202 JOHNSHOPKINSAPLTECHNICALDIGEST,VOLUME23,NUMBERS2and3(2002)
C. J. DUHON
T
TacticalDecisionAidforCECEngageonRemote
Christopher J. Duhon
heCooperatingUnitPositionPlannerisaplanningtooldevelopedforcalculatingcooperative Aegis engagements against low-elevation cruise missile threats where theshootingshipmaynotholdthethreatwithitsonboardsensors.Thetoolallowsoperatorstocreate scenarioswithup toeightAegiscruisers,viewengageability regions innear–real time, and then refine the scenario to enhance coverage against a threat. BattleGroupCommanderscanthenplanshippositionstotakeadvantageoftheCooperativeEngagementCapability’sremoteengagementpotential.Thetoolintegratesseveralexist-ingmodelsforremoteengageability,atmosphericradarpropagation,andfirmtrackperfor-mance;combinesthemwithanup-to-dateintelligencedatabase;andprovidesagraphicaluserinterfaceembeddedintheCommonDisplayKernelsoftwarepackage.
INTRODUCTIONThe Cooperative Engagement Capability (CEC)
providesrevolutionaryairdefensecapabilitiestoNavysurface warfighting platforms by distributing sensor,weapons,decision,andengagementdataamongbattlegroupmembers.1CECoperationalprinciplesofcompos-itetracking,precisioncueing,andcooperativeengage-mentsresultinsignificantlyextendedbattlespacesandengagement zones. Composite tracking allows trackstobeformedandmaintainedmoreaccuratelybymerg-ingsensordatafrommanyplatforms.Precisioncueingallowsforearliertargetacquisitionbyonesensorbasedoninformationfromsensorsonotherships.Withcoop-erativeengagements,ashipcanengageandfireuponathreatusingremotedatafromanothership,evenifthatfiringshipdoesnotseethethreatwithitslocalsensors.
CEC’sabilitytoperformAegisengagementsusingremote sensor data provides a unique, distributed
weapon system capability. Because of CEC, a com-mandercanaltershipplacementstoincreasethebattlegroup’seffectivenessagainstthreatsandtobetterpro-tect the battle group’s assets. These ship placementsmay not be standard or conventional. For example,CEC-equipped ships, known as Cooperating Units(CUs), can be placed farther apart than non-CEC-equippedships.ToassisttheBattleGroupCommandand Air Defense Coordinator in planning CU posi-tions,wehavedevelopedaCUPositionPlannerthatrunsonaCECdisplayconsoleinthecombatsystemofCECships.Usingthisdisplay,anoperatorcancreateexperimental planning scenarios and, within a fewseconds, can view the battle group’s effectiveness inengagingthreatsusingCEC.
Aboard Navy ships, the CEC interfaces withradars, other sensors, the combat system, and ship
JOHNSHOPKINSAPLTECHNICALDIGEST,VOLUME23,NUMBERS2and3(2002) 203
TACTICAL DECISION AID FOR CEC ENGAGE ON REMOTE
personnel,amongmanyotherassets.Interactionwithship personnel is accomplished via the CEC displaysystem.SomeCECdisplayfunctionsmaybeintegratedwith an existing combat system display, such as theAdvanced Display System aboard Aegis cruisers, ormay reside in stand-alonecomputer systems (Fig.1).Ineithercase,thedisplaysmustallowtheoperatorstointeractwithCECandtoperformnecessaryCECfunc-tions. In addition, these displays abstract and digestvoluminousamountsofdataandpresentthesedatatothe operators to assist in their decision-making pro-cesses. These display tools are collectively known astacticaldecisionaids.TheCUPositionPlannerisonesuchaid.
ThePlannerbringstogether(bothinreal-timesoft-wareandpreprocesseddatafiles)severalexistingtools,addsnewcapabilitiesforintegratingandmergingthosetools,andprovidesanadvancedgraphicaluserinterface(GUI).Theexistingtoolsincludeatmosphericpropaga-tionmodelsfordeterminingradarperformanceinevap-orativeductingenvironmentsovertheocean’ssurface,theaccreditedhigh-fidelityAPL-developedAN/SPY-1BradarsimulationforpredictingSPYperformanceagainstlow-flyingthreats,andacooperativeengagementmodelfordeterminingengageabilityagainstthosethreatswhenusingremotedata.TheCUPositionPlannerdoesnotattempt to predict engageability using local data, butinstead predicts engageability when sensor data origi-natefromashipotherthantheshooter.
