Location-Aware Environmental Monitoring For Mobile Workers

Table Of Contents

1 Introduction
1.1 Location Aware Environmental Monitoring and GIS Application
1.2 Aspects of “Mobile Human Computer Interaction”
1.3 Implementation Supporting Mobile Work
1.4 Structure of the Thesis

2 Positioning Technologies and Applications
2.1 Overview Positioning Techniques
2.2 Cell of Origin (COO)
2.3 Time of Arrival (TOA), Time Difference of Arrival (TDOA)
2.4 Angle of Arrival (AOA)
2.5 Satellite Based Positioning Techniques
2.6 Global Positioning System
2.7 Technical Background
2.8 Measurement of Pseudo Ranges
2.9 Differential GPS (DGPS)
2.10 Satellite Based Augmentation Systems (SBAS)
2.11 Global Navigation Satellite System
2.12 Combination of GPS and GLONASS
2.13 NMEA0183 Standard
2.14 Geographic Information System Overview
2.15 GRASS
2.16 ESRI GIS

3 Related Work: GIS Projects and Location-Aware Implementations
3.1 Accuracy in Real-World Positioning
3.2 GISPAD Implementation Overview
3.3 FOGEPOS Implementation Overview
3.4 Comparison and Relation to Our Approach

4 Project Background: Ecological Field Studies
4.1 Institutional Background
4.2 Umweltbundesamt’s Integrated Monitoring Observation Work
4.3 Monitored Processes and Recorded Parameters
4.4 Data Collection Process and Data Architecture
4.5 MORIS Database

5 Prototype Concept and Overview
5.1 View and Purpose of the Location-Aware Prototype
5.2 Challenges and Functions of the Location-Aware Prototype
5.3 ESRI ArcPad Design Decisions
5.4 ESRI Map and Layer Definitions
5.5 Analysis of Requirements for Implementation
5.6 Final Prototype GIS Functionalities
5.7 Capturing Tree Monitoring Data
5.8 Capturing Permanent Plot Data
5.9 GIS Logging Functionality

6 ESRI ArcPad Object Model Prototyping
6.1 Introduction of ESRI ArcPad Prototyping
6.2 ArcPad Customization Overview
6.3 Development Environment for ArcPad
6.4 Customizing ArcPad - ESRI XML Files (ArcXML)
6.5 Own Prototyping Work and Prototype’s Customization
6.6 Configuration Files and Involved Code
6.7 Selected Prototype’s Functions and Code Fragments
6.8 Prototype Installation Details

7 Evaluation and Field Test
7.1 Field Test’s Area Overview
7.2 Equipment Used for the Field Test
7.3 Qualitative GPS Results
7.4 GPS Track Log Data and Projection Transformation
7.5 Statistical Analysis of Gathered Positioning Values
7.6 Evaluating Usability and Interface Design

8 Conclusion and Outlook

9 Bibliography

10 Appendix
10.1 UN-ECE Integrated Monitoring Zöbelboden [www.umweltbundesamt.at]
10.2 MORIS Specification (Chart Version) [www.umweltbundesamt.at]
10.3 ESRI Object Model [www.esri.com]
10.4 ESRI Shapefile Specifications (Shortened Version) [www.esri.com]
10.5 SiRFstarII GSC2x Technical Documentation [www.sirf.com]
10.6 Gauss-Krüger Projection for Austria [www.statistik.at]

List of Tables

Table 1: Max. Predictable Positioning Errors (Probability 95%) [Roth05].

Table 2: GPS – Measurement Error Sources [Park+96].

Table 3: Combined Typical Positioning Accuracy [Woes+05] [Park+96].

Table 4: NMEA0183 Sentence: $GPGGA.

Table 5: List of Field Protocols used at Zöbelboden.

Table 6: ArcPad Customization Overview.

Table 7: ESRI Framework Object Overview.

Table 8: Handheld Specification.

Table 9: GPS Receiver’s Specification (SysOnChip Inc).

Table 10: GPS Analysis - Nr of Satellites (martin to gerda/karin, values in m).

Table 11: GPS Analysis of Device karin's Values (in m).

Table 12: GPS Analysis – karin’s and martin’s Values Extremum Filtered.

Table 13: GPS Analysis – Karin’s and Gerda’s Values Extremum Filtered.

Table 14: GPS Analysis – Criteria Comparison and Combination.

List of FIgures

Figure 1: Zöbelboden Monitoring and Evaluation Area [Humm+06].

Figure 2: Coordinate Systems [Roth05].

Figure 3: Schematic GPS Satellite Constellation [Aero03].

Figure 4: GPS – Determination of Position [Roth05].

Figure 5: SBAS (Red-WAAS, Green-EGNOS, Blue-MSAS) [www.esa.int].

Figure 6: GIS - Basic Functionality Classification.

