In this chapter the authors discuss the feasibility of sensor node localization by exploiting the inherent resources of WSN technology, such as the received signal strength (RSS) of the exchanged messages. The authors also present a brief overview of various factors influencing the RSS, including the RF-signal propagation and other topology parameters which influence the localization process and accuracy. Moreover, the RSS variability due to internal factors, related to the hardware implementation of a sensor node, is investigated in order to be considered in simulations of RSS-based outdoor localization scenarios. Localization considerations referring to techniques, topology parameters and factors influencing the localization accuracy are combined in simulation examples to evaluate their significance concerning target positioning performance. Finally, the RF propagation model and the topology parameters being identified are validated in real outdoor localization scenario.
TopIntroduction
The most distinctive features of Wireless Sensor Networks (WSNs), in relation to the other wireless networks, are the large number of the sensor nodes involved (often in the range of hundreds and thousands), their small size and relatively lower price. Furthermore, due to the battery-based power supply of the sensor nodes, there are critical constraints for their operation and energy consumption, which impose the requirement for optimizing the use of their hardware modules and resources.
The spatial localization of the sensor nodes is a fundamental problem in the area of WSNs. An attractive way for estimating the location of a wireless object is by using the Received Signal Strength (RSS) attenuation with the distance, as this does not require any additional hardware. This is possible due to the fact that in most sensor node radios, the Received Signal Strength Indicator (RSSI) is a standard feature, which allows the RSS value to be obtained automatically together with the received message.
As we consider outdoor environments, spatial localization via Geo-Positioning System (GPS) is also feasible option. The main disadvantages of the GPS-based localization methods are the incurred cost, the size and the energy consumption of sensor nodes equipped with GPS receiver, the limited accuracy of localization in some locations and environments, as well as the dependence on services controlled by foreign governments. In particular, equipping every WSN node with a GPS receiver increases both the size and cost of the nodes and in consequence, the overall application cost. Furthermore, adding extra hardware increases the energy consumption of the sensor nodes, which in the context of WSNs shortens the operational life-time of the network. In certain potential areas of interest, the localization accuracy offered by geo-positioning systems could be insufficient due to local government restrictions.
On the other hand, the RSS is not bound by these restrictions, but it is known for being noisy, unstable, variable and difficult to use in practice. To achieve better understanding of the nature of these difficulties and limitations, and to identify the range of applicability of the RSS in localization scenarios, a thorough study on the RSS behavior and its dependence on various factors is essential. In connection to this, the following problems are considered of primary importance:
- •
The RSS behavior in outdoor environment has not been researched thoroughly at medium distances, when a variety of factors change, resulting in poor understanding of the limitations and the feasibility of performing RSS-based localization. For that reason, a detailed analysis of the RSS behavior under various conditions is required.
- •
The log-normal path loss model (LPLM), which is the most frequently used propagation model for RSS-based localization, is rather general and does not take into account some important factors that influence the RSS, such as the height of the nodes from the ground. Therefore, alternative propagation models need to be considered and evaluated experimentally.
- •
The existing deployment algorithms provide solutions for the coverage and connectivity problems, with respect to the number of neighbor nodes, without taking into consideration the factors influencing the communication among the network nodes. Most assumptions and simplifications being made towards such solution contribute to the common belief that the communication between two sensor nodes is good until a certain distance and worsens after it, which is not always correct. A communication link should be considered as guaranteed when the RSS is sufficient, so that the sent packets will be received with packet loss rate below a certain threshold.
- •
In most of the published research on RSS-based localization, the reported results are based more on simulations rather than on real experimentations. Authors commonly assume that the variability of RSS is more or less a statistical quantity, instead of measuring and modeling it. To contribute to the better understanding of the RSS variability, a detailed experimental study is essential.
Dealing with these problems is of primary importance for the successful use of the RSS as an estimator of distance in applications requiring network nodes localization.