Compressive Sensing Based Device-Free Multi-Target Localization Using Quantized Measurement

Abstract

Device-free localization (DFL), requiring no extra devices equipped with a target, is an important field of research on the Internet of Thing (IoT). Energy efficiency issue is essential for the development of the IoT, but seldom of the existing papers are focus on it. So we investigate this issue with quantized data of only several bits under the compressive sensing (CS) framework, which can both reduce the required wireless link number and the bit number in the DFL scheme. First, through exploiting the discrete property of CS theory, we calculate the discrete measurement probability bypass computing the complex or uncalculated measurement probability density function (pdf), which can well represent the measurement distribution characteristic. Second, we design a unique quantization scheme for each wireless link according to their measurement probability and build a novel DFL model by analyzing the quantization error. Third, a new DF-QVBI algorithm is proposed to recover the target location, which can make great use of the quantization error. Finally, numerical simulations show the superiority and robustness of the proposed method.