Method for yielding a database of location fingerprints in WLAN

Binghao Li, Y. Wang, H.K. Lee, Andrew G. Dempster, Chris Rizos

IEE Proceedings - Communications · 2005 · 251 citations · 19 references

Concepts

TL;DR

Location fingerprinting in WLAN positioning has attracted much attention, yet generating the database of fingerprints remains a key challenge because conventional methods ignore spatial correlation among adjacent reference points and require laborious training. The authors aim to develop a kriging‑based approach that improves estimation accuracy while reducing workload and training time. The method exploits kriging to model spatial correlation among adjacent reference points, enabling more accurate estimation with fewer training samples. This makes the fingerprinting technique more flexible and easier to implement.

Abstract

Location fingerprinting in wireless LAN positioning has received much attention recently. One of the key issues of this technique is generating the database of fingerprints. The conventional method does not utilise the spatial correlation of measurements sampled at adjacent reference points, and the training process is not an easy task. A new method based on kriging is presented which can not only achieve more accurate estimation, but can also greatly reduce the workload and save training time. This can make the fingerprinting technique more flexible and easier to implement.

References

19