NAVAL POSTGRADUATE SCHOOL

MONTEREY, CALIFORNIA

AD-A281 747

DTIC.
THESIS ELECTE
S JUL 2 01994D1

G
A Limited Analysis of some Nonacoustic Antisubmarine
Warfare Systems

94-22178 by

Daniel Gerald Daly

March 1994

Thesis Advisor: M.D. Kite

Co-Advisor: R.N. Forrest

Approved for public release; distribution is unlimited

94 7 14 0, 9
 REPORT DOCUMENTATION PAGE Form approved OM B No. 0704-188

Pbic reporting burden for this collection of i ormation is estimated to average I hour per response. incluing the time for rev iewing
instructions, searching existing data sources. gathering and maintaining the data needed, and completing and rev.iewing the collection Of
information. Send comments regarding this burden estimate or any ocher aspect of this collection of information including suggestions for
reducing this burden, to Washington Headquarters services. Diretorate for Information operations and Reports, 1215 Jefferson Da%is High%%a
suite I1209. Arlington. %'A22202-4302-, and to the Office of Management and Budget. Paperwork Reduction Project (0704-0188). Washington.
Dc 20.%03
1. AGENCY USE ONLY (1AwcBhank) 2. REPORT DATE 13. REPORT TYPE AND DATES COVERED

I March 1994 Master's Thesis

4. TITLE AND SUBTITLE 5. FUNDING NUMBERS

A Limited Analysis of sonme Nonacoustic Antisubmarine Warfare Systems

6. AUTHOR(Sj
Daly, Daniel G.

7. PERFORMING; ORG ANIZATION NAME(S) AND ADDRESS(ES) & PERFORMING ORG;ANIZATION

REPORT NUMBER
Naval Postgraduate School
Moniterey, CA 939-13-5000

9. SPONSOR ING/MONITORI NG AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSORING/MONITORING

NUMBER
IAGENCY REPORT

11. SUPPLEMENTARY NOTES

The views expressed in this thesis are those of the author and do not reflect the official policy or posi,,on of the Department of

Defense, the I,*S. Government. the Canadian Department of National Defence, or the Canadian Government

12a. DISTRIBUTION/AVAILABILITY STATEMENT 12b. DISTRIBUTION CODE

Approved for public release, distribution is unlimited.

13. ABSTRACT (Maximumn 200 words)

The problem of AntiSubmarine Warfare early in the next century is examined. Nonacoustic detection method%including
magnetic anomaly detection, laser radar, and hydrodynamic detection are examined. A simple analysis of their
relative effectivenessi is made.

14. SUBJECT TERMS 15. NUMBER OF PAGES

Anti-Submarine Warfare, Submarine Detection, non-acoustics. MA-D, Laser Detection of 16. PRICE CODE
Stibmarines

17. SECURITY CLASSIFIJ-18&SECURITY CLASSIFI- 19. SECURITY CLASSIFI- 20. LIMITATION OF

CATION OF REPORT CATION OF THIS CATION OF THIS ABSTRACT
PAGE ABSTRACT
Unlsiid Unclassified Unclassified l

NSN 7540-0I-280-5500) Standard Form 29S (Re%. 2-89)

Prescribed b%ANSI Std 2.39-18S
 Approved for public release; distribution is unlimited.

A Limited Analysis of some Nonacoustic Antisubmarine Warfare Systems

by
Daniel Gerald Dalv
Major, Canadian Air Force
B.Sc., Roval Roads Military College, Victoria, 1981

Submitted in partial fulfillment of the

requirements for the degree of
MASTER OF SCIENCE IN APPLIED SCIENCE
from the

NAVAL POSTGRADUATE SCHOOL

March 1994

Author:__ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

Daniel G rald aly

Approved by:

T sAdio
 ABSTRACT

The problem of Anti-Submarine Warfare early in the next century is examined.

