Indoor localization is of great importance for a range of pervasive applications, attracting many research efforts in the past decades. Most radio-based solutions require a process of site survey, in which radio signatures of an interested area are annotated with their real recorded locations. Site survey involves intensive costs on manpower and time, limiting the applicable buildings of wireless localization worldwide. In this study, we investigate novel sensors integrated in modern mobile phones and leverage user motions to construct the radio map of a floor plan, which is previously obtained only by site survey. On this basis, we design LiFS, an indoor localization system based on off-the-shelf WiFi infrastructure and mobile phones. LiFS is deployed in an office building covering over 1600m2, and its deployment is easy and rapid since little human intervention is needed. In LiFS, the calibration of fingerprints is crowdsourced and automatic. Experiment results show that LiFS achieves comparable location accuracy to previous approaches even without site survey.

/pdf/locating-in-fingerprint-space-wireless-indoor-localization-wrq0pihyq7.pdf

Locating in fingerprint space: wireless indoor localization with little human intervention

WiFi localization, the task of determining the physical location of a mobile device from wireless signal strengths, has been shown to be an accurate method of indoor and outdoor localization and a powerful building block for location-aware applications. However, most localization techniques require a training set of signal strength readings labeled against a ground truth location map, which is prohibitive to collect and maintain as maps grow large. In this paper we propose a novel technique for solving the WiFi SLAM problem using the Gaussian Process Latent Variable Model (GPLVM) to determine the latent-space locations of unlabeled signal strength data. We show how GPLVM, in combination with an appropriate motion dynamics model, can be used to reconstruct a topological connectivity graph from a signal strength sequence which, in combination with the learned Gaussian Process signal strength model, can be used to perform efficient localization.

WiFi-SLAM Using G aussian Process Latent Variable Models

Indoor localization is of great importance for a range of pervasive applications, attracting many research efforts in the past decades. Most radio-based solutions require a process of site survey, in which radio signatures of an interested area are annotated with their real recorded locations. Site survey involves intensive costs on manpower and time, limiting the applicable buildings of wireless localization worldwide. In this study, we investigate novel sensors integrated in modern mobile phones and leverage user motions to construct the radio map of a floor plan, which is previously obtained only by site survey. Considering user movements in a building, originally separated RSS fingerprints are geographically connected by user moving paths of locations where they are recorded, and they consequently form a high dimension fingerprint space, in which the distances among fingerprints are preserved. The fingerprint space is then automatically mapped to the floor plan in a stress-free form, which results in fingerprints labeled with physical locations. On this basis, we design LiFS, an indoor localization system based on off-the-shelf WiFi infrastructure and mobile phones. LiFS is deployed in an office building covering over 1,600  m $^2$ , and its deployment is easy and rapid since little human intervention is needed. In LiFS, the calibration of fingerprints is crowdsourced and automatic. Experiment results show that LiFS achieves comparable location accuracy to previous approaches even without site survey.

Smartphones Based Crowdsourcing for Indoor Localization

Robotics and autonomous systems

Smart services are an important element of the smart cities and the Internet of Things (IoT) ecosystems where the intelligence behind the services is obtained and improved through the sensory data. Providing a large amount of training data is not always feasible; therefore, we need to consider alternative ways that incorporate unlabeled data as well. In recent years, deep reinforcement learning (DRL) has gained great success in several application domains. It is an applicable method for IoT and smart city scenarios where auto-generated data can be partially labeled by users’ feedback for training purposes. In this paper, we propose a semisupervised DRL model that fits smart city applications as it consumes both labeled and unlabeled data to improve the performance and accuracy of the learning agent. The model utilizes variational autoencoders as the inference engine for generalizing optimal policies. To the best of our knowledge, the proposed model is the first investigation that extends DRL to the semisupervised paradigm. As a case study of smart city applications, we focus on smart buildings and apply the proposed model to the problem of indoor localization based on Bluetooth low energy signal strength. Indoor localization is the main component of smart city services since people spend significant time in indoor environments. Our model learns the best action policies that lead to a close estimation of the target locations with an improvement of 23% in terms of distance to the target and at least 67% more received rewards compared to the supervised DRL model.

