Xuemei Guo

An Exponential-Rayleigh Model for RSS-Based Device-Free Localization and Tracking

A common technical difficulty in device-free localization and tracking (DFLT) with a wireless sensor network is that the change of the received signal strength (RSS) of the link often becomes more unpredictable due to the multipath interferences. This challenge can lead to unsatisfactory or even unstable DFLT performance. This work focuses on developing a new RSS model, called Exponential-Rayleigh (ER) model, for addressing this challenge. Based on data from our extensive experiments, we first develop the ER model of the received signal strength. This model consists of two parts: the large-scale exponential attenuation part and the small-scale Rayleigh enhancement part. The new consideration on using the Rayleigh model is to depict the target-induced multipath components. We then explore the use of the ER model with a particle filter in the context of multi-target localization and tracking. Finally, we experimentally demonstrate that our ER model outperforms the existing models. The experimental results highlight the advantages of using the Rayleigh model in mitigating the multipath interferences thus improving the DFLT performance.

A real-time device-free localization system using correlated RSS measurements

Device-free localization (DFL) with wireless sensor networks (WSN) is an emerging technology for target localization, which has received much attention in the area of Internet of Things. Received signal strength (RSS) measurements are the key to realize DFL and mainly affects the localization performance. Most existing approaches need to measure the RSS of all the wireless links in WSN, which take much time on measurement process and localization algorithm due to the large amounts of RSS data, thus they are inefficient, especially in the case of target tracking. In this paper, by making full use of the consecutiveness of motion, we present an efficient measurement strategy based on a small set of correlated wireless links. Furthermore, a lightweight compressed maximum matching select (CMMS) algorithm is proposed to localize target, which only needs a small-scale matrix-vector product operating for one estimation. The proposed approach can significantly reduce the number of RSS measurements and improve the real-time capability of the DFL system. Experimental results demonstrate the superior performance of the proposed method in the context of target localization and tracking.

A Novel Infrared Motion Sensing System for Compressive Classification of Physical Activity

Infrared radiation changes (IRCs) induced by human activity can provide important information about activity patterns. This paper presents an IRC-based compressive classification method for recognizing the physical activities of interest in home-based assisted living. To fully capture the IRC compressively, a multi-view infrared motion sensing system is developed, which consists of three IRC sensing modules, that is, one module on the ceiling and two modules on opposite tripods facing each other. A pilot study is conducted in the context of classifying six typical physical activities with the incorporation of the classification techniques, including hidden Markov model and support vector machine, which demonstrates the effectiveness of our system.

Heterogeneous Bayesian compressive sensing for sparse signal recovery

This study focuses on the issue of sparse signal recovery with sparse Bayesian learning in the context of a heterogeneous noise model, called by the heterogeneous Bayesian compressive sensing. The main contribution is to exploit the capability of noise variance learning in performance improvement and applicability enhancement. Experimental results on synthetic and real-world data demonstrate that heterogeneous Bayesian compressive sensing has superior performance in terms of accuracy and sparsity for both homogeneous and heterogeneous noise scenarios.

An Exponential-Rayleigh signal strength model for device-free localization and tracking with wireless networks

We present a new statistic signal strength model, called Exponential-Rayleigh (ER) model, for device-free (DF) target localization and tracking issues in this paper. It is a single target measurement model for radio frequency (RF) based on received signal strength (RSS) measurement in outdoor regions. The model is a non-linear function between RSS measurements and target motion state. It consists of three parts: the largescale exponential attenuation part, the small-scale Rayleigh enhancement part and the noise. Different from the proposed models, while reserving the large-scale attenuation, we mainly present the small-scale Rayleigh enhancement model in ER model. The Rayleigh part depicts the multi-path caused by single target so as to reduce the multi-path error. In the context of localization and tracking experiment using particle filter, we validate the effectiveness of ER model.

Floor Pressure Imaging for Fall Detection with Fiber-Optic Sensors

Falls are a major health risk for the elderly, decreasing their ability to live independently. This article proposes a novel feature-specific floor pressure imaging system that uses a smart floor embedded with fiber sensors. The potential applications include indoor human-activity analysis for intelligent surveillance. In particular, the authors focus on the application of fall detection in a bathroom scenario. Their fall detection method consists of two steps: floor pressure imaging for target posture classification and the fall event decision based on the classified target postures. Fall detection experiments validated the proposed method. This article is part of a special issue on domestic pervasive computing.

A hierarchical RSS model for RF-based device-free localization

One common challenge in device-free localization (DFL) with radio frequency sensor networks is that the variations of received signal strength (RSS) become less sensitive to shadow fading and the performance will be degenerated in a cluttered environment. To address this problem, this paper develops a hierarchical model that contributes to refining the description granularity of RSS variations and enhancing the sensitivity of RSS variations to shadow fading. The novelty is twofold. First, the Exponential-Rayleigh model and the diffraction-based model are integrated to form a two degree-of-freedom parametrization that can capture not only the target-induced multipath interferences but also the diffraction fading contributions. Second, the link dependent characterizations of RSS variations are explored by parameterizing the model coefficients in terms of fade level. The proposed model is experimentally validated in the context of radio tomographic imaging (RTI).

A Smart Fiber Floor for Indoor Target Localization

This article presents a proof-of-concept study for a smart fiber floor. The task of the smart floor is to create a floor pressure distribution map with an embedded fiber sensor array for indoor target localization. Among various potential applications, the authors are interested in building a floor-centric human-following system, in which the smart floor can be used to guide a robot to track a person of interest in a computationally efficient way. The key to such a smart floor is the deployment strategy of fiber sensors. Two aspects are important in developing this deployment strategy. One is the sensing efficiency, which is motivated by using the minimal amount of fiber sensors or reducing the amount of redundant fiber sensors. Another is the data efficiency in terms of the capability of data self-association. To address these two concerns, the authors developed labeled uniquely decipherable (LUD) code to specify the deployment configuration of fiber sensors. Compared with the standard UD code, the LUD code allows for exploring the use of heterogeneous target sensitivity of fiber sensors in further reducing the number of the required fiber sensors and offering the capability of data self-association. This article is part of a special issue on smart spaces.

A Diffraction Based Modified Exponential Model for Device-Free Localization with RSS Measurements

Radio frequency (RF) based Device-free localization and tracking (DFLT) monitors the change in received signal strength (RSS) measurements to locate the targets without carrying any electronic devices in the sensing area covered by a RF sensor network. This paper presents a new modified exponential model to accurately describe the relationship between the RSS measurements and the target state, which can effectively predict the variation of RSS when the target is present on line-of-sight (LOS) path or non-line-of-sight (NLOS) path. Based on the diffraction theory, we first show that the RSS attenuation caused by target mainly depends on two factors: the target-nodes distance and the target-link distance, which can be exploited to depict the change of RSS on LOS and NLOS, respectively. By taking into account these two factors, we then develop our model, and validate it with single link experiments. We finally explore the use of the proposed model with particle filter for DFLT, and demonstrate that our model can improve the DFLT performance by conducting actual experiments.

A compressed infrared motion sensing system for human-following robots

This paper develops a compressed infrared motion sensing system for mobile robots to detect and localize a moving human target in the vicinity. The proposed sensing system consists of three compressed infrared bearing sensor arrays for generating the bearing measurements of a human target from three perspectives. Then, human location is inferred by fusing these bearing measurements with least square method. The bearing sensor array is composed of pyroelectric infrared (PIR) sensors, and we employ the compressed sensing paradigm for the bearing sensor array design for reducing the number of required PIR sensors. A sensing system prototype is developed and evaluated in the context of human-following with mobile robots.