Nonacoustic detection methods including magnetic anomaly detection, laser radar, and

hydrodynamic detection are examined. A simple analysis of their relative effectiveness is

made.
Accesion For
NTIS CRA&I
DTIC TAB
Unannounced 0
Justification

By .....................
Distribution/
Availability Codes
Avail and ior
Dist Special

iii
 TABLE OF CONTENTS

INTRODUCTION .............................................................................................. 1
A. I
BACKGROUND .............................................
B. OBJECTIVES ................................................................................... 3
C. SC O PE .............................................................................................. 3
D. ORGANIZATION ............................................................................ 3
II. THE THREAT ............................................................................................. 5
A. SUBMARINE TYPES ..................................................................... 5
1. Nuclear Submarines .............................................................. 5
2. Conventionally Powered Submarines ............................... 6
B. AIR INDEPENDENT PROPULSION DEVELOPMENTS .......... 8
1. Closed Cycle Diesel Engines .............................................. 8
2. Fuel Cells .............................................................................. 9
3. Stirling Engines ..................................................................... 10
4. Small Nuclear Reactors ......................................................... II
C. SUBMARINE "SIGNATURES ................................................... 12
D. SCENARIOS .................................................................................... 14
1. Conventional Threat ........................................................... 15
2. Nuclear Threat ..................................................................... 15
HI. SOME NON-ACOUSTIC DETECTION SYSTEMS ................................ 17
A. MAGNETIC ANOMALY DETECTION (MAD) .......................... 17
1. MAD encounter simulation ................................................. 20
2. Random Search Using MAD .............................................. 22

i'"
 B. HYDRODYNAMICS ..................................................................... 22
I. Bernoulli Hump ................................................................... 23
2. Kelvin Wake ........................................................................ 24
3. Vortex Shedding ................................................................... 25
4. Sweep Width Using Wake Detection ............................... 27
5. Random Search Using Wake Detection ........................... 30
C. LASER DETECTION SYSTEM .................................................... 30
1. Overview of Laser Radar (Lidar) ...................................... 30
2. Random Search Using Lidar ............................................... 35
D. ELECTROMAGNETIC SUPPORT MEASURES (ESM) .......... 35
APPENDIX A - COMPUTING ADVANCES ...................................................... 37
APPENDIX B - WAKE DETECTION MATLAB PROGRAM ............................... 41
LIST OF REFERENCES ......................................................................................... 43
INITIAL DISTRIBUTION LIST ............................................................................. 45
 LIST OF FIGURES

Figure I - Typical Pressurized water marine reactor ............................................ 6

Figure 2 - German Type 209 Diesel Submarine .................................................... 7
Figure 3 - Closed-cycle diesel engine ................................................................... 9
Figure 4 - Fuel cell ................................................................................................... 10
Figure 5 - Stirling Engine ............................................................................................. II
Figure 6 - 'Slowpoke' Nuclear reactor .................................................................. 12
Figure 7 - Submarine Signature ............................................................................ 13
Figure 8 - Active Sensors ......................................................................................... 13
Figure 9 - Magnetic noise in an ASQ-10 MAD system due to a moderate
magnetic storm, compared with a submarine-generated
signal ..................................................................................................... 19
Figure 10 - Lateral Range versus probability of detection at CPA for a
straight line encounter between a MAD sensor and a
subm arine ............................................................................................ 21
Figure I I - Shape of far-field surface disturbance (Kelvin wake) over a
moving submarine (Dimensions of the Ohio-class) ........................ 24
Figure 12 - Encounter geometry in the plane of the wake for a straight
line encounter between a wake and a detection system's
sensor ................................................................................................... 28
Figure 13 - Sweep width versus wake length with a fixed detection range
of 100 meters ...................................................................................... 29

\'i
 I - I

Figure 14 -Sweep width versus wake length with a fixed detection range
of 1000 meters .................................................................................... 29
Figure 15 - Absorption coefficient (m- 1 ) versus wavelength (nanometers) .......... 31
Figure 16 - laser return versus time ....................................................................... 32
Figure 17 - Areas in Lidar formulae ........................................................................ 32
Figure 18 - Theoretical lidar return from a submarine, clear water ..................... 34
Figure 19 - Laser return from a submarine, turbid water ..................................... 35

"ii
 ACKNOWLEDGMENTS

My thanks to Professors Kite and Forrest for agreeing to work on this topic with me.
The insights I have developed during this thesis research have proved very enlightening.
I am indebted to the many fine faculty members who taught me during my time in the
United States. I would also like to thank my classmates of Antisubmarine Warfare section
IX-23 for their support during a difficult time in my life, and for reminding me that there is

life outside of schxl.

