Direct Localisation using Ray-tracing and Least-Squares Support Vector Machines

This paper evaluates a novel scheme for direct 2D localization that employs least-squares support vector machines, using ray-tracing data. The scheme does not require non-line-of-sight identification or mitigation, which means it can be applied under any conditions. The approach requires perfect knowledge of base-station positions and the ray-tracing data is location specific. This approach shows that when an outage probability of twenty percent is considered, the mobile's location can be determined to within 15m of accuracy in a dense urban environment. Usage and application contexts for this approach are also provided.