Notes

Outline

Target Motion Analysis for the Localization of Subsurface Targets Stephen Haptonstahl Northern Illinois University December 3, 1999

Disclaimer or How Different Cultures Say “I don’t know” Politician:  “We have a Congressional committee investigating that issue.” Programmer:  “You can’t do that in Windows.  That only works in UNIX.” Consultant:  “I can provide that information, but it will cost you more.” Military:  “I’d tell you, but it’s classified, so then I’d have to kill you.” Math student:  “We never talked about that in class.” Math Professor:  “That’s beyond the scope of the course.”

The Company:  Your Country US Navy 369,220 sailors in uniform (1 officer/6 enlisted) Who joins the Navy? 316 Ships & Subs – almost half underway Operating in every part of the world Other branches Total # of people Allied forces # of other nations

Target Motion Analysis Strategy: Prevent enemy submarines from getting close enough to destroy your ship Tactic: Keep the sub “occupied” dodging helicopter-launched torpedoes. Problem: Where do you send your helicopter? The Captain wants an answer in 30 minutes!

Describing Location in Maritime Warfare Bearing and Range from ownship  - polar coordinates Bearing (BRG): Compass direction (true, not magnetic) from ownship to target in degrees (“mills” used in gunnery – 6400 mills = 360º) Range (RNG): Distance to target in yards or nautical miles Relative reference frame – must correct for ownship motion to get true (WRT Earth) motion Latitude (N-S) and Longitude (E-W) Geo-fixed reference frame Nautical Mile (NM) Defined to be 1/60 degree latitude (equator to pole:=5400 NM) Equal to about 6000 feet, 2000 yards, or 1.1 statute mile

Primary Sources of Information Active sonar Bearing, range, perhaps depth of target – course and speed Very limited range Counterdetection (perhaps 10X sonar range) Amorous marine life Passive sonar Greater range No counterdetection issues (other that normal) No range information – no course and speed Must use TMA to get range, course, speed

Other Sources of Information Visual – periscopes leave wakes Lookouts (ours or on other ships) (BRG & est. RNG) Pilots (est. lat & lon) Sonobuoys “Yardstick” – range from buoy “Pointer” – bearing from buoy “Cadillac” - both MAD – Magnetic anomaly detector Very short range, but can’t mistake a whale for a sub EW - Reception of their radar or radio emissions – BRG only Intelligence SOSUS – Sound Surveillance System Various classified sources

TMA Team Composition Evaluator South & North Plotters Time/Bearing Plotter Time/Frequency Plotter R/T talker and Sharps Input Sonar/EW/Intel Priorities set by CO Output Location of targets Course/speed recommendations

Line of Sound Line of Sound (LOS): A moving reference line joining ownship and the target

Line of Sound – Evaluator’s Plot Purpose:  Determine the course and speed of the target

DRT & Geo-fixed Plot Recognizing LOS Geometries Input: almost everything Speed strips – get course, speed, range This is where all the information is compiled, where the Captain will look for a picture of what’s going on

Time-Bearing Plot – Range Calculations CPA at graph inflection point convexity determines whether opening or closing Single-leg Ekelund Doesn’t require ownship maneuver Requires an estimated STA Double-leg Ekelund Uses info before and after ownship maneuver Yields accurate range at a time near the maneuver Often target’s relative motion allows this technique Spiess Useful when target has low bearing rate (<1º/min) (not common) Cross-fix using only one ship

Single-Leg Ekelund

Doppler Effect – Time-Frequency Plot Using rt = d, we can determine the perceived change in frequency caused by STI & SOI Sw = Speed of sound in sea water,   »1664 yds/sec SI = STI + SOI fr = received frequency f0 = emitted frequency fcorr = f0 affected only by target motion Plot fr, then calculate SOI to get fcorr Changes in fcorr are caused by Changes in STI caused by shifting LOS geometry Target maneuvers (best way to detect target maneuvers)

Applying the Doppler Formula Assume ownship fixed, or correct for SOI fcorr increases as STI does

Line of Sound Determination

What If We Know the Target’s Speed? Sources Blade count + ID of class = speed Intelligence “We believe a Kilo is transiting from Murmansk to Cuba over x days, so expect a minimum speed of y knots.” Geo-fixed plot (speed strips; lead geometry) What we get If we have max(fc) (perhaps a natural transition from overlead to lag) then we can get f0 Evaluator can improve LOS diagram to better estimate course Geo-fixed plot can accurately fix strips to get course and range

What If We Know the Emitted Frequency (f0)? Sources Inflection point of fc “Crazy Ivan” (like in Hunt for Red October): Target turns through 360º to check for contacts in his baffles (wake).  We get f0 halfway between max(fcorr) and min(fcorr).  Also get contact speed. We get Very accurate course Warren (freq) range

Water is Thicker than Vacuum Convergence Zones Sound moves along paths of least resistance Salinity, temperature and pressure all change with depth and affect sound propagation Balance struck is a set of distinct solutions, each a path

The Layer The sharp temperature gradient at the layer causes most sound to be reflected

Technology on the Horizon Expert systems – AI based TMA Can we do it? Is it a good idea? Bottom bounce Multiple instances of the same sound coming in at slightly different times from different angles Ambient noise We see with ambient light, why not apply this idea to sonar? Improved active sonar has reduced counterdetection range

Advanced Techniques and Further Questions Tactics What are good maneuvers to recommend that will: Maximize information on the target Minimize counterdetection Zigzag plans EMCON How do we respond to target maneuvers? What’s the best we can do with these formulas?  Can we get more from less?