mopdf.com

Overview of Platforms and Combat SystemsHard killCATCSoft killControlUSS NimitzAMD capabilitiestoday*– AN/SPS-48G*– AN/SPS-49A*– AN/SPQ-9B*– Mk 9 T/I (4)*– RAM– ESSM– CIWS (3)– AN/SPN-43*– AN/SLQ-32A(V)4– AN/UPX-29– SSDS– CEC– TDLRapid, incrementalimprovements to pacethe threat via theFire Control LoopImprovement Program(on an as-needed basis)elinse deBa graupNetworksUSS Gerald R. Ford(CVN 78) 2021Capability trade-offs andengineering challengesfor the future– Advanced onboard electronic attack– Advanced offboard electronic attack– Next-generation weapons schedulingalgorithms– Hard kill/soft kill integration– Next-generation sensor netting andintegrated fire control– Low-cost multifunction X-band radar– Resilient, cyber-resistant combat system– Next-generation combat information center– Next-generation close-in weapon systemMultifunction radar (DBR)AN/SLQ-32(V)6/SKCHard killSoft killModeling, simulation,and critical experimentsto select most capabilityat affordable cost* This function will be replacedby the multifunction radar.Figure 3. Examples of planned aircraft carrier SSDS combat system evolution and potential capability trade-offs. (Amphibious ships,which also have the SSDS combat system, are also evolving with related improvements and capability trade-offs.) CATC, carrier air trafficcontrol.onboard electronic attack and an improved inventoryof decoys.The SSDS-based combat system on aircraft carriersand amphibious ships has historically relied on asuite of older sensors (some initially designed in the1960s) that have undergone periodic modernizations.Radar surveillance and target tracking are providedby the AN/SPS-48G, AN/SPS-49A, and AN/SPQ-9Bradars. Additional surveillance and tracking as well asillumination for semiactive missile homing are providedby the Mk 9 fire control system. Carrier air traffic controlis supported by the SPN-43. With the new aircraft carrierUSS Gerald R. Ford (CVN 78), these functions will bereplaced by the new Dual-Band Radar (DBR) (Figure 3).This new multifunction radar being developed for theCVN 78 is a combination of the X-band AN/SPY-3 andS-band AN/SPY-4. However, alternative radar designsare being considered for subsequent aircraft carriersCVN 79 and CVN 80 as well as for new amphibiousships. The multifunction radar will accomplish thelong-range surveillance and track functions of theAN/SPS-48 and AN/SPS-49 radars, provide data forcarrier air traffic control (currently provided by theAN/SPN-43), and provide the horizon surveillance andtracking capability of the SPQ-9B radar and the firecontrol functions of the Mk 9 tracker/illuminator. Themultifunction radar will enable better control of ESSMmissile trajectories and more accurate handover to theESSM seeker, improving ESSM capability against antiship cruise missiles.Selecting the most capability at affordable cost is achallenge in development of any new combat systembaseline. Figures 2 and 3 show candidate systems andcapabilities for future baselines of Aegis and SSDS,respectively. APL performs modeling and simulation andcritical experiments to inform the selection of an affordable subset of these systems and capabilities for newbaselines. In addition to the major baseline upgrades,the Navy continues to explore techniques for deployingnew capabilities rapidly on an as-needed basis. Aegis andJohns Hopkins APL Technical Digest, Volume 35, Number 2 (2020), www.jhuapl.edu/techdigest93

