Abstracts of previous student projects

4th semester

2nd semester, 3rd semester

2010

Improving the Position Accuracy with DGPS and EGNOS

By: Andrei Marinescu and Dragos Catalin

This thesis presents a study upon obtaining higher accuracy positions than those of standard GPS. The research has been accomplished through means of EGNOS, DGPS and ionosphere models. A fusion of the available methods is employed towards obtaining best results. All data are processed in Matlab through real-time functions, and the results of long time interval experiments are discussed. 

2007

Analysis of Data From the GPS Reference Station at AAU Using GAMIT

By: Isaac Nii Noi Tetteyfio 

The project concerns determination of the velocity of the GPS reference station at Aalborg University as a result of some geophysical phenomena taking place. This is done by analyzing GPS data obtained at the site over period of time using GAMIT/GLOBK software package. In the project, we describe the data archiving, processing procedures, and provide main results in terms of position time-series and velocities. We conclude that the reference station is in motion but to obtain accurately the motion demands certain time span of data to be processed.

Precise GPS Position and Attitude

By: Volodymyr Petrovskyy and Vyacheslav Tretyak

The need of accuracy in GPS positioning is constantly growing. Carrier phase observations are ambiguous and much more precise compared to code phase observations. This project presents a software, which provides precise positioning and attitude determination in post processing mode of a moving tractor in a field. The results showed the position and attitude accuracy of one centimeter and 0.5º, respectively. Furthermore, to present the nature of observations, code phase and carrier phase observations are implemented in a software receiver. However, the carrier frequency accuracy of the PLL output in the tracking channels produced lower position accuracy obtained from code phase observations in this software receiver.

Monitor for a GPS Receiver

By: Olivier Bourrel

Most questions on the Global Positioning System (GPS) are related to the actual status of the system. For a given position on the surface of the Earth, the actual number of visible satellites, their trajectories, the time of their rising and setting, the actual receiver position, the actual GMT, etc are the parameters of interest. The MONITOR displays all these issues on a carefully designed graphical user interface. The establishment of the MONITOR is based on a complete understanding of how GPS works. This conceptual understanding is converted into a comprehensive computation.

A Software-Defined Galileo Receiver

By: Gediminas Veiverys

This master thesis focuses on implementation of a software-defined receiver for the Galileo system. The base of the presented work is an existing SDR GPS receiver that is augmented to support the new Galileo system. Galileo was engineered to be inter-operable with GPS but many issues must be addressed in order to make a second generation GPS receiver work with Galileo signal. Some theoretical background of the receiver as well as the encountered system differences and solutions to adapt to them are given in the report. The execution speed and memory usage requirements of the existing algorithms are improved in order to accommodate the more resource demanding Galileo signal processing. Finally the report is summarized by showing results obtained by processing a real Galileo signal.

2006

GPS Signal Tracking using Software Vector Delay Lock Loop

By: Sarfaraz Anwar Khan

In some situation the received signal strebgth in a GPS receiver is very low. This might be due to large attenuation, high speed, rough receiver movement etc. Inthis case the receiver may loose track of the satellite hence the position will not be available at some critical point which may cause accidents, uncertainty in positions and limits better path-planning. It is therfore necessary to investigate methods in which weak signal could be tracked. In this report we are adressing the problem of signal availability by using a robust signal tracking mechanism called Vector Delay Lock Loop (VDLL) design and implementation in Software. The VDLL is capable of tracking the GPS signal even when the signal is not tracked by conventional DLL. It processes all observable GPS satellite signal at once in a vector such that the loop feeds backs after EKF, which filters the the user position. In this way dynamics of the receiver also become inherent to the tracking system.

EGNOS - Availability Analysis and Accuracy Research

By: Kazimieras Bagdonas

This master's thesis represents the work done during a final project at the Danish GPS Center at Aalborg University, Denmark. It covers four topics: NAVSTAR GPS and DGPS techniques are presented generally and EGNOS with Extended Kalman filtering (EKF) are analyzed comprehensively.
First the general information about GPS is presented together with the basic DGPS algorithms which are implemented as a reference solution in comparison to the researched ones. It is followed by detailed information about EGNOS and the theory behind KF and EKF. In a second part detailed information covering EGNOS implementation and the design of EKF with a motion model for a vehicle is discussed. The EKF is operating on C/A code GPS pseudoranges with applied EGNOS corrections to estimate the positioning solution. A vehicle motion model is implemented to aid the estimation process. After that two types of conducted test are presented. The first one analyses the EGNOS service availability in northern Europe urban environment leading to the conclusion that geostationary satellite is insufficient as a stand alone provider of the service and supplementary means like SISNeT should aid the broadcast of the correction messages. Second test research the performance of the EKF. It was ascertained that discussed EKF is not capable of solving the problem just using the C/A code observables in kinematical mode and additional measurements like Inertial Navigation System (INS) must be added for filter to work properly.