The Planner is currently installed and operationalaboardUSSJohn F. Kennedy(CV67),Eisenhower(CVN69), Wasp (LHD 1), Cape St. George (CG 71), Anzio(CG68),Hue City(CG66),andVicksburg(CG69),and
attheSurfaceCombatSystemsCenter,WallopsIsland,andtheNavalSurfaceWarfareCenter,DamNeck.
OVERVIEWThe CU Position Planner allows the operator to
definebattlegroupscenariosbyinteractivelyaddingandpositioningAegiscruisers(theonlyCUtypeforwhichthetoolcurrentlyworks),selectingaspecificthreat,andestablishingadefendedpointwhichthethreatisattack-ingandtheCUsareprotecting.Clickingabuttononthe tool’sGUI initiates the engageability calculationsand,withinafewseconds,displaysregionswhereasuc-cessful engagement can occur. These regions indicatewhere the given threat can be engaged by some CUinthebattlegroupusingremote(non-local)data.TheoperatorcanthenexperimentallyrearrangetheCUstoenhanceengageability.
The problem of determining engageability for theentire battle group is broken down into successivelysmallerproblems,theresultsofwhicharerecombinedtoprovidethefinalanswer.Battlegroupengageabilityis determined by calculating engageability for eachindividual CU in the scenario and then combiningthose results graphically on the display. In this way,eachCU isgiven theopportunity tobe the shooter;i.e.,battlegroundengageabilityE forNCUscanbeexpressed as the union of CU engageability Ei forallCUs:
E Ei
N
i==U .
1
IndividualCUengageabilityisfurtherbrokendownbyconsideringeachoftheotherCUsinthescenarioasremotedataprovidersoneatatime.Thatis,Ei canbedefinedas
E E j iij
N
ij= ≠=U , ,
1
whereEijisengageabilityforCUishootingwithremotedata from CUj. This approach can be seen in Fig. 2in which three CUs—Cape St. George (CSG), Anzio(ANZ),andHue City(HUE)—formthebattlegroupscenario, which results in six individual engageabilitycalculations.
Foreachshooter/providerpair,engageability iscal-culatedthroughouta100100nmisquarearoundtheshooterandorientedtowardtheprovider(Fig.3).Pointsat1-nmiincrementsareselectedandtestedforengage-ability,whichresultsin101101=10,201pointstestedfor each pair. At each point a series of engageabilitytests is performed. If all tests pass, then the threat is
Figure 1. A typical stand-alone CEC display system. Shown here is a Sun Ultra 1 workstation running in a CEC development lab-oratory at APL. Shipboard systems are similar, with a trackball replacing the mouse.
204 JOHNSHOPKINSAPLTECHNICALDIGEST,VOLUME23,NUMBERS2and3(2002)
C. J. DUHON
saidtobeengageableatthatpoint;ifanyonetestfails,thenthethreat isdeterminedtobenotengageableatthat point. These calculations take approximately 1 spershooter/providerpair.
COMPONENTSAs battle group engageability is broken down into
setsofsmallerproblems,engageabilityiseventuallycal-culatedatanindividualsamplepointwithasequenceoftests.Thesetestsapplyvariousradarandweaponsystemperformancemodels,whicharedescribedhere.
Firm TrackBecausetheCUPositionPlannerdetermineswhether
ashootingCUcansuccessfullyengageathreatwhenpro-videdengagementsupportdata fromaremoteCU,theremotedataprovidermustbeabletoseethethreatwithits AN/SPY-1B radar. Output from APL’s high-fidelitySPYfirmtrackmodelisqueriedateachsamplepointtodetermine if in fact the threatcanbe seenby thedataprovider.Athreatissaidtobeheldinfirmtrackiftheradarhasseenitseveraltimesandhasidentifieditasarealobjectdistinctfrombackgroundclutter.
Before the CEC display software is delivered, thefirmtrackmodelisrunforaselectedsetofthreatsandresultsarestored intotextfilesthatare includedwitheachsoftwaredelivery.Thesefiles,queriedatruntime,areparameterizedbythreattype,atmosphericevapora-tiveductheight(asspecifiedbytheoperator),andthethreat’s cross-range distance from the providing CU.Cross-range distances, also known as closest points ofapproach (CPAs), are in 2-nmi increments. For eachincrement,twodownrangedistancesaregiventhatindi-catewherethethreatwillbecomefirmtrackandwhereitwillbedropped.Itisbetweenthesetwodistancesthatthethreatwillbeseenbythedataprovider’sSPYradar.