W. G. BathEndo-missilesAlgorithms,computing plant, and codeWarfighter interface/training/planning Target identification,typing, anddiscrimination Weapon selectionand scheduling Sensor dataassociation,fusion, andtrack management Engagementdecision– Missiles/guns– Electronicwarfare Soft-killcoordination Weapons control– Hard kill– Soft killWeaponsElectronic warfare Sensorprocessingand controlGunsSensorsOffboard sensors(through networks)Missile launchersExo-missilesShipboard radars/IFF(either organic or through networks)Combat information centersNetworkingCommunicationsWeapon control linksIlluminatorsFigure 4. A general combat system. Actual combat systems have a subset of the components pictured. Successful engagementsrequire coordinated operation of many combat system components. IFF, identification friend or foe.SSDS use the Aegis Speed to Capability and the FireControl Loop Improvement Program, respectively, torespond to urgent needs in the fleet.Air and missile defense is a complex process involvingthe coordinated operation of equipment and computerprograms. Figure 4 shows a general ship combat system.The workhorses of sensing on a ship are its shipboardradars—particularly the multifunction radars. Theseradars are augmented by other shipboard radars servingspecific purposes. In addition, ships can access offboardsensors located on other ships, aircraft, land sites,and space via secure communications. Sensors arecontrolled and sensor resource use is managed to meetthe overall defense needs. Individual measurementsmade by the entire sensor set are associated, and in someinstances fused, with other sensor data. In all cases,tracks are generated. Each track should correspondto one physical object. A track is the combat system’ssum total knowledge of an individual object, includingits kinematics—e.g., vector position and velocity; theclassification of the object (aircraft, cruise missile,ballistic missile, clutter, debris, etc.); the type of theobject (e.g., if it is a cruise missile, which cruise missiletype is it); and when applicable, the identity of theobject (e.g., friend or foe).Figure 5 illustrates the association and tracking problem. In any part of the world on any given day, there94 is generally a priori context information available tothe war fighter. This context will define who the likelyenemy is, what sort of threats he has in his inventory,and, in general terms, how he is likely to attack. Withintoday’s combat systems, this information is held as “doctrine,” a collection of rules that define how the combatsystem will respond to sensor information. For example,today’s identification doctrine defines, given the context, which additional pieces of sensor evidence are necessary to conclusively identify the target. The next likelyinput to the combat system is some early indication fromISR (intelligence, surveillance, and reconnaissance)that an attack is coming; this early indication alerts thecombat system to the object’s presence and often identifies the object, but it does not necessarily provide precisekinematics or low latency. Today, there is little quantitative integration of contextual and ISR data with organicsensors. The quantitative integration of a priori contextand ISR is a challenge and growth area for new combatsystem designs.Once targets are within sensor range, the combatsystem receives sensor measurements (e.g., onboard oroffboard radar) indicating more precise kinematics atlow latency, but these data may or may not include features for identifying the object. One of the challengesis to correctly associate all of these pieces of data into“tracks.” As measurements are associated to form tracks,Johns Hopkins APL Technical Digest, Volume 35, Number 2 (2020), www.jhuapl.edu/techdigest

Overview of Platforms and Combat SystemsXYA priori context: predicting likely object locations,types, and behaviorISR indicating object presence and often identity,but not necessarily with precise kinematics or lowlatencySensor no. 1 measurements (e.g., onboard oroffboard radar) indicating more precise kinematicsat low latency, but may or may not include featuresfor identifying the objectSensor no. 2 measurements (e.g., onboard oroffboard radar) indicating more precise kinematicsat low latency, but may or may not include featuresfor identifying the objectObject position (ground truth)sensor networks such as the CEC. The principal challenge is the diversity of the data received and the need tomake one unit’s track management process interoperable with multiple units’ track management processes. Forexample, each source will generally have a different wayof characterizing the accuracy of the kinematic trackdata, and some sources may provide incomplete characterizations. Similar diversity exists in the characterization of target identity and type. Different units designedin different time frames and with different missions willhave different rules and algorithms for supporting thecreation of a common track numbering and identification system. In addition, the networks may deliver datawith different time delays, biases, and data dropouts.The process by which all these sources are reconciledinto a single usable track picture is generally called trackmanagement and has been an active area of researchand development at APL for many years.A good example of the metrics for the single trackpicture is given by the Single Integrated Air PictureMetrics1 developed by the joint services (Army, Navy,Air Force, and Marine Corps) for air track (vice ballistic missile) tracking (Figure 7). Note that the metricscover both track kinematics and attributes. In addition,Figure 5. A two-dimensional representation (x, y) of a multi dimensional tracking problem. In this example, three targetsare close enough together to challenge association and filteringalgorithms.the track kinematic state is calculated (and used forsubsequent associations). Track filtering refers to thealgorithms that transform a sequence of measurementsinto such a track state and is discussed in the article byS. A. Hays and M. A. Fatemi in this issue. Figure 6 showsnotional track states that have been calculated by associating and filtering the measurements in Figure 5. Inthis illustration, the tracking process has worked well.The number of tracks in Figure 6 equals the number ofobjects, the track states converge over time to the actualobject positions, measurements from different sensorshave been associated correctly, and the tracks can beextrapolated into the future to accurately predict targetposition. However, the tracking process can be challenged in all these areas by large sensor measurementaccuracies, low sensor update rates, highly unpredictableobject motion, and object spacing. In the case of multiple sensors, measurement biases and different sensormeasurement dimensions are also challenges. Overcoming these challenges remains a subject of research incombat system design.A combat system must merge, fuse, and deconflictmany sources of track data to produce a single usabletrack picture for decision-making. This includes all localsensors as well as track data from tactical data links suchas Link 16/11 and measurement and track data fromXYA priori context: predicting likely object locations,types, and behaviorISR indicating object presence and often identity,but not necessarily with precise kinematics or lowlatencySensor no. 1 measurements (e.g., onboard oroffboard radar) indicating more precise kinematicsat low latency, but may or may not include featuresfor identifying the objectSensor no. 2 measurements (e.g., onboard oroffboard radar) indicating more precise kinematicsat low latency, but may or may not include featuresfor identifying the objectObject position (ground truth)Object track calculated in combat system usingcontext, ISR, and sensor measurementsFigure 6. The combat system calculates tracks representing abest estimate of the object kinematics. This figure depicts quantitative integration of contextual and ISR data with organic sensortracking—a challenge in the design of new combat systems.Johns Hopkins APL Technical Digest, Volume 35, Number 2 (2020), www.jhuapl.edu/techdigest95