Software-Defined Radio for GNSS

By: Gediminas Veiverys

This master thesis focuses on implementation of a software-defined receiver for the Galileo system. The base of the presented work is an existing SDR GPS receiver that is augmented to support the new Galileo system. Galileo was engineered to be inter-operable with GPS but many issues must be addressed in order to make a second generation GPS receiver work with Galileo signal. Some theoretical background of the receiver as well as the encountered system differences and solutions to adapt to them are given in the report. The executions speed and memory usage requirements of the existing algorithms are improved in order to accommodate the more resource demanding Galileo signal processing. Finally the report is summarized by showing results obtained by processing a real Galileo signal.

2005

Distrbuted Kalman Filters on FPGA for Ship Positioning

By: Usman Ahmed Zahidi

GPS is the most successful alternative of finding positioning, time and their derivatives. Similar to other engineering diciplines, GPS positions are subjected to errors due to noise which is incurred by several error sources such as Ionospheric and Tropospheric delays, Signal Multipath or clock errors etc. These errors may result in wrong information to any user vehicle, which could endanger life.Instant filtering is therefore inevitable. Kalman Filter is the most suitable filtering technique for GPS, as it can instantly filter the GPS observation. It is a linear adaptivee filter butsome variants could be used in non-linear domain. One of the major drawbacksof this filter is time-complexity, as well as several matrix multiplication operations are involved. This thesis presents the Distributed Kalman Filter design and implementation on FPGA for ship positioning. In order to to filter the positionin real-time a distributed filter is designed. FPGA is used as an implementiation platform and kinematics of ship is involved for filter design. Immense gain in performance is observed, however the level of task distribution is kept minimum due to the resource constraints on FPGA. Considerable error filtering results are also reported in this thesis.

Software GPS Receiver

By: Kostas Dragūnas, Vitalijus Linikas and Xiangfu Meng

The software receiver is an enabling technology for flexible implementation of a receiver in which significant amount of the signal processing is accomplished in software rather than in hardware. This report documents a prototype stationary software GPS receiver working in a post processing mode. Receiver is able to do GPS signal acquisition and tracking, to decode navigation message, and to compute pseudoranges. As an addition the different position computation algorithms such as least squares, weighted least squares, averaging, discrete Kalman filter and extended Kalman filter are analyzed and tested.

Discrete Kalman Filtering for Ship Positioning

By: Stephen Kwabena Kportufe and Anders Skeie

Effective routing system to reduce the amount of traffic and accidents at sea is a matter of great concern to ship navigation. The aim was to obtain optimal positions of the ships path and the continuous availability of positions when missing data and outliers occurs. Positional information available from GPS receivers are erroneous but are considered the best position for the Kalman filtering process. The Kalman filter is applied on measurements from two GPS receivers at sea. Different filter models are developed to suit different situation of the ship path. A Kalman filter bank is created to accommodate all the dynamics of the ship path. Several filter types are implemented such as Linear and Extended Kalman filtering, the Bayes filter and smoothing.
In results, the linear Kalman filter, the Extended Kalman filter (EKF), and the Bayes filter are analysed and comparisons made under different conditions. The linear Kalman performed closed to the Bayes filter under certain conditions. It was realized that there was no clear cut model which operates successfully under different motion models. The bank of filters was adjudged to be the best approach to the filtering problem.

2004

Studies of the Ionospheric and Tropospheric Delay by GPS observations in Denmark

By: Marianne Knudsen and Rune Tvillum

This project concerns the inospheric and tropospheric delay in Denmark. It consists of three intermediate aims, which are listed as follows: 1: To investigate whether it is possible to distinguish between short and long baselines 2: To investigate whether it is possible to improve the estimation of the tropospheric delay at RTK position by a tropospheric model 3: To investigate whether GPS observations can be utilized in Numerical Weather Predictions (NWP) In the project we conclude that it is not possible to define boundaries between short and long baselines. Also the tropospheric delay at a RTK station will not be standard RTK. Finally we document that GPS observations can be utilized in NWP.

Design of a single Channel GPS Software Receiver

By: Nicolaj Bertelsen and Peter Rinder

Traditionally, most GPS receivers are based on dedicated hardware for the most time consuming signal processing algorithms and multi-purpose processors for the remaining parts of the necessary calculations. With the eternal increase in computer processing power, it is now possible to implement an entire GPS receiver on a multi-purpose processor. This is referred to as a GPS software receiver. The Present Master Thesis contains in the design and implementation of a single frequency GPS software receiver. The necessary hardware for the software receiver includes a GPS antenna and a GPS RF front-end dnated by Simrad Danmark A/S. A National Instrument data acquisition board was used to digitize the analog data supplied by the front-end A GPS signal simulator was designed and implemented in Simulink to be used during algorithm design and test. The necessary algorithms needed to acquire and track the GPS signals were successfully designed and implemented in Matlab.

