Radio Tomographic Imaging using Extremely Resource Constrained Devices

Abstract

This research is aimed to be a proof of concept for a Radio Tomographic Imag- ing(RTI) system designed using extremely resource constrained devices. Unlike any other Radio Tomographic Imaging systems implemented using transceivers having special attributes such as Received Signal Strength Indicator (RSSI), Ul- tra Wideband(UWB) Receiver's which will cost more and consume more power, this method will explore the possibility of adopting a feature that is available even in low end transceivers, that is - packet loss ratio, as a probing mechanism for detecting obstacles. In this proof of concept, nRF24L01 modules were used as transceivers, and Ar- duino UNO modules were used as controller nodes which provide the processing power required for the system. The usage of slave nodes possessing less inbuilt processing capability which can wirelessly communicate with controller nodes not only makes this Radio Tomographic Imaging implementation more cost and en- ergy e ective, but also highly portable. As low cost, low energy consuming hardware is used in an outdoor application, the statistical likelihood of packet loss becomes greater when there is an obstacle in front of the device. Also, as the resolution of the system depends on the dimen- sions and the topology of the RTI mesh, adaptation of advanced mathematical methods comprising of several noise ltration techniques became paramount for building a reliable system. Despite the complexity of some of the subroutines used, the overall simpli ed nature of the mathematical models and algorithms used in the implementation also means the system is scalable to any number of sensors and/or controller nodes with ease. This model also allows exibility to easily implement the same methodology for probing obstacles in a space with di erent geometric con gura- tion than the simple square shaped con guration used for this proof of concept. The low power consumption also creates a limitation of range which can easily be o set by the cost advantage o ered by the system.