Wireless distance estimation with low-power standard components in wireless sensor nodes

TL;DR

This paper evaluates the use of standard low-power wireless sensor node components, specifically the CC430 transceiver, for accurate distance estimation using time of flight measurements, avoiding the limitations of signal strength methods.

Contribution

It demonstrates the feasibility of using existing standard hardware for time of flight-based distance estimation in wireless sensor nodes, with detailed analysis of errors and optimal settings.

Findings

Time of flight measurement accuracy depends on various parameters.

Standard CC430 hardware can be effectively used for distance estimation.

Optimal settings vary with environmental conditions.

Abstract

In the context of increasing use of moving wireless sensor nodes the interest in localizing these nodes in their application environment is strongly rising. For many applications, it is necessary to know the exact position of the nodes in two- or three-dimensional space. Commonly used nodes use state-of-the-art transceivers like the CC430 from Texas Instruments with integrated signal strength measurement for this purpose. This has the disadvantage, that the signal strength measurement is strongly dependent on the orientation of the node through the antennas inhomogeneous radiation pattern as well as it has a small accuracy on long ranges. Also, the nodes overall attenuation and output power has to be calibrated and interference and multipath effects appear in closed environments. Another possibility to trilaterate the position of a sensor node is the time of flight measurement. This has the advantage, that the position can also be estimated on long ranges, where signal strength methods give only poor accuracy. In this paper we present an investigation of the suitability of the state-of-the-art transceiver CC430 for a system based on time of flight methods and give an overview of the optimal settings under various circumstances for the in-field application. For this investigation, the systematic and statistical errors in the time of flight measurements with the CC430 have been investigated under a multitude of parameters. Our basic system does not use any additional components but only the given standard hardware, which can be found on the Texas Instruments evaluation board for a CC430. Thus, it can be implemented on already existent sensor node networks by a simple software upgrade.