Software GPS receiver

By: Ramūnas Dirmeikis, Darius Plaušinaitis and Jon Johansen Wedel

Software GPS receivers have been popular for some years now in scientific circles, but have yet not had a commercial breakthrough. Although this receiver type offers new ways of marketing because of the upgradeability and flexibility, it also has a few drawbacks: lack of speed and large storage requirements. This leads to huge amounts of data that needs to be processed in a very short time interval. One of the ways to address these two problems could be to sample the GPS signal with a quantization level of 1 bit. This would reduce the amount of storage required and possibly pave the way for other signal processing functions optimized for 1 bit quantized GPS signals. In the project software GPS receiver is analyzed, designed and implemented for processing a 1 bit quantized signal. This is done by splitting the functions of a GPS receiver into several main parts. Each part was tested separately using real data collected from Simrad front end. Test result show that each part is working within the requirements set for this project. Based on the test result, it is concluded, that there might be opportunities in using 1 bit quantized signals for software receivers, but more work and tests is needed before a final statement can be given about the validity of this approach.

Optimal filter to estimate the non-linear movement of a vehicle

By: Frank Kofi Ocran and Saulius Pušinskas

With the recent development in GPS precise positioning using RTK technology, a receiver can measure a vehicle’s position within a few centimeters.GPS measurements like other survey measurements are not devoid of errors. However, the system utilizes the concept of DGPS to eliminate common errors but some random errors and positional offset due to computational latency still exists. Filters which have the ability to reduce noise and extreme errors in GPS measurements and aid prediction are the solution.This project will attempt to retrieve the behavior of the motion from a prerecorded data. Virtual trajectories will be generated with similar behavior and motion parameters of several types of vehicles. Several approaches of the Kalman filter together with other techniques will be examined on the trajectories. The optimal filter which shows the best performance when filtering the nonlinear trajectories with some noise added will be chosen from the set of filters tested.Test results obtained show that the filters have their own strong sides and weaknesses.  The best performing filters have been pointed out and investigated under what conditions in particular they perform best.

2003

Aeroplane Attitude Determination Using GPS

By: Tue Klaus Kyndal

This project concern the real-time positioning and attitude determination of a flying aero plane using GPS measurements together with some related problems in flight navigation and system setup in photogrammetry and remote sensing. I was given the opportunity to work with data collected from a Javad AT4 (4x2x20) channel receiver. This receiver is especially designed for attitude determination and real-time applications. The attitude determination system is based on RTK differential GPS algorithms that estimate three body frame baselines. The attitude information is derived manipulating these baselines by projecting them onto a local horizontal plane. The system uses Kalman and Bayes filters to track the three baselines and their float ambiguities. Goad's wide lane technique is used to fix the float ambiguities prior to the baseline estimation. The final system will compute the body frame position and attitude in reference to a local horizontal plane, and a number of navigation parameters, such as the plane heading and correct bearing in relation to the drift coursed by a crosswind component. Download project.
Download related paper.

Web-based GPS Processing Service

By: Stephen N Asamoah 

Modern day processing algorithm of GPS data for sub-centimeter level accuracy demands that two sets of expensive dual frequency geodetic receivers are used, and length of the baseline restricted to 15 km or less.

In addition, processing of data by use of commercial state-of-the-art software packages that is able to post-process raw data to achieved the desired results could be very awkward than the data capture itself. These make GPS survey technique less attractive compared to other techniques.

For this reason an Internet based automatic processing service could be very helpful to users in many ways.

This project contains a documentation and a demonstration of how a web site that is capable of accepting raw GPS data and process it automatically can be very useful to users. In that, user will no longer need to buy two expensive dual frequency receivers and also do not need to install or understand the software packages for processing GPS data.

Programming languages such as HTML, PHP, Javascript, PostgreSQL, Matlab, and UNIX Shell was extensively used in design and implementation. The project also contains a study of all online GPS processing services.

