In ambient assistive living (AAL) it is of a paramount importance to be able to detect, localize and estimate the activities of persons living at their homes. Specially, estimating an accurate person's localization and detailed movement patterns in everyday activities, are very valuable for health monitoring and safety assessment. The indoor positioning and navigation (IPIN) research community mainly concentrates on the use of radio beaconing for trilateration with WiFi, Bluetooth or UWB, acoustic beaconing, inertial pedestrian dead-reckoning, and the fusion of several approaches, using fingerprinting or Bayesian estimators. In the area of activity recognition (AR) authors use different kind of sensors such as smart floors, binary sensors attached to common objects, or infer the proximity to objects using Bluetooth beacons. In this paper we want to join these two different field approaches (IPIN and AR) by proposing an indoor localization method that make use of smart floor information, binary sensors, and the signal strength received at a smartwatch coming from BLE beacons deployed in a smarlab. We use the smart floor as a ground truth in order to estimate location accuracy of persons in a totally unobtrusive way. The localization results, for a person moving in the smart livinglab during 10 days, showing accuracies below 1.5 meter in 80% of the cases. The proposed approach can help the tracking of multiple persons living together and also serve as a complement to improve the performance of location-aware activity recognition algorithms.

Indoor positioning is one of the biggest challenges of many Location Based Services (LBS), especially if the target users are pedestrians, who spend most of their time in roofed areas such as houses, offices, airports, shopping centres and in general indoors. Providing pedestrians with accurate, reliable, cheap, low power consuming and continuously available positional data inside the buildings (i.e. indoors) where GNSS signals are not usually available is difficult. Several positioning technologies can be applied as stand-alone indoor positioning technologies. They include Wireless Local Area Networks (WLAN), Bluetooth Low Energy (BLE), Ultra-Wideband (UWB), Radio Frequency Identification (RFID), Tactile Floor (TF), Ultra Sound (US) and High Sensitivity GNSS (HSGNSS). This paper evaluates the practicality and fitness-to-the-purpose of pedestrian navigation for these stand-alone positioning technologies to identify the best one for the purpose of indoor pedestrian navigation. In this regard, the most important criteria defining a suitable positioning service for pedestrian navigation are identified and prioritised. They include accuracy, availability, cost, power consumption and privacy. Each technology is evaluated according to each criterion using Analytic Hierarchy Process (AHP) and finally the combination of all weighted criteria and technologies are processed to identify the most suitable solution.