| Detected Language | English |
| Rights | Copyright Shahnaz Kouhbor |
Page generated in 0.0023 seconds