'.iii
 I. INTRODUCTION

A. BACKGROUND
It may seem difficult to justify any spending on Anti-Submarine Warfare (ASW),
especially in a tight budgetary climate now the Cold War has been "won." Why use
scarce resources on what may seem like fighting the last war again? Yet, although the
threat of submarine forces to the re-supply of Europe is no longer credible, the threat of
submarines to other activities has increased, and the world has become less predicable,
and to some, more dangerous as a result.

Submarine building and export, far from fading in the post-Cold War era, is in fact

a growth industr'. The submarine by its covert nature remains a potent and desirable
tool in the political arsenal. Benedict compares the military sales in the time frame 1982-
85. and 1986-89 in constant 1989 dollars, and notes they decreased by 30-60 percent

for most major weapon systems, except for submarine deliveries, which increased by
30 percent [Ref. 11. Consider the following excerpt from the United States Naval
Institute Proceedings:

In 1992, export sales accounted for 3 2 7c of Russian military production. The

industrial lobby is currently trying to push up the quotas so that they would be
allowed to sell 40 to 50% cf their production overseas [Ref. 21.

The vessels sold by the former Soviet Union for export were not always first-
class. For Russian sales, this has changed:

In the past, the Russians-concerned that their secrets would be compromised-

were reluctant to export their most advanced weapons. They might fall into
Western hands, or their combat use by export clients might provide Western
intelligence with an understanding of their operation. Ineither case, their future
effectiveness in Soviet hands would have been degraded by the development of
Western countermeasures. With the end of the Cold War, this disincentive has
disappeared [Ref. 2 :p.391.

I
 For example, Red Star, on 16 October 1992 carried an "ad'crtisement" for a

TANGO-class diesel boat [Ref. 31. Thus, so-called Third-World powers' may be able

to acquire state-of-the-art submarines.

Increased automation in the latest export submarines means that their small crews 2

need not be as well-trained in order to operate them. Also, the reductions in military
forces world-wide have left a large number of professional submarine sailors, both

nuclear and conventional, looking for employment. Consequently, it is not

inconceivable that one or several submarines could be crewed by mercenaries. In
addition, not all societies (or leaders of them) share the same reverence for life that
Western societies do, and the idea of a submarine crew accepting martyrdom in a
political statement, whether it be for sinking a Western combatant or a cruise ship,
cannot be discounted.
The ASW threat is not going away, but may be growing as more capable platforms

come into less "friendly," less stable hands. Concurrently, economic realities suggest
that Canada's military budget will continue to decrease, leaving few resources to invest
in new or technologically risky ASW systems. Why is ASW research of immediate

concern? Not only are diesel-electric submarines poised to make an order of magnitude
jump in capability, but it is generally accepted by ASW professionals that current
systems developed for deep water ASW do not adequately address the present and
future shallow-water problem. It is therefore critical to examine current ASW systems
and ASW systems under development to see how best to counter this threat.

1
Third-World: "Term thai refers to the > 100 developing countries of Africa, Asia. Latin America, and
Oceania Typically.former colonies with traditional cultures, agrarian economies, high birthrates, and
widespread poverty. Term..coined after World War II to characterize an emerging group of nations that
did not align themselves politically with either of the two powerful groups of industrialized nations:
The Capitalist World or the Communist Bloc." [Ref. 41

2 l-xample: The TR-1700 diesel-electric submarine has a crew of only 26. [Ref. 51

2
B. OBJECTIVES
This thesis examines the problem of Anti-Submarine Warfare, particularly in

shallow water. In the near term, defined as the next ten years, submarines in serv ice
and currently building will be the threat.

