Menu:
Improving localisation in indoor wireless networks
Investigators: Tadhg Deasy & William Scanlon
The project involves the examination - via computer simulation - of two main types of WLAN based localisation to determine their performance in several different test bed layouts.
The first approach considered was deterministic in nature, whilst the second approach used a probabilistic Gaussian kernel;
both rely on Received Signal Strength Indication (RSSI) measurements to locate the client.
In general, using either technique the error in the position estimate was in the order of 6 m rms.
To improve performance, two novel refinement algorithms were developed and compared with an algorithm presented in the open literature. These algorithms depend on previous position estimates to improve the current estimate, each with varying degrees of success. One algorithm in particular, the Constrained Movement algorithm, gave a 40 per cent reduction in rms error when used in combination with deterministic localisation. Building on the simulation work and refinement algorithm development, a thorough measurement campaign was executed in two distinct test beds. This allowed a statistical comparison of simulated and measured radio maps as well as an investigation of the viability of using a simulated radio map to reduce the time required to set up the localisation scheme.
Finally, a new method of reducing the handoff delay for clients moving between Access Points (AP) using a Kalman filter is proposed. Using the client’s current estimated position, the filter can predict the client’s intended destination and track their velocity with reasonable accuracy. Using a simple criterion, the filter can successfully predict the next AP in 75 per cent of cases.
Publication Output - Examples
- T.P. Deasy & W.G. Scanlon, 'Simulation or measurement: the effect of radio map creation on indoor WLAN-based localisation accuracy,' Wireless Personal Comms., Vol. 42, 4, pp. 563–573, Sept. 2007. [doi: 10.1007/s11277-006-9211-x]
- T.P. Deasy & W.G. Scanlon, 'Stepwise algorithms for improving the accuracy of both deterministic and probabilistic methods in WLAN-based indoor user localisation,' Intl. Jnl. Wireless Information Networks, Vol. 11, 4, pp. 207–216, October 2004. [doi: 10.1007/s10776-004-1234-1]
- T.P. Deasy & W.G. Scanlon, 'Accuracy improvement algorithms for prediction of user location using receive signal strength indication in infrastructure WLANs,' 15th IEEE Intl. Symp. Personal, Indoor & Mobile Radio Comms. (PIMRC), Vol. 3, pp. 1757–1761, Sept. 2004.
Support
This project was funded by the European Social Fund (ESF).