Location determination using IEEE802.11 network radio signals and deterministic algorithm

Abstract

A location determination system can find the location of a mobile node by using sensor technology, location model and estimation algorithm. The outputs of such a system can be fed into another system like an IDS or a location aware information system. In indoor environments, the employment of IEEE 802.11 radio signals for this purpose has several advantages such as simplicity of the system, low cost of ownership and resiliency. The main target of this thesis is to find the optimum strategy for location determination using IEEE 802.11 radio signals and deterministic algorithm. To achieve this goal first, the underlying technology and existing algorithms are studied. Second, by statistical and numerical analysis of data obtained from several signal strength measurements, the influential factors on RSSI behavior are probed. With this knowledge, finally a number of improvements to the existing algorithms are proposed. Existing deterministic methods summarize a set of RSSI samples in a single scalar value. In this research the proposed combinational algorithm uses a combination of maximum and mean for summarization of a set of numeric data. This approach leads to improvement in the accuracy of the estimates. Moreover, an algorithm for detection and dealing with the problem of aliasing is suggested. By using an additional source of information in the RF map and a voting mechanism this algorithm deals with aliasing problem effectively and leads to elimination of large error distances in location estimation. Several simulated experiments show that suggested enhancements improve the accuracy of the existing algorithms up to 33%. The most efficient method is deterministic algorithm using maximum RSSI for summarization of RSSI values with the proposed anti aliasing technique applied. This approach manages to locate the mobile node within 2 to 6 meters from its true location.