Received Signal Strength Fingerprinting-Based Indoor Location Estimation Employing Machine Learning

Abstract

The fingerprinting technique is a popular approach to reveal location of persons, instruments or devices in an indoor environment. Typically based on signal strength measurement, a power level map is created first in the learning phase to align with measured values in the inference. Second, the location is determined by taking the point for which the recorded received power level is closest to the power level actually measured. The biggest limit of this technique is the reliability of power measurements, which may lack accuracy in many wireless systems. To this end, this work extends the power level measurement by using multiple anchors and multiple radio channels and, consequently, considers different approaches to aligning the actual measurements with the recorded values. The dataset is available online. This article focuses on the very popular radio technology Bluetooth Low Energy to explore the possible improvement of the system accuracy through different machine learning approaches. It shows how the accuracy–complexity trade-off influences the possible candidate algorithms on an example of three-channel Bluetooth received signal strength based fingerprinting in a one dimensional environment with four static anchors and in a two dimensional environment with the same set of anchors. We provide a literature survey to identify the machine learning algorithms applied in the literature to show that the studies available can not be compared directly. Then, we implement and analyze the performance of four most popular supervised learning techniques, namely k Nearest Neighbors, Support Vector Machines, Random Forest, and Artificial Neural Network. In our scenario, the most promising machine learning technique being the Random Forest with classification accuracy over 99%.

WLAN RSS-Based Fingerprinting for Indoor Localization: A Machine Learning Inspired Bag-of-Features Approach

“…Classification of raw RSS data features is carried out through KNN [ 39 , 43 ], probabilistic [ 44 ], SVM [ 45 ], DT [ 46 ], ensemble learning-based classification model [ 47 ], and the discriminant analysis classifier (DAC) [ 48 ], and comparisons are made with the proposed BoF-enabled approach. These are the most common machine learning models used in RSS fingerprinting systems [ 49 ]. The KNN method estimates labels based on k neighbor samples in the data.…”

Indoor Localization System Using Fingerprinting and Novelty Detection for Evaluation of Confidence

“…In [37], it is used BLE with multiple anchors and multiple radio channels to improve the reliability of power measurements. The accuracy-complexity trade-off affects the four most popular supervised learning techniques (k-Nearest Neighbors, Support Vector Machines, Random Forest, and Artificial Neural Network).…”

Automated Calibration of RSS Fingerprinting Based Systems Using a Mobile Robot and Machine Learning

“…The first approach treats the localisation process as a classification problem, in which the location of the device must be assigned to a particular class being either a singular point [ 11 , 28 ] or a larger area such as a room [ 29 ]. The proposed methods use different types of classifiers: Support Vector Machine (SVM) [ 11 ], Random Forest [ 11 , 21 , 28 ] and Artificial Neural Networks (including Convolutional ones) [ 11 , 12 , 29 ].…”

“…The latter approach is to use the ML to solve a regression problem, where the results are the coordinates of the localised device. The studies described in the literature typically utilise Neural Networks [ 21 , 30 , 31 , 32 ] and k-Nearest Neighbours (kNN) [ 11 , 33 ].…”

“…Another way to speed up system calibration is to measure the RSS values in a limited number of reference points and interpolate a complete radio map by fitting a propagation model or training machine learning algorithms [ 11 ]. Such an approach can also be used to improve localisation accuracy in systems, where the fingerprints are gathered in a traditional way [ 12 ].…”