Optimizing placement of nodes in different environments for locator broadband sensor networks

Master's Thesis in Electrical Engineering (System Communication)

Abstract

Optimization of Node Placement in Different Environments for Ultra-Broadband Locator Sensor Networks

locator systems are systems that can detect the location of an object in a space. Such systems usually detect the distance between two objects by using radio communication and determine the location of the object in the area after several measurements from different nodes. Among the most important global locator systems, we can mention global locator systems [1] and Russian military locator systems [2], European citizen locator system [3].

According to the need for high-precision locators as well as positioning in closed spaces, the use of these systems has become less important and systems with more accurate signals have become more important. Among these systems is the use of ultra-wideband signals, which have very narrow pulses and wide bandwidth and can have good time resolution.

Therefore, these systems are proposed as a suitable proposal for precise location of closed spaces.

All location systems have reference nodes, each of which has a specific distance to the receiving object, and the location can be determined by measuring these distances. found the nodes. One of the most important proven factors in the accuracy of location systems is the location of reference nodes. In recent decades, the importance of this issue has been investigated and suitable solutions have been proposed for it. For example, it is effective in reducing the number of sensors where the position of the target is most of the time.

If the probability distribution function of the target in coordinates (x,y) can be assumed to be known, then this function can be used and placement can be done in such a way that the average error is the desired value. In order to reach the target distribution function, the assumption that the target location process is orgadic in time can be used and attributed to its probability distribution in a way. Therefore, by studying the conducted research and deriving the criteria for comparison and accuracy of proper distance finding using broadband signals, it is hoped that by reducing the number of reference nodes and using the behavioral characteristics of the target in the network, proper accuracy can be achieved. Such nodes must have a specific location so that they can estimate the location of the target node. In this text, assuming that the reference nodes have a known location wherever they are placed, these nodes are tried to be placed in a place where they can estimate the following two conditions:

First: they can estimate the location of the target node in the whole area under the watcher.

Second: the average error of estimating the location of the target node is also minimal in the entire area.

Among other factors which can be effective in positioning accuracy is the location of reference nodes in the monitored environment. But the lack of which is evident in most of the researches is that if the reference node is mobile, a suitable placement for the reference nodes is not suggested. In this research, in addition to providing a suitable solution to optimize the location of reference nodes, it has been used in several environments and its results have been observed. Finally, a comparison between the random distribution of reference nodes like other researches and the proposed method has been made, and it has been determined that a smaller number of reference nodes can be used to achieve the desired accuracy by using this method. It is to be able to optimally obtain the location of reference nodes by using broadband location systems to achieve the desired accuracy.

Introduction

Today, the use of sensor networks has become very common and practical.. Sensor networks are collections of sensor devices that are placed in an environment with a special arrangement and pursue a specific goal by trying to cover the entire environment. The goal in sensor networks may be to sense temperature for certain environments, to sense smoke to prevent fire, or to sense a certain type of gas. But one of the most important parameters that can be detected by sensors is location and time, which is very useful. Due to the increase in positioning efficiency in various applications, the need for such systems is increasing day by day. The expansion of such systems has attracted the attention of researchers and manufacturing companies. For example, the location of cars in controlled and military areas, and the monitoring of traffic in different places and the control of passenger fleets are among the recent applications of positioning that use GPS positioning systems. is not Because these systems have satellite communication and need a direct line of sight with the receiver, and the receiver must be in contact with four satellite reference nodes at the same time. Such limitations make it almost impossible to use these systems indoors. In addition, these systems require high transmission power in the locator module to exchange information with reference nodes [1], and the use of these modules using battery power is difficult. One of the other important problems of these systems is their relatively low accuracy of several meters [1] .

In more precise applications such as mobile robots[2], soccer player robots and location of people in closed security spaces of large hospitals, it is not possible to use GPS systems and systems made of sensor networks are used. Therefore, local wireless networks[3] are used for indoor spaces. But these systems are also not efficient in terms of accuracy. Broadband signaling is recommended in the table below for the required accuracy [2]. It achieved this by using ultra-wideband signals.The basis of ultra-wideband communication is the transmission of information by very narrow pulses (in the time domain) that have a low energy level. According to the rules established by the American Radio Regulatory Organization [1], broadband telecommunication systems can work in the frequency band of 3.1 to 10.6 GHz (of course, this frequency range is specific to the United States and in Europe the lowest frequency for broadband telecommunication systems is 2.4 GHz). In order to be called a broadband signal, it must have at least one of the following properties:

It has a bandwidth of at least 500 MHz.

The following relationship applies to the broadband signal.

(2-1)

in which the bandwidth, central frequency, minimum and maximum frequency of the transmitted signal.

The use of this technology in the fields of wireless communication is increasing day by day. To prevent interference with existing telecommunication systems, the US Radio Regulatory Authority has specified the allowable effective isotropic radiated power [2] (EIRP) for each frequency band. Considering that this band is very wide, these bandwidths are dedicated for applications such as WSN and others. Therefore, the use of the band has power limitations for ultra-broadband applications. Of course, the regulatory center has been stricter in the bands in this range that are reserved for other specific uses and purchased by the equipment manufacturing companies than the ultra-broadband users.