AgenericfirmtracktableisdepictedinFig.4a.TheshadedregionsindicatewheretheproviderCU,whichisinthecenter,willholdathreatinfirmtrackforvary-ingCPAs.Figure4billustrateshowthefirmtracktableisqueriedforaparticularsamplepoint.Inthisexample,withCSGshootingondataprovidedbyANZ(iteration1fromFig.2),thesamplepointinquestionisheldasafirmtrackbyANZ.Thus,ANZisabletosupportCSG’sengagement.
PropagationAnengageabilitycalculationatasamplepointcon-
sidersmanyfactors,oneofwhichisadeterminationofwhethersufficientAegisilluminatorenergyisreceivedbyStandardMissile-2(SM-2).ThisilluminatorenergyissentoutbytheshootingCU,reflectedoffthethreat,anddetectedbySM-2.TheCUPositionPlannermustdeterminehowmuchoftheilluminatorenergycanbeseenbySM-2.
Illuminator energy is attenuated by several factors,including distance between shooter and threat, dis-tancebetweenthreatandSM-2,andradarcross-sectionof the threat. Additionally, the energy can be both
X
Samplethreat path
CSG
ANZ
HUE
Defendedpoint
Individual engageability calculations
Iteration Shooter Provider1 CSG ANZ2 CSG HUE3 ANZ CSG4 ANZ HUE5 HUE CSG6 HUE ANZ
Figure 2. Battle group engageability broken down into shooter/provider pairs. A sample scenario comprising three CUs, a defended point, and a nominal threat trajectory is shown. Engage-ability for the entire battle group is calculated by separately con-sidering each CU as a shooter and, for each shooter, considering every other CU as a remote data provider, as shown in the accom-panying table.
Figure 3. Sample points (spaced regularly every 1 nmi) in a sample region around a shooter. This represents iteration 1 from Fig. 2, with CSG shooting and ANZ providing remote data. An engageability calculation is performed at each point in the 100 100 nmi square around CSG. The sample region is oriented so that the provider is considered lying due east of the shooter.
50 nmi
ANZ
CSG
JOHNSHOPKINSAPLTECHNICALDIGEST,VOLUME23,NUMBERS2and3(2002) 205
TACTICAL DECISION AID FOR CEC ENGAGE ON REMOTE
weakenedandstrengthenedatdifferentpointsbyprop-agationthroughtheatmosphereneartheoceansurface,aswell asby reflectionsoff theocean surface.OutputfromtheAPLTroposphericElectromagneticParabolicEquationRoutine(TEMPER)radarpropagationmodel2isusedtodeterminethislatterformofattenuation.
As with firm track data, TEMPER models are runforvariousatmosphericconditions,andtheresultsarestoredintextfilesanddeliveredwiththeCECdisplaysoftware.Thefilesareparameterizedbythreataltitudeandthedistancebetweenthethreatandtheshooter’silluminator, and their values are queried at run time.Values in the files indicate whether the illuminatorenergy is increased(e.g.,byconstructive interference)ordecreased(e.g.,bydestructiveinterference).Figure5providesavisualizationofonesuchTEMPERpropaga-tionfile.
Remote EngageabilityTheCUPositionPlannerdoesnotattempttodeter-
mine engageability when the shooter is tracking thethreatwithitsonboardradarsystems.Instead,thetool’spurposeistoshowtheusertheenhancedengageabilityaffordedbyCECwhen radarand sensordataarepro-videdtotheshooterfromaremoteAegiscruiser.T.P.NguyenofAPLhasdescribedcalculations thatdeter-mine whether an individual sample point is engage-able. A model that implements those calculationsfor remote engageability has been developed3 usingMATLAB (a commercial software package for math-ematical numeric processing, visualization, and simu-lation). The MATLAB model was subsequently con-verted to C++ for integration into the CU PositionPlanner.
