Improving cooperative trajectory mapping applications with encounter-based error correction

Abstract

Cooperative trajectory mapping is an emerging technique that allows users to create such a map by using data that are collected from each other's mobile phones. This technique has been proposed for many applications, such as people localisation, public transportation tracking and traffic monitoring. Unlike the traditional Global Positioning System, cooperative trajectory mapping only requires information about the departure distance and moving direction from the previously reported position. This avoids problems such as the high-energy consumption of GPS and weak GPS signals in indoor conditions. However, the new technique also brings about other problems, such as measurement errors in cooperative trajectory mapping, which is when a measurement error causes the spatial relations among users to be wrong. We propose an encounter-based error correction algorithm to efficiently reduce measurement errors. Extensive simulation experiments are carried out to validate our solutions.

Keywords::

• cooperative trajectory mapping
• encounter
• measurement error
• mobile phones
• relative error

Acknowledgement

This research was supported in part by NSF Grants ECCS 1128209, CNS 1065444, CCF 1028167, CNS 0948184, CCF 0830289 and REU 1156574.

Notes

^1. Email: [email protected]

^2. Email: [email protected]

^3. As mentioned by [6], although there are several Bluetooth-based neighbour discovery techniques [20], the low detecting rate of short-lived encounters is a problem. In this paper, we assume that all encounters can be detected.