Progressive RSS Data Augmenter With Conditional Adversarial Networks

Liangfeng Chen,Haibo Tan,Bo Lv,Shutao Zhang

doi:10.1109/access.2020.2971269

Abstract

Accuracies of most fingerprinting approaches for WiFi-based indoor localization applications are affected by the qualities of fingerprint databases, which are time-consuming and labor-intensive. Recently, many methods have been proposed to reduce the localization accuracy reliance on the qualities of the established fingerprint databases. However, studies on establishing fingerprint databases are relatively rare under the condition of sparse reference points. In this paper, we propose a novel data augmenter based on the adversarial networks to build fingerprint databases with sparse reference points. Additionally, two conditions of these networks are designed to generate data effectively and stably, which are 0–1 sketch and Gaussian sketch. Based on the networks, we design two augmenters with different cyclic training strategies to evaluate the augmenting effects comparatively. Meanwhile, five quantitative evaluation metrics of the augmenters are proposed from two perspectives of the artificial experiences and the data features, and some of them are also used as the gradient penalties for generators. Finally, experiments corresponding to these metrics and localization accuracies demonstrate that the data augmenter with the 0–1 sketch adversarial network is more efficient, effective and stable totally.

Highlights

The localization awareness is a trend for the hyper-connected society to grow rapidly
A fingerprint database is constructed with received signal strength (RSS) heard from access points’ (APs) in an indoor area, while signal receivers are at the centers of the predefined grids (each grid center is defined as a reference point (RP))
Reasons are: a) all networks are good as long as the training data is enough, b) the fake data is valid except GAN, and c) accuracies are depended on k-nearest neighbors (KNN) after RPs are enough

Summary

INTRODUCTION

The localization awareness is a trend for the hyper-connected society to grow rapidly. L. Chen et al.: Progressive RSS Data Augmenter with Conditional Adversarial Networks original localization information are learned efficiently over the conventional methods. Reference [60] proposes the RMapTAFA scheme to construct a radio map to improve localization performance of both pedestrian trajectory tracking and stationary point positioning. These methods could decrease the reliance on the accuracies of the established fingerprint databases.

DATA AUGMENTER BASED ON GAN

PRELIMINARY OF DATA SAMPLES

EXPERIMENTS AND RESULTS

CONCLUSION