Figure 7: ESRI Component View [www.esri.com].

Figure 8: GISPAD Scenario and GUI (www.forstgis.eu).

Figure 9: GISPAD Ruhrgas Scenario (www.ascos.de).

Figure 10: FOGEPOS Scenario (www.ascos.de).

Figure 11: Zöbelboden Area (www.umweltbundesamt.at).

Figure 12: Design of a Permanent Plot.

Figure 13: Zöbelboden Observation Overview.

Figure 14: MORIS Database User Interface Example[Schen+05].

Figure 15: Data Structure of Collected IM Parameters.

Figure 16: MORIS Data Model Overview.

Figure 17: Integrated Monitoring – Prototype System Architecture.

Figure 18: GIS Layer Definition [www.esri.com].

Figure 19: Prototype Interface and Single Toolbar.

Figure 20: Zooming In and Out.

Figure 21: Create Tree Objects.

Figure 22: Selection of Objects.

Figure 23: Removing Objects.

Figure 24: ArcPad - Entering / Altering Tree Data.

Figure 25: ArcPad - Entering / Altering Permanent Plot Data.

Figure 26: ArcPad Customization - Involved Files and Tools.

Figure 27: ESRI Object Model Overview [Arcb02].

Figure 28: Snippet - Customized ArcPad.apm.

Figure 29: Snippet - Customized ArcPad.apm.

Figure 30: Snippet - Customized ArcPad.apx.

Figure 31: Algorithm – Selecting Trees and POIs.

Figure 32: Algorithm – Adding Tree Objects.

Figure 33: Algorithm - Adding Multimedia Files.

Figure 34: Mobile Device Used for Field Test.

Figure 35: GPS Receiver SiRF II+ Used for Field Test.

Figure 36: 3D View of Test Area with Spatial Marked Waypoints.

Figure 37: Closer Look of Field Test’s Main Outlier Sections.

Figure 38: Gauss-Krüger Projection [www.statistik.at].

Figure 39: Aggregated 2D View from Device Logs.

Figure 40: Distance of Investigated and Estimated Plots based on GPS values [Humm+06].

Figure 41: Detailed Map IM Area.

Figure 42: MORIS Chart.

Figure 43: Object Model Framework.

Figure 44: Shapefile Specs (a).

Figure 45: Shapefile Specs (b).

Figure 46: SiRF Tech Specs (a).

Figure 47: SiRF Tech Specs (b).

Figure 48: GK M28 / M31 / M34.

1 Introduction

The motivation of this master’s thesis is to investigate the usefulness of mobile computing for researchers in the field. Especially we focused on two main aspects, positioning accuracy and the overall usability in a field environment. The project was initiated as a part of the “Austrian GRID” (AGRID, www.austriangrid.at) initiative within a dedicated mobility work package (WP M-7b: Performance Engineering for Mobility in the GRID). The AGRID initiative is based on a national consortium in Austria to set up and to support grid computing in general and to provide coordination and collaboration between computer scientists for GRID technologies and developers of scientific computing applications. The aim of the consortium is to support and enhance advanced computing technologies in cooperation with well-recognized partners in grid-dependant application areas.

In the context of this AGRID project, this thesis is related to a close cooperation between the University of Vienna “Institute of Distributed and Multimedia Systems” and the “Federal Environment Agency of Austria” (Umweltbundesamt) with focus on a long term ecosystem monitoring program. The involved area of the monitoring activity shown in Figure 1 is called Zöbelboden, which is an environmental observation area in the national park “Northern Calcareous Alps” situated in Upper Austria. This area is part of the UN ECE’s “Integrated Monitoring Program” and was set up to provide long-term data collection of pollution and ecological data.

The main objective of the joint AGRID subproject (from now on simply referred to as project) is to support and assist mobile workers in the case of collecting ecological data in extreme natural environments. Sarah Nusser [Nuss+03], [Nuss+04] conducts that especially mobile systems may be too complex and circumstantial during scientific field usage when researchers are interacting with objects or processing different tasks. In our current project we want to investigate if it is possible to provide a usable mobile client for this specific need, thereby deriving a scientific user’s requirements, performing field tests, and investigating GPS accuracy in the field.

Currently, the collection of field data at the Umweltbundesamt is processed without support of mobile technology using paper filed protocols entered to the database afterwards, which results in a lot of paperwork and therefore costs a lot of time and resources. We expect that using a mobile device in the field provides high potential to save time and costs.

Abbildung in dieser Leseprobe nicht enthalten

Figure 1: Zöbelboden Monitoring and Evaluation Area [Humm+06].

To exploit mobile technologies in a real world scientific application, we design and implement a useful und usable location-aware mobile client for immediate and user friendly data collection in field situations with provided opportunity of direct data integration into Umweltbundesamt’s databases and therefore in a grid enabled and grid supported infrastructure.