GRAPHICAL USER INTERFACETheoperator interactswith theCUPositionPlan-
nerviaitsGUI(Fig.6).TheGUIallowstheoperatortoenterenvironmentalinformation,threattypeandloca-tion, and anticipated operational area defined by thelocation of the defended point and the CUs consti-tuting the battle group. As objects such as CUs areaddedtothescenario,theyappearimmediatelyinthedisplay’sPlanPosition Indicator(PPI)window.UsingtheModify in PPIbuttons,theoperatorcanrepositionobjectsby simplyclicking themouseanywhere in thePPI.Aspecial featurecalledGet Real Time Position DatacanbeusedtoautomaticallyextractthepositionsofAegisCUsandanyaircraftcarrierfromthereal-timedatamaintainedbyCEC.TheSave…,Restore…,andDelete…buttonsallowtheusertomaintainanarchiveofscenarios.
Threat in firm track
CPA
Threat NOT in firm track
2 nmi
Threat trajectory
Firm track areasignored by planner
Downrange
Provider
CSGshooter
Defendedpoint
(a)
ANZprovider
Downrange
CPA
Sampleregion
Samplepoint
(b)
Figure 4. Firm track tables. (a) Generic table in which each shaded strip represents the area in which a threat, moving from right to left, is visible to the provider’s radar. The actual table con-tains only the upper half, since the values are symmetric. (b) Table for ANZ with sample point in ANZ’s firm track region. The sample region is shown around CSG, with one sample point highlighted. The table is reoriented for each sample point so that the strips are parallel to the threat’s trajectory to the defended point.
Figure 5. One-way propagation factor for a 22-m evaporative duct. This image illustrates how illuminator energy is both ampli-fied and attenuated over the ocean surface. Note that the image is highly out of scale, with the vertical axis in feet and the horizontal axis in nautical miles.
250
200
150
100
50
05 10 15 20 25 30
1.2
1.0
0.8
0.6
0.4
0.2
Range (nmi)
Alti
tude
(ft)
10P
ropa
gatio
n fa
ctor
/10
206 JOHNSHOPKINSAPLTECHNICALDIGEST,VOLUME23,NUMBERS2and3(2002)
C. J. DUHON
INTEGRATIONThe computer applications con-
stituting the CU Position Plannerare set up in a client/server par-adigm. Figure 7 shows this archi-tecture, in which circles are appli-cations, rectangles representUNIXsharedmemorysegments,andarrowsindicate general information flow.
Figure 6. CU Position Planner graphical user interface.
The GUI creates a scenario and stores it into a Common Display Kernelnamed buffer.4 The server application, which performs the actual engage-abilitycalculations,readsthescenariowiththeassistanceofanamedbufferdaemon(nbufd).Thedaemonsprovideaccessmechanisms for readingandwriting named buffers transparently so that applications need not be con-cernedwithwhetheranamedbufferresidesinlocalmemoryorinthememoryofsomeothercomputer.(TheserverneednotresideonthesamecomputerastheGUIandPPI.)Results,storedasasequenceoflatitude/longitudepairsindicatingwhereanengagementagainsttheuser-specifiedthreatwillbesuc-cessful,arestoredontotheGUI’scomputerforprocessinganddisplayinthePPI.Thedatageneratorapplicationtransformsthesequenceofpoints(aswellasotherinformationfromthescenario,suchasCUpositions)intoabufferofdrawinginstructionsusingtheGraphicalEntityData(GED)language,5,6whichistransferredtothePPI.ThePPIapplicationrendersthosedrawinginstructionsassymbolsandgraphicsintoanXWindowsdisplay.
As the serverbreaksdowntheengageabilityproblem into successivelysmaller steps, it eventually considers a single sample point (Fig. 3). Asequenceoftestsisperformedoneachsamplepointtodetermineifathreatcanbesuccessfullyengagedatthatpoint.Thetestsperformedinclude
• Firmtrack,whichtestswhethertheprovidercanseethethreat(Fig.4b)• Illuminatoraccuracy,whichcheckstoseeiftheshooter’silluminatorcan
pointatthethreatwithsufficientprobabilitytosupporttheengagementusingtheremoteprovider’sdata
• Illuminator power, which determines if there is sufficient illuminatorpowerreceivedattheSM-2radomeforthemissiletosuccessfullyseekandengage
• Seekeraccuracy,whichdeterminesiftheSM-2seekercanfindthethreatwithsufficientprobabilityforasuccessfulengagement
• Dopplershift,whichdetermineswhethertheSM-2seekerwillbeabletodistinguishthethreatfrombackgroundclutter
Thesetestsareruninorder.Asamplepointissaidtobeengageableifandonly ifall testsarecompletedsuccessfully. Ifanytest fails, the remainingtestsarenotperformed,whichsavesonprocessingtime.