Semisupervised Deep Reinforcement Learning in Support of IoT and Smart City Services

Currently the most accurate WLAN positioning systems are based on the fingerprinting approach, where a "radio map" is constructed by modeling how the signal strength measurements vary according to the location. However, collecting a sufficient amount of location-tagged training data is a rather tedious and time consuming task, especially in indoor scenarios -- the main application area of WLAN positioning -- where GPS coverage is unavailable. To alleviate this problem, we present a semi-supervised manifold learning technique for building accurate radio maps from partially labeled data, where only a small portion of the signal strength measurements need to be tagged with the corresponding coordinates. The basic idea is to construct a non-linear projection that maps high-dimensional signal fingerprints onto a two-dimensional manifold, thereby dramatically reducing the need of location-tagged data. Our results from a deployment in a real-world experiment demonstrate the practical utility of the method.

/pdf/semi-supervised-learning-for-wlan-positioning-1281vkm64p.pdf

Semi-supervised learning for WLAN positioning

Using landmark-based navigation instructions is widely considered to be the most effective strategy for presenting navigation instructions. Among other things, landmark-based instructions can reduce the user's cognitive load, increase confidence in navigation decisions and reduce the number of navigational errors. Their main disadvantage is that the user typically focuses considerable amount of attention on searching for landmark points, which easily results in poor awareness of the user's surroundings. In indoor spaces, this implies that landmark-based instructions can reduce the attention the user pays on advertisements and commercial displays, thus rendering the assistance commercially inviable. To better understand how landmark-based instructions influence the user's awareness of her surroundings, we conducted a user study with $20$ participants in a large national supermarket that investigated how the attention the user pays on her surroundings varies across two types of landmark-based instructions that vary in terms of their visual demand. The results indicate that an increase in the visual demand of landmark-based instructions does not necessarily improve the participant's recall of their surrounding environment and that this increase can cause a decrease in navigation efficiency. The results also indicate that participants generally pay little attention to their surroundings and are more likely to rationalize than to actually remember much from their surroundings. Implications of the findings on navigation assistants are discussed.

Influence of landmark-based navigation instructions on user attention in indoor smart spaces

WiFi fingerprinting is currently one of the most popular techniques for indoor localization as it provides reasonable positioning accuracy while at the same time being able to exploit existing wireless infrastructure. To facilitate calibration efforts and to overcome fluctuations in location measurements, many indoor WiFi positioning systems utilize a discrete partitioning, e.g., a grid or a topological map, of the space where the positioning is being deployed. A major limitation of this approach, however, is that instead of considering spatial similarities in the signal environment, the partitioning is typically based on an uniform division of the space or topological constraints (e.g., rooms and walls). This can significantly decrease positioning accuracy when the signal environment is not sufficiently stable across all partitions. Moreover, current solutions provide no support for identifying partitions that are not compatible with the current wireless deployment. To overcome these limitations, we propose AWESOM (Activations Weighted by the Euclidean-distance using Self-Organizing Maps), a novel measure for automatically creating a discrete partitioning of the space where the WiFi positioning is being deployed. In addition to enabling automatic construction of a discrete partitioning, AWESOM provides a measure for evaluating the goodness of a given partitioning for a particular access point deployment. AWESOM also enables identifying partitions where additional access points should be deployed. We demonstrate the usefulness of AWESOM using data collected from two large scale deployments of a proprietary wireless positioning system in a hypermarket environment.

AWESOM: automatic discrete partitioning of indoor spaces for wifi fingerprinting

We present Ma$$iv, an intelligent mobile grocery assistant that provides support for the customer during the entire shopping process To guide the design of Ma$$iv, we conducted a user study that explored customer preferences regarding features in a mobile grocery aid We first describe the study and its results, after which we introduce the design principles and design of Ma$$iv We also describe the features that Ma$$iv supports and discuss functionalities that we are integrating into Ma$$iv As part of the discussion, we describe technical challenges that we have encountered during our development efforts

Teemu Pulkkinen

Papers

Semi-supervised learning for WLAN positioning

Influence of landmark-based navigation instructions on user attention in indoor smart spaces

AWESOM: automatic discrete partitioning of indoor spaces for wifi fingerprinting

Ma$$iv -- An Intelligent Mobile Grocery Assistant

Semi-supervised Convolutional Neural Networks for Identifying Wi-Fi Interference Sources