The prototype is based on ESRI’s Object Model provided by ArcPad Studio Application Builder [Arcb02] and ArcPad mobile GIS software [Arcp04] ensuring direct integration of Umweltbundesamt’s own ESRI ArcGIS [Arcg06] data layers [ArSh98] of the monitoring area at the Zöbelboden plateau.

1.1 Location Aware Environmental Monitoring and GIS Application

The prototype should allow complete data capturing functionality using standard PocketPC or handheld devices with a continuous calculation of accurate positioning values, as already mentioned within Hasegawa’s study on GPS accuracy in forest conditions [Hase+07]. With this we defined the following initial requirements: As already mentioned by Anand et al. [Anan+06], it should be possible to provide a usable mobile Geographic Information System (GIS) and data collection prototype solving positioning interferences and signal problems at extreme geographic locations. All digitally collected data should be easily transferred back to an ESRI database like ESRI ArcGIS or ArcGIS Server [Arcs07] and as a next step to the MORIS database referring to Schenz et al. [Schen+05] and Peterseil et al. [Pete+05], which represents a specific ontology based Oracle database later discussed in this document.

In our case we focus on a single-user mobile environment containing a specific observation area’s data set. For this specific area, we applied the configuration from a distributed system involving a separated data structure (MORIS database) and GIS server structure (ESRI ArcGIS Server). An interesting aspect of the prototype development is the data collection process of various parameters of defined permanent plots and trees situated at the monitoring site, allowing to capture additional data like pictures and other multimedia data and to link it with monitored objects and points of interest.

The usability of this prototype will be determined by interrogating the target users and the gained accuracy of GPS traces will be investigated by collecting geospatial data, in terms of changes of defined permanent observation areas (termed permanent plots) and monitored objects, like trees, at the monitoring site. These investigations represent a feasibility study to decide, if future development of mobile data collection clients is reasonable for this kind of observation work at extreme geographic locations. We summarize the various criteria for this study by the ease of use in free territory, the quality of captured environmental and scientific data and, of course, its achieved quality of positioning values in this kind of environment.

1.2 Aspects of “Mobile Human Computer Interaction”

It is intended to produce a mobile prototype which can be easily used in outdoor situations, especially with a strong focus on mountainous territory. Location-aware mobile clients can be single-user and multi-user systems as mentioned by Ashbrook et al. [Ashb+03] or as referenced by Pham and Wong [Pham+04] such systems can be designed within a multi component system with different structures depending on where data is stored, which server environment handles data transfer and how spatial data and gathered information is processed with a defined client structure. In this project we design a loose coupled single-user mobile client, offering manual data integration capabilities to UBA’s existing data structure and GIS server infrastructure.

Focusing on usability aspects Jason Pascoe et al. [Pasc+00] defined Human Computer Interaction (HCI) in fieldwork environments by four characteristics:

- Dynamic user configuration
- Limited attention capacity
- High-speed interaction
- Context dependency

To consider these aspects within this work, we strongly concentrate on limited attention capacity, high speed interaction and context dependency, because these are the main aspects of an intuitive and convenient use in free field. It is important, that the usage of the prototype should not bother the mobile worker in extreme field situations.

Following initial challenges were defined for the proposed prototype:

- A clearly laid out, easy to use and functional arranged application structure
- A fully functional graphical user interface specially designed for Umweltbundesamt’s needs
- A continuous display of geographical maps and important spatial objects
- The ability to include multimedia-based data (pictures, audio, videos, etc.)
- An interface for integration of collected scientific data into UBA’s ESRI Database and integration into the MORIS database.

1.3 Implementation Supporting Mobile Work

In the following chapters of this document a prototype will be introduced which allows capturing the desired ecological data digitally using a PocketPC aka Personal Digital Assistant (PDA). An easy-to-use application based on ESRI’s ArcPad Application Developer is been developed for use on GPS capable Pocket PCs with Windows Mobile OS which integrates the Umweltbundesamt’s own ESRI data layers. Thus, the mobile device is prepared for data integration into a data GRID infrastructure with an UBA customized database interface for MORIS. This new prototype is a GPS-assisted client, which enables location-awareness and further allows mobile users to augment the collected data with comments, digital photos, digital audio or even small movies of artifacts, like trees and other plants.

A simple algorithm for determining nearest situated points of interest (POI’s), in our case permanent plots, is implemented for use in our prototype depending on the current GPS data received. It calculates the distances of near situated spatial objects and shows it to the mobile user. This functionality is supposed to improve the usage of the mobile client and user’s interaction in future by advancing the overall monitoring process. It is further expected that better positioning hardware and extensions like a compass device or a rangefinder component will also help to improve accurate detection and identification of scientific points of interest for our use case.

