Indoor location estimation using smart antenna system with virtual fingerprint construction scheme

Regarding indoor location estimation, many smart positioning techniques have been proposed and they could be classified into several categories. Fingerprinting is one of the categories and its features are to build the indoor radio map during offline phase and to utilize the radio map to estimate the location during online phase. Creating indoor radio map is the most critical step, where the accuracy of location estimation depends on the distribution and number of reference points (RPs). Usually, the time required to collect received signal strengths (RSSs) is proportional to the number of 'real' RPs. To reduce the complexity of offline process, this paper proposes an efficient scheme, which is based on the well-known signal propagation model, to construct 'virtual' RPs in indoor radio map. The RSSs on virtual RPs are calculated and used to substitute for the training RSSs on 'real' RPs collected during offline phase. The proposed scheme can not only shorten the collecting time but also achieve high accuracy for indoor positioning. The indoor location estimation using smart antenna system (SAS) also requires longer time to collect all RSS information because of multiple antennas. By combining this scheme with SAS, we can easily obtain enough and valid RSSs information to build the indoor radio map in a more efficient way. Experimental results showed that applying virtual fingerprint construction scheme on SAS can decrease 33% 'real' RPs in indoor radio map without sacrificing the positioning accuracy.