AsmentionedintheComponentssection,theCUPositionPlannerinte-grates a number of previously existing models and tools. The tools weredesigned independently by separate groups of engineers, and their outputswerenotdesignedforintegrationintoasingleengageabilitytool.Oneprob-lemresultingfromthiswasthemixtureofcoordinatesystemsusedbythevari-ousmodels.Forexample,firmtracktablesplacethedataproviderCUatthe
GUI
nbufd nbufd Server
DatageneratorPPI
Data
Dis
play
net
wor
k
X
Results
Scenario
Figure 7. Applications constituting the CU Position Planner.
JOHNSHOPKINSAPLTECHNICALDIGEST,VOLUME23,NUMBERS2and3(2002) 207
TACTICAL DECISION AID FOR CEC ENGAGE ON REMOTE
DP = defended point at (0, 0),
Pt = (xPt, yPt),
Sh = shooter’s position (xSh, ySh),
Pr = provider’s position (xPr, yPr), and
Sh, Pr, SP, and Pt are two-dimensional vectors.
i = (1, 0),
= cos1(i • SP/ |SP |),
= if ySP < 0,
DP = Rot(SP, ), and
Pt ′ = [(x cos y sin ), (x sin + y cos )] + Sh,
where
Rot[(x, y), ] = [(x cos + y sin ), (x sin + y cos )].
center,whereastheremoteengageabilityMATLABsim-ulationlocatestheshooterinthecenter,withthedataprovideralwayspositioneddueeast.
Rather than redesigning each tool to conform toa display-imposed coordinate system—a process thatwouldhaveinvolvedsubstantialrisk—theCUPositionPlannerselectsthelocationofthedefendedpointasitscenterandconvertssamplepoints,shooters,providers,andotherpositionsfromthatsystemtothecoordinatesystemrequiredbythevariousmodels.
ThecommoncoordinatesystemusedbythePlannerisshowninFig.8.Coordinatesforengageabilitycalcu-lationsforonesamplepointPtforaparticularshooter/providerpairareshowninFig.9.(VectorsfromSh′toPt′andPr′aredefinedintheusualway.)
Forfirmtrackcalculations,theCPAanddownrangeforeachsamplepointinthesampleregionmustbecal-culatedtodetermineifthesamplepointlieswithintheprovider’sfirmtrackregion.CPAanddownrangecalcu-lationsareshowninFig.10.CPAissimplythedistancefromtheprovidertothethreat’strajectory.Downrange
Pt
Pt
Pr
Sh
Sh
SP
Pr
DP
Sampleregion
x
y
PtPr Sh
DP
x
y
Pt
Pt
Pr
Pr
DPx
y
PC
PtCC
C = proj(Pr, Pt ),
PC = C Pr,
CPA = |PC|,
PtC = C Pt,
DR = |PtC|,
S = PtC • (Pt ), and
downrange =
where
proj(A, B) is the projection of A onto B, anddownrange values are positive if the threat is approaching CPA; negative if it has passed CPA.
DR if S ≥ 0,DR if S < 0,⎧⎨⎩
Figure 10. Calculations for CPA and down-range for firm track determination. The CPA is the perpendicular distance from the data provider to the threat’s trajectory. The down-range value indicates the threat’s position relative to CPA. PtC, the vector from the threat’s current position to CPA, is slightly offset for clarity.
Figure 8. Common coordinate system with one engagement sample point Pt.
Figure 9. Coordinate system for engageabil-ity calculations with the shooter at the center and the provider located due east. This system was chosen to match the remote engageability MATLAB model, which was originally designed with threats flying toward the shooter at the center.
208 JOHNSHOPKINSAPLTECHNICALDIGEST,VOLUME23,NUMBERS2and3(2002)
C. J. DUHON
Figure 11. PPI image with sample scenario. CSG, ANZ, HUE, and the defended point are arranged similar to Fig. 2 off the coast of Puerto Rico. The white dots represent sample points that the tool has determined to be engageable.
isthedistancefromthesamplepointtoCPA.Ifdown-range is positive, then the threathasnot yet reachedCPA; ifnegative, the threathas alreadypassedCPA.Thus the calculated downrange value can be directlycomparedtothevaluesinthefirmtrackdatatable.
