SMARTTAP

Tapping 802.11 Access Point Infrastructure for Fine-Grained Indoor Location

 Partecipanti

Mappa

 Word cloud

Esplora la "nuvola delle parole (Word Cloud) per avere un'idea di massima del progetto.

 Obiettivo del progetto (Objective)

'On today’s mobile devices, location plays a pivotal role, enabling a raft of current applications such as navi- gation, location-based neighbourhood discovery, “check-in” style applications and targeted advertising. How- ever, the location stack on smartphones today largely relies on GPS, which does not work indoors. Today, phones scan for nearby WiFi access points (APs) and use the measured received signal strengths (RSS) to derive coarse indoor location. Owing to random fluctuations in received signal strength, this approach has an inherently-limited accuracy (5–10 metres) and can sometimes result in gross errors of up to 20 metres, resulting in a poor and unpredictable user experience. From conversations with potential users including museums, convention centres, and retailers, we have learned that a 5–10 metre accuracy does not suffice for the applications people want. For example, museums want to provide visitors with a mobile application that can show content specific to the current room, and nearest exhibit. Major retailers and brands are interested in providing location-based content and targeted offers to users shopping in stores, with a reliable location fix at the aisle level or better. The ability to provide a location fix with less than a metre accuracy will enable these applications, and also open up potentially exciting new applications such as augmented reality. There exist approaches that use ultrasound, RF, or Bluetooth hardware to localise to within less than a metre. However these typically do not work across smartphone platforms, and require the widespread deployment of specialised infrastructure that must be powered and networked to a backhaul. This creates a huge barrier to commercial success because it increases acquisition cost. Additionally, each sale requires going through a lengthy process of clearance from IT and facilities departments in the organisation buying the solution. We note an exciting opportunity to take the following fresh approach to indoor location that provides high accuracy and can work with the existing wireless AP infrastructure in the building. Most indoor spaces today already have such infrastructure, or plan to deploy it to provide visitor Internet access, and WiFi is available across smartphone platforms without exception. We also note the increasing prevalence of multi-antenna APs, mainly to bolster capacity and coverage with multiple-input, multiple-output (MIMO) techniques. We propose SmartTap, an indoor location system that uses MIMO-based angle-of-arrival (AoA) techniques at the AP to track wireless clients in real time as they roam about a building. When a client transmits a single frame over the air, each SmartTap AP receives the packet and performs an AoA analysis on the incoming signal. A backend server then fuses the AoA information from each AP together in order to get a physical location fix on the client. As clients roam around the building, the backend server updates the location of each. AoA is superior to the RSS based methods in use today because it is more accurate, more responsive, and requires fewer APs, making it more robust in buildings with sparse WiFi coverage such as retail stores. The expected outcome of the SmartTap project is a demonstration of the system localising clients in real time that can be shown to potential customers. We will develop a hardware module (the SmartTap) using off- the shelf components to tap into the baseband processor of already-deployed APs, with the goal of one-metre indoor localisation accuracy. We are confident that the cost savings, leveraging of existing infrastructure, and lack of need for extensive internal clearance processes will make SmartTap more attractive to potential customers than the status quo.'