Received-Signal-Strength Threshold Optimization Using Gaussian Processes

Abstract

Abstract-There is a big trend nowadays to use event-triggered proximity report for indoor positioning. This paper presents a generic received-signal-strength (RSS) threshold optimization framework for generating informative proximity reports. The proposed framework contains five main building blocks, namely the deployment information, RSS model, positioning metric selection, optimization process and management. Among others, we focus on Gaussian process regression (GPR) based RSS models and positioning metric computation. The optimal RSS threshold is found through minimizing the best achievable localization root-mean-square-error formulated with the aid of fundamental lower bound analysis. Computational complexity is compared for different RSS models and different fundamental lower bounds. The resulting optimal RSS threshold enables enhanced performance of new fashioned low-cost and low-complex proximity report based positioning algorithms. The proposed framework is validated with real measurements collected in an office area where bluetooth-low-energy (BLE) beacons are deployed.

Received Signal Strength-Based Indoor Localization Using Hierarchical Classification

“…In our experiment, the direct use of KNN, multilayer NN and SVM for flat classification-based object position estimation are selected as the benchmarks for performance evaluation. Specifically, the multilayer NN has 8 fully-connected hidden layers with 10,18,27,35,22,16,15 and 22 neurons, respectively, and they adopt the tanh activation function. The cross-entropy loss is employed in the training process, where 85% of the training data is used for optimizing the connection weights while the remaining 15% is applied for cross validation to avoid overfitting.…”

“…Firstly, Wi-Fi access points (APs) have been extensively deployed in indoor environments; secondly, measuring Wi-Fi RSS is readily available in the current Wi-Fi terminals. Many regression techniques have become available for Wi-Fi RSS indoor positioning and they include the distance-based [13,14] and Gaussian Processes (GP)-based techniques [15,16]. In fact, the use of the standard log-normal model for indoor positioning (see, e.g., [17]) can be considered as a regression technique as well.In this work, we shall take a different approach and consider the fingerprint-based indoor localization method.…”

An Implementation Approach and Performance Analysis of Image Sensor Based Multilateral Indoor Localization and Navigation System

“…They derived the Cramer-Rao lower bound of position estimation for improving the overall performance. Meanwhile, a generic framework of RSS has been introduced to enrich the positioning performance [24]. Alternatively, the combination of spatiotemporal constraints in RSS fingerprint-based localization has achieved the same purpose [25].…”

Online Sparse Multi-Output Gaussian Process Regression and Learning

“…Their parameters together with other model parameters such as the measurement noise precisions are collectively called hyperparameters that also need to be learned from the training data. GP regression has found diverse applications in indoor localization [3]- [5], image processing and visual tracking [6], [7], change point detection [8] and sensor (robotic) network management [9], [10]. It has also been used for wireless channel prediction/identification [11], [12] and state transition dynamics modeling in Bayesian filtering [13]- [15].…”