2002

Open-loop Processing in the Context of GNSS based Radio Occultations

By: Mindaugas Paukštė, Mindaugas Stonys, and Jevgenij Gagac

This project is focussed on Radio Occultation data processing issues in the framework of Global Navigation Satellite System. GNSS signal are measured by the means of a GNSS receiver for operational NWP (Numerical Weather Prediction) and global climate research. A brief description of GNSS signals formation and propagation is given. In this project the GPS signals are discussed. The signal measurement tools, such as a GPS receiver are described. The main focus is on a GPS receiver operating in Open Loop mode architecture. The error models affection a GPS signal while travelling are investigated and discussed, as well at the errors of the measurements produced by the GPS receiver while processing. The software model of the GPS receiver operating in Open Loop mode is created. The reference model is created relying on GRAS instrument specifications. A thorough discussion on Ground Processing of integrated I and Q values issues is given. The assessment of real world data and simulation with the models of it is done. Simulations with the created software model, introducing different errors on a signal and in the channel are done. The influence of errors on integrated I and Q signal is discussed and summarized. The software model simulation results show, that the model is well preforming and can be used as an interpretation of a real Open Loop GPS receiver for preliminary research on the Radio Occultation processes.

Accurate Positioning in Real-Time Kinematic

By: Inga Paulauskaitė

This master's thesis represents work done during a final project at the Danish GPS Center, Aalborg University. The aim of this project was to achieve accuracy similar to what can be found with the supplied software base of the actual devices. This report is divided into five important chapters. The first chapter presents one of the fastest methods for finding the position with one receiver: Bancroft algorithm. Then we introduce the LAMBDA method which is used for solving the ambiguities (when using two receivers) and a study on when to fix the ambiguities (theoretical and empirical criteria). After that, three Kalman filters are developed and compared in a virtual trip. Finally, we paid our attention on the implementation aspect and all the problems which needed to be solved. The results showed that the proposed solutions yielded a high accuracy either in static or in kinematic.

2001

Adjustment of the Lithuanian GPS network using GAMIT

By: Darius Popovas and Kristina Radzevičiūtė

The interest of getting precise point positioning was growing together with the new technology development. The Earth surface measurements for geodetic control networks started with very simple measurements, as a triangulation, or trilateration, but anyway it lead to sufficient accuracy of the networks. Complicated and high technology instruments helped to achieve better accuracy, until NAVSTAR Global Positioning System (GPS) revolutionized the tasks of navigation and surveying. Today, the most accurate geodetic networks, covering even the whole continent, are established with the help of GPS. Lithuania was one of he countries taking a part in GPS network establishment for northern countries in 1992. This network was connected to European network. The measurements for the densification of the Lithuanian GPS network were taken in 1993 and all calculations and adjustments were done. The main task of the theses is to readjust the Lithuanian GPS network. A previous adjustment was done in 1993, when the technical base to accomplish this task was limited. Eight years later with the new powerful software and hardware it became possible to implement advanced tasks. The data from 1993 was reprocessed first separately session by session, and later combined into final adjustment taking into account the correlations between different sessions. The results were compared with the previous ones from 1993. Download project.

Monitoring GPS Observables and Positions

By: Saulius Prizginas and Dalia Prizginienė

The purpose of this project is to procure GPS measurements, to get the reasonable outcome after postprocessing of the data and finally to present it. The task for presentation is to build up the interface where various GPS data and results can be seen. It has to be simply understandable and readable, attractive and easy to use. In this project we decided to concentrate on presenting five fields of GPS: Receiver position, Orbits of the satellites, Satellite health, Dilution of Precision values and Signal strength. We present these topics by creating a Graphical User Interface (GUI) so general that any user can use the program for processing his/her data and presenting his own results.

Real-Time Kinematic GPS

By: Adeniyi Iyiade and Kofi Owusu-Nkasah

Until quite recently positioning by differential GPS was only feasible in post-mission, but today the need for real-time positioning has increased tremendously. This study develops a real-time kinematic GPS. As GPS measurements like other survey measurements are not devoid of errors, the study utilises the concept of differential GPS to eliminate common mode errors. The final part of the work utilises an extended Kalman filter as an attempt to reduce the reception noise. The end results show a sub-centimetre level positioning, which is the gal of this project work.

Remote Tracking System For Alzheimer Patients

By: Audronė Kvedarienė

The requirements for the system were established, considering the peculiarities of the disease and different possible solutions provided. The emphasis was on the development of the equipment, which will be used for the private purpose. The detailed algorithms were proposed for this solution. Under several assumptions the prototype equipment was developed and initially tested. Verification showed that the designed patient side device works properly, but at the caregiver side performs only the basic ideas therefore further significant developments are required. However the extensive improvements and testing indispensable for the overall equipment before implementing the proposed solutions to the final product.

Useful links:

> Project library

Danish GPS Center, Fredrik Bajers Vej 7C, DK-9220 Aalborg Ø, Denmark, Tel +45 9940 8362, E-mail: gpscenter(at)gps.aau.dk