Master Thesis ProposalJohan MalmströmSeptember 12, 20021BackgroundKnowing your position is, in many applications, very important. There are severalways of solving this problem. INS1 and GPS2 are two different types of navigationsystems which can tell the position of the user. The two systems both have weaknesses, e.g. GPS has a relatively low update frequency and INS has a time-growingerror. If the two systems are integrated their complementary properties can give abetter performance.There are two ways of combining the INS and GPS data. The first way is to doa decentralized integration where the GPS and INS data are processed separately.The GPS position is only used as a supporting system to estimate the drift in theINS position. The second way is to do a centralized integration. Here all the rawdata from the GPS are used in the integration.The first solution is easier to implement but is a sub-optimal solution since not allavailable data is used. The second solution uses all GPS data for estimating theINS error. It is also possible to use the GPS data even with less then four satellitesavailable. In the first solution signals from four satellites are required to calculatea position in the GPS.One drawback with using GPS is that it is very sensitive to jamming, due to thelarge distances to the satellites. An effective way to improve the robustness againstjammers is to use a adaptive directional antenna for the GPS. The satellite positions are known so the antenna gain can, in some way, be maximized in thesedirections.2Problem definitionIn this master thesis project a tightly coupled Kalman filter, that uses centralizedintegration between GPS and INS data, will be implemented. This filter will useboth raw GPS data (pseudo-range and phase information) and information froman INS (gyros an accelerometers). The algorithms should be expressed in an EarthCentered Earth Fixed (ECEF) coordinate system.1 Inertial2 GlobalNavigation SystemPositioning System1
3 Method2The simulation environment should also include the possibility to arrange differentcombinations of jammers and to change type of antenna for the GPS. The performance will be investigated for the case with an omni-directional antenna andan adaptive antenna. The gain for adaptive antenna will be optimized for twosituations, only satellite positions known and both satellite and jammer positionsknown.3MethodThe raw GPS and INS data will be integrated using a tightly coupled Kalman filter.A theoretical model for the relation between the states used in the filter and theobservable variables will be made.To simulate a GPS receiver satellite orbits must be calculated. Several errors termsoccur, both in the GPS and in the INS. E.g. the ionosphere introduces a unknowndelay to the clock signal from the satellites. The errors that are possible (andworth) to estimate and compensate for will be carefully chosen.The Adaptive antenna will mainly be tested in a simulation environment developedby FOI. The Electronic warfare department at FOI in Linköping will be contactedfor guidance in this area.4OrganizationThe Master thesis is performed at the Swedish Defense Research Agency (FOI).The supervisor at the agency is Bengt Boberg. Examinator and supervisor at theS3 department at KTH is Ph.D. Magnus Jansson.4.1Contact informationNameJohan MalmströmBengt BobergMagnus email@example.comPh. no. (work, cellular)08-55 50 ? ?, 070-654 22 2608-55 50 36 56, 070-92 77 32908-790 84 43Project planningThe thesis project has been separated into four sub-projects (see Gantt scheme, appendix B). The sub-projects are Planning & documentation, INS/GPS integration,Adaptive antenna simulation and Examination. Each subproject has milestonesand/or tollgates (see also Timeplan, appendix C).The INS/GPS integration is common with another Master thesis performed at FOI.The work with these parts will therefore partly be done together.
5.15.1Milestones3MilestonesSeveral milestones are connected to the project. Milestones are points in time whena certain result should be achieved. Putting in milestones makes it possible to seeif the project advances as planned.12345675.2123456Output DocumentMaster thesis proposalTheoretical filter modelSimulation report, INS/GPSAdaptive antennas for GPSSimulation report, Adaptive antennaDraft of Final reportFinal reportWeek3739434650512TollgatesTollgateStatus report 1Status report 2Status report 3OppositionOral PresentationWeek38434846-492-3ReportingThe progress of the project will be reported in Status reports with regular intervals(see 5.2 Tollgates). The status reports describe the current status of the project.A final report will summarize the performance and result of the project. It will bewritten in English and typeset using LATEX.A first coarse draft of the content in the Final report is found in appendix A.All documentation connected to the project (e.g. Status reports and time plans) willbe published on the web address http://www.e.kth.se/ e98 jom/master thesis/.
A Skeleton for Final reportASkeleton for Final reportIntroduction– Background— INS— GPS– ObjectProblem definition– SpecificationMethods– Theoretical modelImplementation–Integration of GPS/INS–Adaptive antennaResults– Simulation resultsConclusions– Future workReferencesAppendix4
Gantt chart40Preparation for OppositionOppositionPrepare PresentationOral presentationFinal reportTotal number of hoursNumber of hoursTotal timeExamination208Adaptive antenna simulationInvestigation/Literature studiesVisit FOI LinköpingImplementing matlab toolboxToolbox documentationSimulationEvaluationTotal time243INS/GPS integrationLiterature studiesSatellite empherisis generationTheoretical filter modelFilter implementationMatlab documentationSimulationEvaluationTotal time239Planning & documentationProject planningStatus reportingTotal timeActivity3530535036Master Thesis in Signal Processing at the Swedish Defence Research Agency 4310352055051403010December052Updated 2002-09-12402515130201022January2376004Author: Johan Malmström, firstname.lastname@example.orgBGantt
Time planWeek 4745678Tollgate3:11:41:1 Master Thesis proposalPlanning & documentationWeek 391:14:1567895689Week 5272:2 Simulation report, INS/GPS2:1 Theoretical filter model3:24:2Week 4423453:2 Simulation report, Adaptive antennas3:1 Description of adaptive antennasAdaptive antenna simulation1January 2003Week 1684:39Week 31:312356Week 44NovemberWeek 45789104:1 Opposition4:2 Draft of Final report4:410 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26Examination7Week 22:210 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 3110 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31INS/GPS integration44Week 434:4 Oral presentation3Week 5131:2Week 424:3 Final report2Week 502Week 411:4 Status report 31DecemberWeek 492:11OctoberWeek 40Updated 2002-09-12Author: Johan Malmström, email@example.com:3 Status report 21:2 Status report 1Week 3810 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30Week 489Week 3711 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30November (cont)Week 46Milestone2002-09-122002-09-123126 27 28 29 30 312SeptemberWeek 36AugustWeek 35Master thesis project 2002CTime plan
Gantt chart Updated 2002-09-12 Master Thesis in Signal Processing at the Swedish Defence Research Agency (FOI) Author: Johan Malmström, firstname.lastname@example.org 35 36 37 38 39 .