1.4 Structure of the Thesis

After a first introduction in chapter 1, we introduce fundamental positioning technologies and related applications in chapter 2 to give the background of knowledge needed for this thesis. As we used the stable and already widespread GPS positioning technique within our project, we are going to discuss the GPS technology in detail as well as basics of other positioning techniques in this area. Chapter 3 focuses on related work and already implemented GIS projects in order to show possibilities and approaches in this sector to provide an overview of different approaches in the same thematic area.

Chapter 4 concentrates on our project with the Umweltbundesamt and the involved information needed within ecological field studies. In the following chapter 5 we are going to introduce the prototype’s concept and an overview of its desired functionality. Based on the requirements defined, chapter 6 informs about our concrete ArcPad prototyping work and software based background of the mobile prototype. It shows the techniques and methods used to implement and customize our prototype for Umweltbundesamt’s needs.

Within chapter 7 we discuss the used technical equipment and mobile devices, the evaluation of the prototype within the field tests and gathered statistical information of the positioning quality. As outcome of this analysis we define metrics for positioning quality by analyzing positioning values gathered by different devices.

Finally, in chapter 8, we show our conclusions and possible outlooks for future work in the area of mobile positioning and location-aware applications.

2 Positioning Technologies and Applications

Accurate positioning is very important for the present work. Thus, here the fundamentals for positioning techniques are explained. As there are many different hardware based positioning techniques, systems and services available, we focus in the following part of this chapter on a short overview of fundamental positioning techniques and in the second part we concentrate on satellite based positioning techniques and specific enhancements relevant to our project.

Further there are many software based Geographic Information System (GIS) solutions and software packages available for server applications, desktop use, or mobile use especially for field worker. These systems basically enable analysis and handling of geospatial information. Hence we introduce in the third part of this chapter important GIS software packages. There is a wide range from commercial software packages to open source distributions, providing spatial and geographical services in conjunction with hardware based localization techniques.

2.1 Overview Positioning Techniques

Positioning techniques, systems and services provide geospatial information and therefore location-awareness to the end-user. Many applications can be enhanced by positioning systems offering various geospatial problem solving solutions either in personal, scientific, business or industrial use conditions. There are many different forms of location-based techniques available like cell based, infrared based, wireless LAN based or even satellite based localization systems. Each of these location-based services utilizes different base technologies (referring to Jürgen Roth [Roth05]). We want to describe three primary technologies for positioning which are Cell of Origin (COO), Time of Arrival (TOA) which also incorporates Time Difference of Arrival (TDOA), and as a last key technology Angle of Arrival (AOA).

2.2 Cell of Origin (COO)

This technique is usable within positioning systems based on a cell structure. Wireless transmission and telecommunication systems are often restricted to a limited range of coverage, which allows to assign a mobile terminal to a cell according to the area the user currently resides in. Therefore identification and measurement of position is derived from the cell information given by the system (as mostly used within GSM or UMTS networks).

This technique is often combined with the measurement of signal strength. Simply this method of calculating the position with signal strength needs sender devices due to known positions. As often used within wireless local area networks (WLAN), some access points exactly known are measuring the various signal strengths to a client and then approximating the position of the device.

2.3 Time of Arrival (TOA), Time Difference of Arrival (TDOA)

TOA and TDOA basically depend on the delay of electromagnetic waves and therefore on the speed of light which is about 300.000 km per second. Modern measurement devices enable to exactly quantify run-time periods of radiowaves and accordingly it is possible to compute nearly exact distances and the position of sender and receiver (as used within satellite based positioning techniques). When using this technology in connection with GSM or UMTS networks, it is also called Enhanced Observed Time Difference (EOTD).

2.4 Angle of Arrival (AOA)

When calculating positions by measuring angles we need a point-to-point or beam radio link system to determine the direction of the signal source. When receiving this targeted radio signal, it is possible to measure angles. This is done by measuring spatial reference objects, by combining measured angels it is possible to calculate the angular separation and therefore to estimate the position of the receiving device. For example antennas with direction characteristics allow measuring the direction of the received signal. This method is usually used by antenna arrays which can simultaneously measure different directions of a signal and, thus, estimate the position of a mobile device.

2.5 Satellite Based Positioning Techniques

In the next sections we focus on satellite based systems and services since they are used in the thesis project. For a better understanding we will now introduce the major technologies in this field, beginning with the wide spread NAVSTAR Global Positioning System (GPS) [Long+81]. Secondly we will pass on with its improved part Differential Global Positioning System (DGPS) corresponding to Beisner et al. [Beis+96] or as a counterpart to Satellite Based Augmentation Services (SBAS), improving the overall positioning quality with additional correction signals. As a third important point we introduce the rarely used Russian positioning system Global Navigation Satellite System (GLONASS), which differs in some points to the GPS system. All mentioned satellite based positioning services grant real-time accuracy in free environments between 10 to 25 meters. With additional data required from separately geodynamic services (e.g., International GPS Service (IGS) for Geodynamics Centre) or additional satellite services (like SBAS) for GPS positioning systems, it is possible to grant real-time accuracy up to few meters and post-processing accuracy within some centimeters regarding to Zarraoa et al. [Zarr+98].

