Accuracy in WiFi Access Point Position Estimation Using Round Trip Time

Abstract

WiFi Round Trip Time (RTT) unlocks meter level accuracies in user terminal positions where no other navigation systems, such as Global Navigation Satellite Systems (GNSS), are able to (e.g., indoors). However, little has been done so far to obtain a scalable and automated system that computes the position of the WiFi Access Points (WAP) using RTT and that is able to estimate, in addition to the position, the hardware biases that offset the WiFi ranging measurements. These biases have a direct impact on the ultimate position accuracy of the terminals. This work proposes a method in which the computation of the WiFi Access Points positions and hardware biases (i.e., products) can be estimated based on the ranges and position fixes provided by user terminals (i.e., inverse positioning) and details how this can be improved if raw GNSS measurements (pseudoranges and carrier phase) are also available in the terminal. The data setup used to obtain a performance assessment was configured in a benign scenario (open sky with no obstructions) in order to obtain an upper boundary on the positioning error that can be achieved with the proposed method. Under these conditions, accuracies better than 1.5 m were achieved for the WAP position and hardware bias. The proposed method is suitable to be implemented in an automated manner, without having to rely on dedicated campaigns to survey 802.11mc-compliant WAPs. This paper offers a technique to automatically estimate both mild-indoor WAP products (where terminals have both Wi-Fi RTT and GNSS coverage) and deep-indoor WAP (with no GNSS coverage where the terminals obtain their position exclusively from previously estimated mild-indoor WAPs).

Improving Fingerprint-Based Positioning by Using IEEE 802.11mc FTM/RTT Observables

“…This information is normally not provided by the APs in private networks. The authors in [ 27 ] propose a double-iteration approach to mitigate this issue, where the APs positions are computed in a first stage by using a GNSS solution. Then, in a second stage, these positions feed a hybrid GNSS/FTM-based Wi-Fi position system to compute the user’s position, at the cost of inheriting the errors that come with APs positions.…”

Using Perspective-n-Point Algorithms for a Local Positioning System Based on LEDs and a QADA Receiver

“…In this context, an indoor local positioning system (LPS) is often required and designed to estimate the pose of people or devices in environments where GNSS solutions have attenuation/coverage problems and multi-path effects that limit their performance [2]. For that purpose, different sensory technologies have also been applied in previous work (acoustic [3,4], infrared [5,6], radio-frequency [7,8], etc.). In particular, the use of optical signals (infrared or visible light) [9] has spread significantly due to their low cost, long lifetime and their presence in most of today's infrastructures [10].…”