Abstract:Recently, source localization is becoming a major research focus. The majority of the existing studies focus on the design of received signal strength (RSS)-based localization methods. However, when in the face of complicated environments with severe fading, RSS-based localization methods achieve relatively inferior accuracy performance, compared with fingerprinting-based localization methods. Nevertheless, traditional fingerprinting-based localization methods are subject to the condition that the source transmit power is known, which cannot be directly used in passive localization cases where the sensing nodes do not have the prior information on the source. In addition, the received sensing data may contain errors and then affect the location precision due to various abnormal conditions, such as device failure and malicious cases. In this paper, we propose a novel robust relative fingerprinting-based passive localization algorithm via a data cleansing approach. First, we figure out the fingerprint correlations property and introduce a new relative fingerprint framework. The key idea is that by exploring the correlations between the source fingerprint and the reference fingerprint database, the correction factors can be achieved to apply the fingerprint idea into the passive localization case. Second, we formulate a generalized modeling of the abnormal data in localization problem and propose a data cleansing approach which utilizes the sparse property of the abnormal data. Based on this, the negative influence of abnormal data can be further eliminated. Third, considering the sparse property of the source position, we use the sparse Bayesian learning in the matching process for the purpose of achieving more precise estimated source position. Simulation results demonstrate that the proposed algorithm achieves higher accuracy performance in passive source localization in terms of eliminating the abnormal data impairment.

An improved sensing data cleaning scheme for object localization in edge computing environment

Collision-Aware Churn Estimation in Large-Scale Dynamic RFID Systems

Collision-Aware Churn Estimation in Large-Scale Dynamic RFID Systems.

Leveraging Spatio-Temporal Redundancy for RFID Data Cleansing

A Model-Based Approach for RFID Data Stream Cleansing.

Efficiently Managing Uncertain Data in RFID Sensor Networks

STEP: A Time-Efficient Tag Searching Protocol in Large RFID Systems

Object Detection and Localization by Passive RF Backscattering System

Less is More: Efficient RFID-Based 3D Localization

Flexible and Time-Efficient Tag Scanning with Handheld Readers

A Range-Free Localization of Passive Rfid Tags Using Mobile Readers

A Framework For Processing Uncertain Rfid Data In Supply Chain Management

Robust Relative Fingerprinting-Based Passive Source Localization Via Data Cleansing

Complete and Fast Unknown Tag Identification in Large RFID Systems

An Efficient Method for Optimizing RFID Reader Deployment and Energy Saving

Model-driven data cleaning for signal processing system in sensor networks

Fuzzy logic-based approach to detecting a passive RFID tag in an outpatient clinic

RFID Dynamic Performance Measurement System Embedded in Multiscale Deep Learning

Adaptive and dynamic RFID tag anti-collision based on secant iteration

An RFID multi-Tag anti-collision algorithm based on Successive Interference Cancellation with power adaptive regulation

Efficient Object Localization Using Sparsely Distributed Passive RFID Tags