In order to provide correct positioning data within satellite navigation we are going to introduce coordinate systems to reference a geographic point on the earth’s surface in a definite way. There are three main forms of referencing geospatial information displayed on Figure 2 [Roth05].

One of the most used method referencing this kind of information is a geodetic coordinate system using an ellipsoid for modeling the earth’s surface, like the “World Geodetic System 1984” (WGS84) used for GPS positioning (refer to Figure 2 Part 1). Another but rather rarely used method (e.g., used for intermediate calculations) is a geocentric or cartesian coordinate system. This method specifies the position within three axes (refer to Figure 2 Part 2). The most accurate method is the transverse mercator projection.

This projection divides earth’s surface into different segments on a plain map projection which ensures very high accuracy of the projection (refer to Figure 2 Part 3, planar or transverse mercator projection). There are two important coordinate systems to mention within this projection method. On the one hand the Gauss-Krüger geographic coordinate system (GK) which is specifically subdivided and aligned to different geographic maps and on the other hand the Universal Transversal Mercator coordinate system (UTM) which defines a complete projection sight of the whole world.

Abbildung in dieser Leseprobe nicht enthalten

Figure 2: Coordinate Systems [Roth05].

If different coordinate systems are used for projection, we need to point out specific transformations to correctly align spatial objects on maps. The difference between various coordinate systems is commonly referred as a datum shift, which will be dealt with in context to our project in chapter 7.

Though satellite navigation provides the most accurate instant positioning method usable for every outdoor usage on earth locations, we should mention several error sources that influence satellite navigation. These errors are caused and affected by ionospheric propagation, receiver clock errors, multipath receiving problems, tropospheric propagation, ranging errors, multipath mitigation or onboard clock errors.

2.6 Global Positioning System

The Navstar Global Positioning System (GPS) development initially started in the late 1960's with its concept of creating a worldwide localization system [Pace95]. In the 1970's a Joint Program was initiated within the Department of Defense (DoD) and the Defense System Acquisition and Review Council (DSARC) [Long+81]. A schematic view of GPS satellite’s constellation is shown in Figure 3.

Abbildung in dieser Leseprobe nicht enthalten

Figure 3: Schematic GPS Satellite Constellation [Aero03].

In February 1978 the first GPS satellite was launched followed by many more satellite launches [Pace95]. Generally the GPS service needs at least 24 satellites in orbit to support accurate positioning all over the world [Lang+06]. The satellites have to be replaced after a given time of usage in order to keep the system working and alive. It should be mentioned that before the year 2000 the GPS service was consciously downgraded for civil usage, so a horizontal accuracy was given from "no worse than" 100 meters. After this time, due to a presidential directive from the United States’ president, the "Selective Availability" (SA) was turned off and now it is possible to achieve a standard horizontal accuracy of 10 to 25 meters without additional correction signals.

Described more exactly, GPS consists of three basic components. The first component defines a set of satellites, currently 32 operational satellites (as given by 15th of March 2008) which are orbiting in about 20.200 kilometers above the earth's surface. These satellites are transmitting ranging signals (two different microwave frequencies in the radio spectrum) for defined parts of earth's surface. The second component are maintenance stations, a set of ground monitor stations and upload facilities in order to guarantee the permanent function of the GPS service. The third component is the GPS signal receiving unit, divided in civil and military use. Each satellite transmits a unique digital code which has to be exactly timed with an atomic clock [Lang+06] received by the GPS enabled device. This therefore called GPS receiver picks up the signal by an external or internal antenna. This signal then is computed with other code sequences inside the receiver’s logical processing unit. These signal codes are lined up and matched and the receiver determines the travel time from a satellite to its position, this timing is taken for measurement and is converted into distances using the absolute speed of radio waves (which equals the speed of light 300.000 km / second) [Lang+06].

For measuring any location we theoretically need at least three satellite signals, but in real situations we require at least four or more signals for an exact measurement of location including altitude information which is called a three dimensional GPS fix (3D fix). To achieve a suitable positioning quality in forested and mountainous situations four and more satellite signals are needed for localization. The receiver determines its latitude, longitude, and altitude while synchronizing its clock with the GPS time standard [Ray+05]. This method of computing and determining a specific location with the measurement of distances (refer to pseudo ranges in section 2.3.2) with satellite signals is known as TDOA [Muel68]. We have to keep in mind that this is fundamentally different to AOA which involves the measurement of angles and is often mentioned within Location Based Services (LBS) or Mobile Location Services (MLS) on GSM networks, referring to Dao et al. [Dao+02], Zeimpekis et al. [Zeim+03] and Gabber & Wool [Gabb+98]. For example triangulation is also often used to localize mobile phones registered in different cells and giving cell specific information to the user (COO & AOA). Moreover this is additionally used within cheaper Assisted GPS (A-GPS) services provided by cellular networks which combine single channel receivers (they share one channel to receiver more different satellite signals) with cellular network’s cell information in order to guarantee a minimum state of accuracy referring to Agarwal et al. [Agar+02].