W. G. BathCombat systems are designedusing error budgets for their critical functions. These budgetsidentify the maximum errors thateach combat system function cantolerate, and they allocate a portion of that maximum error toeach of many contributing factors.Kinematic track state errors aregenerally significant contributorsto the maximum error. To meetchallenging error budget allocations, most combat systems filtermeasurement data differently fordifferent combat system functions (Figure 9). One example ofthese differences is the degree offiltering. Heavier filtering (smallerfilter gains) will weight new mea-96 Prevents confusion leadingto delays and “waste” ofengagement resourcesPreserve short reaction time:prevent delays due toreidentifying targets and/orrestarting engagementsEnables resource coordination:– Shooter to shooter– Shooter to provider– Tracker to trackRange to zinterceptCloser range(a)Threat 1Threat 2Threat 3Smaller errors(b)Errorsin targetkinematicsThreshold neededfor successfulengagementMorecertain (c)Certaintyin targetidentity andcharacteristics(d)SensorresourceuseThreshold neededfor successfulengagementMore certain Determination of acceptableweapon launch times andintercept points (scheduling)Tracks on all targets availablefor engagementFigure 7. Typical metrics for the combat system air track picture.1 Determination of intent Decision to engageImportance of metricsTrack management metricsCompleteness: The air picture is completewhen all objects are detected, tracked, andreported.Clarity: The air picture is clear when it doesnot include ambiguous or spurious tracks.Continuity: The air picture is continuouswhen the tracks are long-lived and stable.Kinematic accuracy: The air picture iskinematically accurate when the positionand velocity of a track agrees with theposition and velocity of the associatedobject.ID completeness: The ID is complete whenall tracked objects are labeled in a stateother than unknown.ID accuracy: The ID is accurate when alltracked objects are labeled correctly.ID clarity: The ID is clear when a trackedobject has no conflicting ID states.Commonality: The air picture is commonwhen the tracks held by each participanthave the same track number, position,and ID.More resource usethe metrics measure the degree ofcommonality between the trackpictures on different ships and aircraft. This commonality is essential for sharing of engagement andidentification data.Once tracks exist, they becomethe organizing tool for the engagement sequence. The success of theengagement depends on the fidelity of the track on the target beingengaged. As the target closes inrange to its objective (Figure 8a),more sensor measurements aremade, resulting in continualimprovement (Figure 8b) in theaccuracy of the track kinematics(e.g., position, velocity, and acceleration) and in the certainty intarget identity and characteristics(Figure 8c). However, most weapons require that additional sensorresources (e.g., different radarwaveforms, higher update rates,high priority in radar scheduling,or in some instances, additionalsensors) be applied to achievea “fire-control-quality track”(Figure 8d) capable of supportingall or part of the following:Fire-controlquality ial detectiontimesFigure 8. A typical engagement sequence against a raid of threats. Sensor data are gathered beginning as soon as a target is detected, eventually leading to sufficient kinematicaccuracy and certainty in identity and characteristics for a successful engagement. Sensorresource needs to meet these thresholds vary throughout the engagement.Johns Hopkins APL Technical Digest, Volume 35, Number 2 (2020), www.jhuapl.edu/techdigest

