Device-Free Localization With Multidimensional Wireless Link Information

As an emerging technique with promising application prospects, device-free localization (DFL) could estimate the location of target within the deployment area of wireless networks (WNs) while eliminating the need to equip the target with a wireless device. However, one major disadvantage of this technique is that it needs several wireless links travelling through the deployment area to guarantee good performance. To overcome this problem, a novel multidimensional wireless-link-information-based DFL scheme is proposed. Different from a traditional DFL scheme that scans wireless links sequentially with one frequency and one transmission power level, the proposed scheme makes full use of multiple frequencies and multiple transmission power levels to enrich the link measurement information. Meanwhile, motivated by the fact that the location information of the target is not only sparse but also changes slowly and continuously over time, we present a novel recursive compressive sensing algorithm to reconstruct the location information from undersampled measurements. The experimental results demonstrate the outstanding performance of the proposed scheme.