GPS is a fundamentally passive receiver based service. GPS devices simply receive satellite signals and do not transmit or bounce signals to any other device or satellite, it is an absolutely anonymous und unlimited (anonymous and unlimited users) system which is also given with the later mentioned Russian GLONASS [GNSS02] service or the future European GALILEO [Boer00] positioning system. The Global Positioning System is a permanent and global positioning system and is used for civil and military purposes. It is controlled by an executive board of the U.S. government (joint civil and military) and is maintained by the U.S. Air Force on behalf of all users.

2.7 Technical Background

To mention technical properties for GPS receivers we start with Standard Positioning Service (SPS), in fact they access the civil C/A-code (Coarse/Acquisition Code) transmitted on the L1 frequency at 1575.42 MHz. Military receivers use the Precise Positioning Service (PPS), a separately and encrypted P(Y)-Code (Precise or Precision Code, Y stands for the encrypted P-Code) which is transmitted on both the L1 frequency with 1575,42 MHz and the additional L2 frequency at 1227.60 MHz. The additional L2 frequency channel allows receiving units correcting atmospheric and other propagation errors in real-time which offer more stable positioning. With this enhancement military GPS receivers work with more accurate measurements offering a higher reliability shown within the DoD defined accuracy of GPS as given in Table 1. P-Code receivers are designed to achieve localization accuracy with a maximum predictable error of 22 meters, while civilian C/A-Code receivers are designed for an accuracy with a maximum predictable error of 25 meters [Roth05]. Before the year 2000 there was an additional forced calculation error of C/A Code receivers added by the U.S. Government (Selective Availability “SA”, already discussed in chapter 2.6) in order to decrease positioning quality as shown in Table 1.

illustration not visible in this excerpt

Table 1: Max. Predictable Positioning Errors (Probability 95%) [Roth05].

2.8 Measurement of Pseudo Ranges

Referring to Jürgen Roth [Roth05] the principle of calculating a user’s position u on the earth’s surface with the measurement of (at least) three, or for better probability four satellites s signal ranges r is defined by the speed of light c and transmission time t.

The range is therefore defined as:

r = c ˙ Δt

If involved clocks are set up to transmit the exact time we would get the requested range with the formula r = c ˙ t = c ˙ (tu - ts), where tu defines the time of the signal at position u (when receiving the signal) and ts defines the time of sent satellite signal. However a received signal is mostly affected by errors within all clocks involved in the system and by transmission interferences (not included in the approximation), we have to approximate the ranges for an effective calculation, which is then called pseudo range p [Roth05].

The following variables are introduced for calculation of pseudo ranges: ts time of sent satellite signal, tu time of received signal from user, ť s local estimated time of sent satellite signal, δts offset from satellite time to system time, ťu local estimated time of received signal from user, δt u offset from user time to system time, Δt exact runtime of signal range, Δť estimated runtime of signal range, c speed of light.

p = c˙ Δť

= c˙ (ťu – ťs)

= c˙ ((tu + δtu) – (ts + δts))

= c˙ (tu – ts) + c ˙ (δtu – δts)

= r + c˙ (δtu – δts)

Assuming satellite clocks are exactly timed to system time we set δts = 0 and range r is expressed as coordinates from satellite and user (assuming a cartesian coordinate system with a zero-point in earths barycenter) we get the following formula:

p = r + c ˙ δtu

= Abbildung in dieser Leseprobe nicht enthalten+ c ˙ δtu

With this formula representing a calculated pseudo range we are able to start a Taylor series approximation approach [Roth05], calculating a position with intersections of the circles that are defined by the approximated positions within overlapping pseudo ranges on earth’s surface as shown in Figure 4.

Abbildung in dieser Leseprobe nicht enthalten

Figure 4: GPS – Determination of Position [Roth05].

The calculation of positions with GPS systems needs the exact position of satellites, the exact range from each satellite to receiver and exact adjusted clocks within all affected GPS parts. An error of only 1 µs within all involved clocks results in a computed positioning error of 300 meter [Roth05].

