Doctoral theses
Current Record: 90 of 121
|
|
| Identifier |
000355146 |
| Title |
Wireless network device position location system |
| Alternative Title |
Σύστημα εντοπισμού θέσης συσκευών ασύρματων δικτύων |
|
Author
|
Παπαδάκης, Στέφανος Εμμανουήλ
|
|
Thesis advisor
|
Τραγανίτης, Απόστολος
|
| Abstract |
Location Based Services (LBS) are continuously gaining popularity and enabling a new area of innovative applications. The basic functionality needed in order to provide LBS is the position location of the user. Many localization systems and techniques are already available, with the GPS being the most widespread. Since GPS is based on a network of satellites it cannot cover indoor environments and many of the positioning systems used indoors require proprietary hardware and software.
The objective of the dissertation is to investigate analytically and verify experimentally through simulations and field trials a novel approach for the development of a positioning system. We consider IEEE 802.11-based wireless network devices and infrastructure in our positioning system, thus enabling the location positioning for the majority of contemporary mobile computing devices. Most of the localization solutions using 802.11 are based on the creation of signal strength maps, a time consuming and labor intensive process. We are proposing and evaluate a novel position location technique, the Signal Strength Difference on Arrival (SSDoA). We use the differences of the received signal strength in three or more observation points that are in known positions to generate a set of equations that provide an estimation of the transmitter's position. These equations define the hyperbolic curves where the transmitter may lie. In parallel the use of fixed and known observation points assists the calculation of the radio-environment characteristic, the radio propagation path loss exponent. The knowledge of the propagation characteristics substitutes the need of signal strength mapping.
The primary contribution of our work is that the developed system does not require any time consuming calibration or specific/proprietary hardware. Also we achieved this without requiring any involvement from the end user, as long as they are actually using their device to transmit any data. This became feasible by introducing the signal strength difference (SSD) to the distance difference relation. Overall this work has produced a novel and flexible location positioning system which enables any IEEE 802.11 client device to be accurately and promptly located.
In order to have a proof of concept we used off-the-self Cisco access points that are part of our building's wireless infrastructure and created a tracking application in Java programming language. The Cisco access points have a special monitor mode where all the wireless packets are captured and the value of the received signal strength is recorded. The application is communicating with the access points, that are essentially the observation points, that push the captured data to the application. The captured data have the signal strength information per packet and the application implements a special packet decoder for the decoding of their information. Then the signal strength information is processed and with the use of SSDoA the position estimate is obtained.
Using simulations we have investigated in detail all the factors that influence the accuracy. These factors are the granularity (one dB steps) in the signal strength measurement, the value of the path loss exponent, the hyperbolic solution method, the initial guess point for the hyperbolic solution method, the sliding time window size, the number and geometry of the observation points. The results revealed that the most important factors are the observation points geometry and their number. Also the use of a tracking filter is essential for the rejection of the deviating estimates. Based on these results we proposed guidelines for the successful implementation of our system. The results of the simulations for our proposed system are quite promising, with an average accuracy of about 2 meters under harsh conditions. Furthermore we conducted some preliminary experiments with the use of our Java application in an existing 802.11 infrastructure with very encouraging results. It is clear that this work could provide, with little effort, a complete working product comparable to the most advanced commercial ones.
|
| Language |
English |
| Issue date |
2009-10-30 |
| Collection
|
School/Department--School of Sciences and Engineering--Department of Computer Science--Doctoral theses
|
|
|
Type of Work--Doctoral theses
|
| Permanent Link |
https://elocus.lib.uoc.gr//dlib/2/5/5/metadata-dlib-b403b98150537167a1ec59e545c7e493_1275633966.tkl
|
| Views |
632 |
Collections
