Positioning system for wireless sensor networks with location fingerprinting

Cao, D. and Goh, H. and Kwong, K.H. and Michie, C. and Andonovic, I. (2009) Positioning system for wireless sensor networks with location fingerprinting. In: Progress In Electromagnetics Research Symposium 2009, 2009-08-18 - 2009-08-21.

[thumbnail of Cao-etal-2009-Positioning-system-for-wireless-sensor-networks-with-location]
Text. Filename: Cao_etal_2009_Positioning_system_for_wireless_sensor_networks_with_location.pdf
Accepted Author Manuscript

Download (604kB)| Preview


Wireless sensor networks (WSN) are networks that deploy hundreds or thousands of wireless sensors in a pre-defined area that can communicate with each other to detect, for example the ambient environment. Each sensor is composed of the four basic elements: transmitting unit, processing unit, power unit and sensing unit. The main task of each sensor is to detect events, perform a restricted set of local data processing tasks and then transmit the data. This technology still in its early stage new researches are being conducted intensively in MAC protocols, network and routing layer, and adaptation into various domains applications. In this proposal, the focus is placed to investigate algorithms in mapping the location of sensor nodes. Knowing the location of the sensor node is critically important; the knowledge of the location of the sensor node that reported a detected event can reduce the time for assistants reaching to the outbreak point. This can potentially save life or can bring the outbreak event under control in shortest time. As the sensor node's physical hardware is mainly comprises of low specification and low cost componentry to facilitate mass production hence affordable to be applied intensively in monitoring zone. This has created a tough challenge in mapping the locations of sensor nodes as the hard-ware can not provide precise timing in calculating time of flight of a packet which is an important parameter in estimating distance between transmitting node and receiving node. In general the sensor node is only equipped with a single antenna which has also rule out the possibility of using techniques rely on angle of arrival packet. Therefore, the research is limited to use the received signal strength as the main source in estimating the travelling distance for the received packet. This paper investigates positioning algorithms that based on received signal strength i.e., location fingerprinting. In positioning systems, location fingerprinting is also referred as pattern matching of radio signature. The advantages of using RF fingerprinting are it does not require any hardware modifications to the sensor node and in comparison to other algorithms it is immune environmental influences that caused signal attenuation such as multipath, fading, reflection, non line of sight, and etc. This paper focuses on challenges that relate specifically to the location mapping of wireless sensor node including radio propagation of low specification WSN hardware, accuracy, operational range and impact of environmental factors. The optimized positioning system for WSN is documented, and results gained from experiment based on IEEE 802.15.4 WSN platform is provided.