There are many sources of errors and so ranging errors are principally grouped into different classes. These errors are in detail, referring to Parkinson et al. [Park+96], ephermis errors (errors in transmitted location of satellites), satellite clock errors (errors of transmitted clocks), ionosphere errors (errors in correction of pseudo range - caused by ionospheric effects), troposphere errors (errors in correction of pseudo range - caused by tropospheric effects), multipath errors (errors caused by reflected signals at receiver's antenna) and of course receiver errors (errors in receiver's measurement of range - caused by thermal noise, software and channel-bias). Refer to Table 2 for a short overview of different sources of errors [Park+96].

illustration not visible in this excerpt

Table 2: GPS – Measurement Error Sources [Park+96].

There are important enhancements for GPS services with additional computed and transmitted correction signals which provide an improvement of accuracy and reliability for GPS. Either this signal is provided by ground stations whose service is generally called Differential GPS (DGPS) service or it is transmitted via additional satellites called Satellite Based Augmentation System (SBAS).

2.9 Differential GPS (DGPS)

A DGPS receiver system receives GPS correction data which was sent form a reference station on the earth’s surface. Each DGPS reference station receives GPS positioning data, processes the gathered information and compares its own precisely known position with the position calculated from GPS satellites. In this way the reference station permanently calculates the differences between the true position known and the calculated GPS positioning data. With this information the DGPS reference station then retransmits the calculated difference via radio stations or a separated long wave band (283.5 - 325.0 KHz) back to a DGPS capable receiver in range of the DGPS transmitting station. This receiver is either an additional receiver or a second receiver integrated in a GPS receiver.

With this added differential GPS technology an average accuracy of three to five meters can be achieve in normal circumstances. But a great drawback to mention with DGPS services is the transmission distance of correction signals. If the distance to the differential GPS receiving unit increases (e.g., signal must not exceed a transmission range more than 200 kilometers from source), the signals provided from the reference station get inaccurate. Another way of broadcasting corrections signals is over the Radio Data System (RDS).

2.10 Satellite Based Augmentation Systems (SBAS)

With Satellite Based Augmentation Systems (SBAS), we summarize different systems and services covering different areas and continents [Woes+07] like WAAS for USA, EGNOS for Europe and MSAS for the Asian region (Figure 5). We now focus on the Wide Area Augmentation System (WAAS) for the North American area and the European Geostationary Navigation Overlay Service (EGNOS). While WAAS is already fully operational since the year 2003, EGNOS is still under development and under testing. EGNOS transmits its signal in WAAS compatible data format over additional satellites (Figure 5) and can be already received with WAAS enabled receivers.

Abbildung in dieser Leseprobe nicht enthalten

Figure 5: SBAS (Red-WAAS, Green-EGNOS, Blue-MSAS) [www.esa.int].

These SBAS signals sent by additional satellites are used to improve standard GPS receiver calculations for positioning with information about influences (due to ephemeris, ionospheric or tropospheric errors referenced in Table 2). The concrete difference between DGPS services and SBAS services is that the latter do not need any additional long wave receiver and of course no additional signals from ground stations are needed. Each SBAS service has to build up a number of Ranging and Integrity Monitor Stations (RIMS, now 25 stations in the U.S. and 10 stations in the EU) which receive GPS data, calculate differences and retransmits GPS correction data to additionally set up Geostationary Earth Orbit (GEO) satellites. These GEO satellites are orbiting in about 36000 meters above the earth’s surface mostly used for telecommunication purposes [Gau+05].

The covered area by SBAS systems therefore depends on where RIMS stations are set up and if GEO satellites are received in the actual target area shown in Figure 5 [Gau+05]). WAAS and EGNOS are able to improve GPS measurement accuracy to a level of one to three meters according to Gauthier et al. [Gau+05], which is supposed to be very good. The European Space Agency (ESA) which is responsible for EGNOS claims to reach positioning accuracy less than two meters when using EGNOS augmenting. But we should take account of, if SBAS enhancements are used in forested or mountainous areas WAAS or EGNOS probably will not work as desired. Wößner and Köhne [Woes+07] say that this is caused due to the fact that there are now three SBAS satellites over Europe with angles between 15 and 40 degrees which are relatively low over the horizon and therefore easily covered by objects like trees, buildings and even mountains.

illustration not visible in this excerpt

Table 3: Combined Typical Positioning Accuracy [Woes+05] [Park+96].

Table 3 (in addition to the statistical probability of accuracy shown in Table 1) provides an overview of average positioning accuracy typical in real world use cases for GPS and assisted technology combinations [Woes+05], [Park+96], [Gau+05].

2.11 Global Navigation Satellite System

GLONASS, the Russian pendant to the Navstar GPS system, is similar to the GPS positioning technology and is also based on trilateration (TDOA). Referencing to Börjesson [Boer00] and GLONASS Center Papers [GNSS02], it also requires at least four satellites for a three dimensional (3D) position fix and a synchronized system time as well. It is also designed for a three component based passive receive only positioning system.