Overview of Platforms and Combat SystemsThe next part of theengagement decision is todeterminewhichweapons(missiles, guns, and/or electronicwarfare) have the capability tonegate the threat and to selectthe weapons to be used basedMedium target (maneuvers)on their inventory and predictedeffectiveness. The engagementHigh target (maneuvers)decision is one of the factors thatdrives the need for precise trackkinematics because the trackkinematic state may need to bepredicted well into the future(e.g., to account for the fly-outtime of a ship-launched missile toa predicted intercept Tracking filtering gainsHeavier filteringLighter filteringscheduling is the concept ofFilter design region for combatFilter design region for combat“depth of fire”—multiple layerssystem functions such as:system functions such as:of defense in range. The most Measurement/track association Engagement decisioneffective defense generally is Seeker handover Intercept point predictionmultiple layers of defense using Illuminator pointing Weapon and resource schedulingdifferent technologies in eachlayer such as long-range hardFigure 9. To meet engagement error budgets, most combat systems process sensor meakill (e.g., naval integrated firesurement data differently for different combat system functions. For example, heavier trackcontrol), followed by hard-killfiltering (smaller filter gains) will produce kinematic estimates with a smaller variance duearea defense, followed by bothto measurement noise and enable longer-term time prediction. Lighter track filtering willhard-kill and soft-kill (electronicproduce kinematic estimates more tolerant of unpredictable target motion (maneuvers).warfare) self-defense. The ballisticmissile defense analogue of thissurements less relative to the current track state and prowould be midcourse defense followed by sea-basedduce kinematic estimates with a smaller variance due toterminal defense.measurement noise. However, these filters are not veryConsider air defense as an example. Assume a raid oftolerant of unpredictable target motion (maneuvers).NT threats. A typical measure of air defense performanceLighter filtering (larger filter gains) will weight new meais the probability of raid annihilation (PRA). For NL layerssurements more relative to the current track state andof defense, each with a probability of killing the targetproduce kinematic estimates with a larger variance due(PK), the mathematical advantage of depth of fire canto measurement noise. Although these filters are morebe easily demonstrated in a very simplified analysis. Totolerant of unpredictable target motion (maneuvers),annihilate the entire raid, each of the NT targets must betheir variances makes them less desirable for functionskilled, and there are NL opportunities to kill each target.that require long-term time prediction.The simplified analysis assumes that all these events areOnce tracks exist, they need to be characterized as tostatistically independent, in which case, PRA is given bytheir type and identity. Is this target a threat attackingPRA [1 – (1 – PK)NL]NT.a defended area (which should be engaged) or anotherThis equation is plotted in Figure 10 for a raid sizeobject such as a commercial airliner or a nonlethal pieceof five threats. Note that achieving a very high levelof debris (which should not be engaged)? In addition,of defense (high PRA) with only a single layer requiresthe greater number of target characteristics that can bea very high probability of kill in that layer. That highknown (e.g., type of threat), the more effective the engageprobability of kill can be difficult to achieve with ament can be. Both determination of type and determinasingle technology (e.g., with a single missile type ortion of identity generally require dedication of additionalsingle electronic warfare strategy) because any defensivesensor resources to achieve the confidence necessary fortechnology has weaknesses that could be exploited bya successful engagement. The identity and characteristicsthe adversary. A layered defense using different techof the track, as well as its kinematics, are compared withnologies in each layer requires a relatively lower proboperational doctrine to make the engagement decision.ability of kill in each layer and generally makes a highKinematic track state errorsErrors in long-term prediction (low target maneuvers)Errors in short-term prediction (medium/high targetmaneuvers)Point where errors are minimized by gain selectionJohns Hopkins APL Technical Digest, Volume 35, Number 2 (2020), www.jhuapl.edu/techdigest97

W. G. Bath1.0Depth of fire 30.8Depth of fire 2PRA0.60.40.200.5Depth of fire 10.60.7PK0.80.91.0Figure 10. Multiple layers of defense using different technologies generally make a high PRA more achievable. This graphdepicts the results of simplified analysis for a raid of five threatsand the assumption that all engagements are statisticallyindependent.PRA more achievable. The different technologies usedin different layers make it more likely that the statisticalindependence assumption is valid and thus more likelythat the gains from multiple layers occur. As a result,most ships have a mixture of both hard-kill and soft-killdefensive technologies as well as different types of hardkill and soft-kill weapons.In addition, depth of fire can conserve inventory. Ifthe engagement is successful in the first layer, and if thatsuccess can be confidently measured, then the resourcesrequired for the subsequent layers do not need to beexpended on that threat.As the engagement proceeds, more combat systemresources of are generally required for success (Figure 8d).A significant challenge in combat system design isdeciding which of these weapons to employ when andhow to schedule combat system resources (e.g., sensors,launchers, illuminators) to accomplish the engagements.The sch

measures such as jamming and decoys. The Cooperative Engagement Capability (CEC) and Tactical Data Link (TDL) networks enable Aegis and other units to fight as a coordinated force. The USS Zumwalt (DDG 1000) brings to the Navy a unique set of volume firepower and precision str