SAMPLE OUTPUTFigure11showsanimageofaCECdisplayPPIwith
results from a sample CU Position Planner scenario.Therelativepositionsoftheshipsanddefendedpointin this scenarioare similar to those inFig.2,andthebattle group has been placed off the coast of PuertoRico. Threat, SPY, and SM-2 performance data arefictitious in the figure. Each white dot in the image
representsasamplepointthatthetoolhasdeterminedtobeengageable.
FUTURE DIRECTIONSFuture engagement planning tools can build upon
the CU Position Planner by overcoming many of itscurrentlimitations.NewtoolsmightallowforairborneCECplatforms(e.g.,theE-2C)toactasdatadistribu-tionsystemrelaysandtoprovideadvancedtrackcueing.Otheradvancescanbemadeintheareaofgreaterthreatcoverage, inbothbreadthanddepth.Deployedbattlegroupswillrequiretheadditionofmanymorethreatsinthedatabaseaswellasmorerobustandhigher-fidelitydataconcerningthosethreats.
The tool’s current design does not take into consid-eration the self-defense capabilities of CEC ships, or ofothernon-CECshipsinthebattlegroup.Futureworkmayinvolve integrating Planner capabilities with shipboardself-defense engagement capabilities to provide a fullerandmorerobustbattlegroupengageabilityresource.
REFERENCES 1“TheCooperativeEngagementCapability,”Johns Hopkins APL Tech.
Dig.16(4),377–396(1995). 2Dockery,G.D.,andKuttler,J.R.,“AnImprovedImpedance-Bound-
aryAlgorithmforFourierSplit-StepSolutionsoftheParabolicWaveEquation,”IEEE Trans. Ant. Propogat.44(12)(1996).
3Guevara,W.J.,Development of Engage on Remote Engageability Con-tours for Low Elevation,TechnicalMemorandumF3D-6-1309,JHU/APL,Laurel,MD(7Jun1995).
4Common Display Kernel (CDK) Software Design Document Volume I: Top Level Design and Common Software Libraries, Version 3.0,Techni-calMemorandumADS-96-003,JHU/APL,Laurel,MD(Feb1997).
5Nesbitt,D.W.,Graphics Entity Data Format Specification,TechnicalMemorandumF3D-3-1695,JHU/APL,Laurel,MD(17Mar1995).
6Duhon,C.J.,An Object Oriented Application Programmer’s Interface to Graphics Entity Data,TechnicalMemorandumF2C-5-530,JHU/APL,Laurel,MD(24Jan1995).
ACKNOWLEDGMENTS:Thedesign, implementation,and testingof theCUPositionPlannerhaveinvolvedtheworkofmanypeople.SoftwaredevelopmentwasperformedbyKarenCase,TheresaSnider(GUI),DouglasHoffman,KevinHinton(data),MargaretMcGarry,JeffreyPeterson(enhancements),andWilmanGuevara(remoteengageability).DataassemblyandsystemtestingwereperformedbyRichardBourgeois,MarsGralia,DaveRichards,TimothyMeushaw,andJeffRicker.TechnicalguidanceandsystemrequirementswereprovidedbyJerryKrill,DanDockery,DoddHuffaker,VishalGiare,JohnMoore,andDanielSunday.
THE AUTHOR
CHRISTOPHERJ.DUHONisamemberofAPL’sSeniorProfessionalStaffandSupervisor of the Advanced Display Technology and Development Section inADSD’sComputerSystemsDevelopmentGroup.HereceivedaB.S.degreeincom-puterscienceandinmathematicsfromtheUniversityofSouthwesternLouisianain 1984 and an M.S. degree in computer science from Texas A&M Universityin 1990. Mr. Duhon joined APL in 1991 and has been involved in the designanddevelopmentofadvancedcommandandcontrolsoftwaresystemsforthesur-face Navy, including Command Support At-Sea Experiment, Common DisplayKernel,andCEC.HeiscurrentlyleadengineerfortheCECdisplaysystem,work-ingonprototypeuser interfacesandgraphicalvisualization.Hise-mailaddress [email protected].