As already introduced with the previous mentioned GPS technology it is based on a ground based control segment, a space based orbiting segment and an user based receiving segment. These components operate together and provide accurate 3D positioning, timing, and velocity data. Actually there are 12 operational GLONASS satellites orbiting around the earth.

The Russian government [GNSS02] commits that there will be about 18 operational satellites by end of the year 2007 followed by promised 24 satellites by end of 2009. With this roadmap, GLONASS will be able to equally fit any need already provided by the GPS positioning system. GLONASS signals are transmitted in the two specific carrier frequencies L1 and L2 sub band. It is a multi component signal, L1 carrier signal is transmitted at 1602 MHz band and L2 carrier at 1246 MHz band. Both carrier signals are a combination of binary ranging codes, digital navigation data and auxiliary meander sequences. Each satellite has its own specific frequency on each L1 and L2 frequency band, which identifies them uniquely.

2.12 Combination of GPS and GLONASS

As, for example, described by Heinrichs et al. [Hein+02] it is already possible to combine GPS and GLONASS systems and there already exist a growing number of commercial combined receivers. So, if we look into the case to combine both GPS and GLONASS positioning systems already possible with a special combination of dual channel receivers together with additional augmenting services like WAAS, EGNOS or MSAS the accuracy of positioning will be increased even in strong forested or extreme hilly environments which now have accuracy problems like our field test area at Zöbelboden. The in some field tests achieved accuracy is under 50 cm (refer to FOGEPOS implementation at the end of this chapter), which defines a new high-performance position measurement. Nevertheless a simultaneous usage as mentioned by Carine Bruyninx [Bruy07] of GPS and GLONASS has to be tested individually in varying environments before reliable usage results in terms of accuracy and reliability can be stated. These dual receivers are significantly more expensive than low-cost receivers.

2.13 NMEA0183 Standard

NMEA0183 is an interface data format developed and specified by the National Marine Electronics Association (NMEA) which is intended to send positioning information in a WGS84 (World Geodetic System 1984) coordinate format (measured in longitude, latitude, and altitude) to computers and other equipment like marine devices. Therefore a GPS receiver's communication is defined within the NMEA standard referenced by Baddely [Badd07]. But there are various NMEA versions and other proprietary formats like Garmin [NMEA06] and Magellan. Ardalan et al. [Arda+00] describe in their work that NMEA specifications may seriously fluctuate within various manufacturers. We focus on the specification version 2 of NMEA0183 standard; a full copy of NMEA0183 standard is available for purchase at NMEA's official web site (www.nmea.org). If we look into this specific NMEA0183 standard, the data received from a communications port of a GPS receiver includes complete position (including longitude, latitude and altitude in WGS84 projection format), velocity and time information computed by the GPS receiving device. Each defined sentence starts with a fixed initial character "$" followed by a two letter prefix "GP" for GPS devices. Then a three letter sequence follows defining the exact contents of the computed NMEA0183 sentence, in our case "GGA".

GGA generic data syntax:

$GPGGA,hhmmss.ss,ddmm.mmm,a,dddmm.mmm,b,q,xx,p.p,a.b,M,c.d,M,x.x,nnnn,*ch

Example field test data:

$GPGGA,161858.590,4750.5404,N,01426.6038,E,1,06,1.9,875.7,M, , , ,0000*0C

Now we introduce this specific “GGA” NMEA sentence used in our project, a valid type of this sentence is shown in Table 4.

illustration not visible in this excerpt

Table 4: NMEA0183 Sentence: $GPGGA.

2.14 Geographic Information System Overview

Information software systems, including geographic and spatial data are called Geographical Information Systems (GIS). They are developed for any need of capturing, analyzing, managing and storing geographic data and according metadata, which are spatially connected and referenced on earth according to the definition given by Tor Bernhardsen [Bern+99].

GIS systems generally define a computer system involving hardware, software, data and applications which are capable of displaying, integrating, editing, analyzing, and storing geographical information. GIS systems even enable manual interactive and automated statistical queries in a specific geographic context as summarized by Henk Ottens [Otte99]. These geographic systems can be used for different purpose, for example in governmental resource and asset management, scientific investigation, environmental impact studies (in context to our project), urban or land planning purposes, cartographic, history, business needs, etc.

illustration not visible in this excerpt

Figure 6: GIS - Basic Functionality Classification.

GIS systems can be classified in three main classes according to their functionality and complexity. A server GIS system provides various and extensive analysis and processing functionality for multiple users, desktop GIS software provides special and customizable analysis and statistical functionality for workstation users and mobile GIS software provides essential and basic functionality for everywhere instant usage with focus on mobile worker’s use. The three types are visualized in Figure 6.

Another way of classifying a GIS system is to classify it based on license issues. GIS software can be distributed as open source software with open code sources under GPL license, as free software with restricted code sources and even as commercial proprietary software.

[...]