To address the near-term threat, some current nonacoustic ASW systems are
examined for their applicability to shallow-water ASW, either as-is or %Ith

modifications. New submarine technology is examined for its effect on the ASW
problem.
Finally, a summary is presented that will attempt to answer the following

questions: In order to maintain an adequate level of capability, which nonacoustic ASW

systems have minimal room for improvement and which systems should continue to be
funded?

C. SCOPE
The research for this thesis was conducted at the unclassified level. The physics
dealing with the ASW problem is, of course, not classified. Classification usually
occurs when dealing with data gained from specific systems, for example, acoustic
source levels from a submarine class, or vessel within the class. No such information

was used in this thesis. All data concerning characteristics of performance, physical
size, and equipment fit were gained from unclassified literature, primarily Jane's

Fighting Ships. The conclusions drawn from the research are the author's, and do not
necessarily represent those of the Canadian Air Force, the Canadian Department of
National Defence, the United States Navy, or the United States Department of Defense.

D. ORGANIZATION

The first part of this thesis presents scenarios which are plausible for use of
submarines against the Canadian Forces or their allies, provides a resumd of some of
the current submarines, details implications of pending technological advances, and

3
discusses signatures which will be exploitable in the timc-frame discussed above. The

second part co%ers non-acoustic methods of submarine detection.

Appendix A examines the computing power that can be expected to be available in

future ASW systems, and Appendix B is the listing of a MATLAB program used in
Chapter Three.

4
 !1. THE THREAT

A. SUBMARINE TYPES
I. Nuclear Submarines
The navics of the United States, United Kingdom, the ex-USSR (Russia and

Ukraine), France, People's Republic of China (PRC), and India operate nuclear pol:.-crcd
submarines. 'The nuclear-powered submarines provide long submerged endurance,
considerable speed, and, in general, more volume for weapons and stores than do

conventionally powered submarines.

In current nuclear submarines, water is heated by passing it through the core of a
reactor in a primary loop under high pressure, thus the appellation pressurized water reactor
(PWR). In this process, contaminants in the water can become highly radioactive. Heat
from the water is transferred to a lower-pressure secondary loop (which is not made
radioactive because no mixing takes place), boiling the water. The resulting steam turns a
turbine, which runs a generator, and, through reduction gearing, turns a propeller shaft.
Some submarines may use generators to run electric motors which turn the shaft; this

avoids noise from the gearing being produced. Some experimentation has been done with
liquid-metal primary loops, but water remains the current system of choice. Nuclear
submarines have essentially unlimited fuel endurance. A number of nations have been
interested in possessing nuclear submarines, including Brazil, Canada, India, Spain,
Pakistan, Turkey, and Argentina [Ref. 6]. The main problems with nuclear propulsion

systems are that they are complex, requiring extensive engineering training and,
additionally, they require considerable infrastructure ashore. A secondary problem,
radioactivity, is becoming increasingly unpopular with many segments of society, as well

5
as being lethal to the crew (and environment) if handled improperly. Figure 1 illustrates a

typical marine nuclear reactor (from Stefanick page 139).

MAIN |hNGIINf 7ROT-LE

I
P ESSU RISER STEAM GENERAT OR RIOJCTIO" GEARING

COW'OIL TURBOGONERAO1 MA
IN TJISINE CLUTCm
ROD
MdOTORS EiLECTROCAL

PRIMAqR SsCO.f.oAV IATTERy

O&ACTOM
COOLING COO.ING
MA N COOANT DJMD LOOP LOOP MAIN CONDENSER
SIELoEO SULK.EAO CONDENSER MOTOR GENERATOR

Figure I - Typical Pressurized water marine reactor

2. Conventionally Powered Submarines
Conventional submarines are operated by the nations mentioned in the
introduction to nuclear power (save the United States, and very recently the United
Kingdom), most of the NATO allies, and 20 Third World nations. Conventionally powered
submarines are best suited for patrolling when great mobility is not required. A modern
conventional submarine operating submerged on battery is very quiet acoustically, and
virtually undetectable. As of I January, 1990, there were 408 operational submarines in 41

countries. Of these, about 200 belonged to Third World nations [Ref. 71. Germany, the
United Kingdom, France, Sweden , the Netherlands, possibly in the near future Australia,

Russia, China, and North Korea export conventionally powered submarines [Ref. 8). The
following sections will briefly describe their operation.

