Indoor Space Classification Using Cascaded LSTM

Abstract

Indoor space classification is an important part of localization that helps in precise location extraction, which has been extensively utilized in industrial and domestic domain. There are various approaches that employ Bluetooth Low Energy (BLE), Wi-Fi, magnetic field, object detection, and Ultra Wide Band (UWB) for indoor space classification purposes. Many of the existing approaches need extensive pre-installed infrastructure, making the cost higher to obtain reasonable accuracy. Therefore, improvements are still required to increase the accuracy with minimum requirements of infrastructure. In this paper, we propose an approach to classify the indoor space using geomagnetic field (GMF) and radio signal strength (RSS) as the identity. The indoor space is an open big test bed divided into different indiscernible subspace. We collect GMF and RSS at each subspace and classify it using cascaded Long Short Term Memory (LSTM). The experimental results show that the accuracy is significantly improved when GMF and RSS are combined to make distinct features. In addition, we compare the performance of the proposed model with the state-of-the-art machine learning methods.

An Overview of Machine Learning Techniques for Radiowave Propagation Modeling

“…As an example, in [31], the authors used training data coming from UAV, Wi-Fi and cellular base stations. In [45], the authors used a fusion of Wi-Fi RSS and geomagnetic field (GMF) data. The authors also found that by using only RSS or GMF data, the accuracy of their ML model deteriorated.…”

“…Many different variations of RNNs have been used for propagation modeling purposes, such as the echo state networks (ESNs), [57], Elman RNNs [53], standard RNNs [32], [58] and gated RNNs that use a gated recurrent unit (GRU) instead of a standard recurrent unit cell [31], [32]. The most important and popular type of RNN is the long short-term memory (LSTM) RNN that can capture longer dependencies in the input data, compared to the other types of RNNs [30], [32], [33], [45], [59]. In propagation modeling problems, the input sequence to the RNN is usually spatial [32] or temporal [31], [58].…”