An accurate and reliable Indoor Positioning System (IPS) applicable to most indoor scenarios has been sought for many years. The number of technologies, techniques, and approaches in general used in IPS proposals is remarkable. Such diversity, coupled with the lack of strict and verifiable evaluations, leads to difficulties for appreciating the true value of most proposals. This paper provides a meta-review that performed a comprehensive compilation of 62 survey papers in the area of indoor positioning. The paper provides the reader with an introduction to IPS and the different technologies, techniques, and some methods commonly employed. The introduction is supported by consensus found in the selected surveys and referenced using them. Thus, the meta-review allows the reader to inspect the IPS current state at a glance and serve as a guide for the reader to easily find further details on each technology used in IPS. The analyses of the meta-review contributed with insights on the abundance and academic significance of published IPS proposals using the criterion of the number of citations. Moreover, 75 works are identified as relevant works in the research topic from a selection of about 4000 works cited in the analyzed surveys.

https://www.mdpi.com/1424-8220/19/20/4507/pdf

A Meta-Review of Indoor Positioning Systems.

In the recent past, we have witnessed the adoption of different machine learning techniques for indoor positioning applications using WiFi, Bluetooth and other technologies. The techniques range from heuristically derived hand-crafted feature-based traditional machine learning algorithms, feature selection algorithms to the hierarchically self-evolving feature-based Deep Learning algorithms. The transient and chaotic nature of the WiFi/Bluetooth fingerprint data along with different signal sensitivity of different device configurations presents numerous challenges that influence the performance of the indoor localization system in the wild. This article is intended to offer a comprehensive state-of-the-art survey on machine learning techniques that have recently been adopted for localization purposes. Hence, we review the applicability of machine learning techniques in this domain along with basic localization principles, applications, and the underlying problems and challenges associated with the existing systems. We also articulate the recent advances and state-of-the-art machine learning techniques to visualize the possible future directions in the research field of indoor localization.

A Survey of Machine Learning Techniques for Indoor Localization and Navigation Systems

Location-based services have attracted significant attention since the concept of Industry 4.0 was proposed and the ‘Internet +’ era began owing to its social and commercial value. Many scholars have attempted to introduce machine learning in indoor fingerprint positioning to improve indoor positioning accuracy, enhance system robustness, reduce costs, and improve the performance of indoor positioning methods. A comprehensive overview of indoor positioning technology, its methods, and classifications is provided. Furthermore, a detailed review of the application of WiFi fingerprinting and machine learning methods in indoor positioning is presented, and the advantages and disadvantages of these methods when applied to indoor positioning are analyzed. This study summarizes the difficulties and challenges encountered by indoor positioning and suggests development directions. 

Overview of WiFi fingerprinting-based indoor positioning

Enhancing the accuracy of indoor visible light positioning systems with simple, real-time, and stable methods is one of the interesting challenges in recent research. In this paper, a relatively minor mean positioning error of 8 mm and a 42-52% improvement in computational time could be achieved within a real space of 1.2 m $\\times1.2$ m $\\times1.2$ m by transcending the serious limitations of the traditional k-nearest neighbors (KNN) algorithm. These disadvantages (slow execution time, high error formation) are a result of finding the nearest neighbors from all the fingerprints, averaging the Euclidean distances, and the excessive passivity of the K value. To overcome the above limitations of KNN, we proposed a maximum received signal strength recognition (MRR) technique and weighted optimum KNN (WOKNN) algorithm, which is a combination of optimum KNN (OKNN) and weighted KNN (WKNN). While MRR was used to reduce the computational time, WOKNN was used to solve the remaining problems. Specifically, OKNN was used to automatically determine the best number of nearest neighbors for each position in the area under consideration, and WKNN helped improve the errors that come from the Euclidean distance averaging process. Based on positive experimental results and a meaningful comparison with various versions of KNN, we demonstrated that the improved conventional KNN algorithm can achieve very high positioning accuracy and is totally suitable for several specific 2-D indoor positioning applications.

/pdf/high-precision-weighted-optimum-k-nearest-neighbors-7a6xxl9n5o.pdf

High Precision Weighted Optimum K-Nearest Neighbors Algorithm for Indoor Visible Light Positioning Applications

Machine Learning (ML) for Optical Communication (OC) is certainly a hot topic emerged recently and will continue to raise interest at least for the next few years. The rate of research development in this area is growing very rapidly. Novelty of this research direction resides mainly in the peculiarity of the application field, rather than in the methodological approaches, which are (at least up to now) state-of-the-art ML algorithms. Reviewing the literature shows that many of the ML algorithms have not yet been used in this area, and many of the OC applications are not considered yet, which reflects the fact that the research topic is pristine. Accordingly, tutorial investigations are quiet necessary in this filed to help researchers be aware about the last progressions and cavities of this field. Although several tutorials have been released recently, they considered this topic from OC view, and neglected ML view. However, it is required to have an investigations about the ML algorithms used in this subject. Accordingly, for the first time, this paper reviews ML for OC literature from ML viewpoint. This view could be really helpful because only OC experts work on ML for OC, and they are not ML experts, so it could really help them to have a comprehensive view on the ML subjects implantable in OC. It has worth to mention that compared with other works, this survey reviews much more investigations; therefore, it has more generality, and gives the reader to have a comprehensive overview on this topic.

A Survey on Machine Learning for Optical Communication [Machine Learning View]

WLAN-based indoor positioning algorithm has the characteristics of simple layout and low price, and it has gradually become a hotspot in both academia and industry. However, due to the poor stability of Wi-Fi signals, received signal strength (RSS) fingerprints of some adjacent reference positions are difficult to evaluate similarity when utilizing traditional distance-based calculation methods. By clustering these RSS fingerprints into one region, the commonly utilized KNN algorithm in the past cannot achieve accurate positioning in the region. For this, we introduce a concept of the insensitive region of the RSS fingerprint and a new algorithm named DBSCAN-KRF. This algorithm can delete noise sample and detect insensitive region, then, different methods are selected to achieve indoor positioning by judging the region of the estimated fingerprint sample, the KNN algorithm is selected when the region is sensitive, and random forest algorithm is selected when the region is insensitive. The experimental results show that the DBSCAN-KRF algorithm is superior while compared with other alternative indoor positioning algorithms.

/pdf/learning-to-improve-wlan-indoor-positioning-accuracy-based-1t9321uljt.pdf

Ya Huang

Papers

Learning to Improve WLAN Indoor Positioning Accuracy Based on DBSCAN-KRF Algorithm From RSS Fingerprint Data