6
 Modem conventionally powered submarines are generally powered by diesel-
electric propulsion systems. Diesel engines are used on the surface or with a snorkel to

charge batteries which are used to power electric motors for propulsion underwater. While
underwater, oxygen is stored in high pressure tanks, and released slowly to sustain the

crew. Carbon monoxide, and other undesirable chemicals, are scrubbed from the air with
chemicals if air cannot be exchanged via the snorkel. Such submarines are very quiet on
battery operation, but they must operate on or near the surface frequently in order to
recharge batteries and replenish their breathing air supply, rendering them vulnerable to
detection by acoustic (engine sounds) and non-acoustic (visual, radar, etc.) means.
Modem diesel-electric submarines are very capable. An Argentine TR-1700 type
submarine, Santa Cruz, was reported to have transited 6,900 nautical miles (NM) from the

Bay of Biscay, 600 NM on the surface, the remainder submerged, to Mar del Plata in 'Q
days. The 6,300 mile submerged run was completed at an average speed of 10 knots
snorkeling an average time of 2 hours a day. She arrived with a fuel reserve of
50 percent. [Ref. 91
Figure 2 illustrates a typical diesel electric submarine (from Benedict "Third
World Submarine Developments").

Figure 2 - German Type 209 Diesel Submarine

7
B. AIR INDEPENDENT PROPULSION DEVELOPMENTS
More recently, advances in several new technologies have drastically changed the anti-
submarine conventionally powered picture, particularly Air Independent Propulsion (AlP)

systems. AlP systems free the conventional submarine from surfacing to recharge batteries.

The four major systems are Closed-cycle Diesel Engines, Fuel Cells, Stirling Engines, and
Low-power Nuclear Reactors. [Ref. 101

AIP is used in combination with diesel-electric technology, using AIP for patrol, and
diesel-electric power when fast transit is required. Redundant systems also give increased
safety for the submarine. Each of the AIP technologies is briefly described below:
1. Closed Cycle Diesel Engines
Recharging batteries with a diesel engine driving a generator has, in the past,
required being on the surface, or with a snorkel mast near the surface. Even if a supply of

air were available at depth, the requirement to discharge exhaust against great pressure has
also robbed the engine of its usefulness. This has changed with the closed-cycle diesel
engine, which recycles exhaust gas to the engine intake and adds oxygen to maintain a

content of approximately 21 percent. The remaining 79 percent of inert gas is made up by

adding argon gas, which provides the right volume and density for the compression heating
and firing stroke of the diesel. Sea water is used to absorb exhaust products from the
engine, and the exhaust-filled sea water is exchanged for clean sea water. The inter-
relationships between argon, oxygen, and flow rate of exhaust-absorbing sea wrater are

controlled by a microprocessor.
This new process uses only 10 to 15 percent of the engine's energy leaving

sufficient energy to propel the ship and charge batteries. Closed-cyclc diesel engines ha' e
also been proposed as emergency back-ups for nuclear submarines. Small prototype

engines (25-kilowatt) have been operated since 1982, and larger units are planned (400-

8
600 kilowatt). [Ref. 11 ] Considerable interest in the process for the deep-water oil

exploration industry' virtually assures continued development.

This system is in use in an Italian midget submarine Mariaia (150-ton), with
more than a week of submerged endurance, at depths of up to 300 meters. [Ref. l:p..581
Figure 3 illustrates a typical closed-cycle diesel engine (from Fox page 29).

-ABC*" D-1~

gntakc

OOICSOL
Pratt.

