Abstract
The GPS signals are not available inside the buildings; hence, indoor localization systems rely on indoor technologies such as Bluetooth, WiFi, and RFID. These signals are used for estimating the distance between a target and available reference points. By combining the estimated distances, the location of the target nodes is determined. The wide spreading of the internet and the exponential increase in small hardware diversity allow the creation of the internet of things (IoT)-based indoor localization systems. This chapter reviews the traditional and machine learning-based methods for IoT-based positioning systems. The traditional methods include various distance estimation and localization approaches; however, these approaches have some limitations. Because of the high prediction performance, machine learning algorithms are used for indoor localization problems in recent years. The chapter focuses on presenting an overview of the application of machine learning algorithms in indoor localization problems where the traditional methods remain incapable.
TopIntroduction
The popularity of the Internet, smartphones, and different kinds of wireless devices has enabled the provision of new services such as indoor localization and tracking systems. Indoor localization is the process of detecting the real-time location of wireless devices in an indoor environment with a bounded error rate. Indoor localization and tracking of mobile objects have extensive and increasing applications in different fields such as healthcare, advertisements, marketing, monitoring, security, building management, surveillance, and warehousing (Karimpour, 2019)(Khelifi, 2019). For example, in a hospital, tracking the assets, patients, and medical staff can increase the service quality and lead to efficient resource planning. As another example, the indoor localization of customers in a big shopping center can help send more efficient advertisements and analyze customers' behavior and shopping interests. Finally, the indoor localization of assets in a big warehouse can help to find the assets or free spaces faster and simpler.
The GPS signals are not available inside the buildings; hence the indoor tracking and localization systems try to use other signals such as Bluetooth Low Energy (BLE), WiFi, RFID, Wireless Sensor Networks (WSN), and Ultra-Wide Band signals (UWB). Receiving the signals from different sources allows the localization systems to merge the information and estimate the location of target assets. Recent advances in wireless communication modules, low energy BLE modules, sensors, memory chips, and processors have emerged a new generation of small and powerful devices that can store and run programs, measure the signal's strength and communicate over radio channels. The wide spreading of the Internet and the exponential increase in the diversity of small hardware allow us to create a network of devices that communicate over the Internet and form an Internet of Things (IoT). The IoT-based tracking systems allow the real-time positioning and monitoring of different assets. The precision of available localization systems generally depends on the underlying hardware and localization method. This chapter reviews the traditional and machine learning-based methods for IoT-based positioning systems and discusses their challenges and limitations.
An ideal indoor localization system should determine the exact location of the desired number of mobile targets inside a large building with minimal energy consumption, small mobile devices, and low cost. The main properties of an ideal indoor localization system are as follow:
Key Terms in this Chapter
Internet of Things: The Internet of Things (IoT) is a system of interconnected, internet-connected devices that are capable of collecting and transmitting data via a wireless network.
Machine Learning: Machine learning, a branch of artificial intelligence (AI), gains the computer systems the automatic self-learning ability via past experiences.
Supervised Learning: In supervised learning, a mathematical and statistical predictive model is constructed using a raw data set that is already tagged with correct labels.
Classification: The most known and commonly used supervised learning method is classification. This method categorizes new unlabeled samples into predefined classes.
Clustering: Clustering is an unsupervised learning technique that groups a set of objects into clusters based on similarity.
Indoor Localization: Indoor localization is the process of detecting the real-time location of wireless devices in an indoor environment with a bounded error rate.
Unsupervised Learning: The unsupervised learning technique trains the system using unlabeled observation data, which has not any prior information about the output value.