Figure 3 - Closed-cycle diesel engine

2. Fuel Cells
Oxygen and hydrogen are combined in a fuel cell in a continuous chemical
reaction that directly produces electricity without any combustion. The German Navy has
tested this system in a Type-205 submarine (450 ton), and published reports suggest I

month submerged operations should be possible. The system is reportedly 50-70 percent
efficient, and, significantly for ASW purposes, is very quiet due to a lack of moving parts
and combustion [Ref. 1:p.58, Ref. 81. Figure 4 is an abbreviated schematic diagram of a
typical fuel cell system (from Benedict).

9
 "
0--.

Figure 4 - Fuel cell

3. Stirling Engines
Stirling engines have been operated by the Swedish Na"y in the AI14--clas

submarine Nacken (1000 tons) since 1989 [Ref. 12]. The system produces heat from

liquid oxygen in an external combustion chamber that is transferred under a constant and

relatively low pressure to the engine via a heat pipe. The external chamber is kept in over

pressure to facilitate overboard discharge of exhaust at depths of 300 meters. This indirect

heating technology makes it possible to seprarate the engine from an external heating

system, which limits risks [Ref. 13]. Reports of submerged endurance up to 2 wveeks, and

the ordering of a new class of submarine, the A 19 Gotland--class (first-of-class laid down

20 Nov ember 1992), reportedly powered by Stirling engines, suggests successful trials

[Ref. 6J. Figure 5 illustrates a typicald Sterling engine (from Benedict).

l0
 A-

Figure 5 - Stirling Engine

4. Small Nuclear Reactors
In contrast to the large reactor providing high pressure steam, small nuclear
reactors provide a "nuclear battery charger." [Ref. 13 :p.53 1 A Canadian design, the
autonomous marine power source (AMPS), has been licensed for unattended operation at

research facilities ashore. It is scheduled to be installed on the French SAGA-I commercial

ocean submersible by 1995 [Ref. 61. If these trials are successful, unlimited slow speed
endurance would result. Sprint speeds would be limited, however. Figure 6 illustrates a
typical low -power nuclear reactor (from Benedict).

11
 NUT

Figure 6 - 'Slowpoke' Nuclear reactor

Any of the above AIP systems will permit a conventional submarine to patrol

with either extremely limited snorkeling, or none at all, during a patrol. A ratio of
snorkeling time to patrol time is termed "indiscretion rate," as a submarine visible from the
surface is considered to be indiscreet.
It is noted that a Russian non-nuclear submarine, the Beluga, was reported in the
November 1991 edition of Jane's Defence Weekly to use an oxygen system (a fuel cell or

Stirling-type engine). Similar in form to the Alfa-class SSN, it was completed in Leningrad
in February 1987. [Ref. 1:p.57]

C. SUBMARINE "SIGNATURES"
A submarine's characteristic signatures that may be exploited by covert (passive)

detection systems are shown in Figure 7. Exploited properly, the submarine may never
discover it has been detected. In hostilities, it greatly enhances the ASW unit's likelihood of
survival, as a hostile submarine's first indication of the presence of an enemy %%
ould bc the
noise of a torpedo in the water. Since a submarine is not aware that it is being observed by

12
a passive system, standard operating procedures and habits can be discerned, which aid the
ASW force in detection tactics both in peace and war [Ref. 141.
H,
BHeuHmp felt.ma4e
AW- use Emep 6.b i d6 / adar/ magac

%Eut1io
er,
nternal Wav

WWakeI

Gas Dckmle Heal Bahiaweace

Figure 7 - Submarine Signature

In addition, ASW forces can overtly seek a submarine using active sensors, which

may or may not result in counter detection. A resume of these sensor is contained in
Figure 8:

Visual
Laser A

Infrared
Radar
Active Acoustics

Figure 8 - Active Sensors

In addition to being classed as either passive or active, detection systems are classed as

being acoustic or non-acoustic.

13
DR SCENARIOS
Where is the "threat'?" Increasing numbers of more capable submarines in less friendly

hands does not directly pose one. The future ASW threat depends on the scenanos under
which hostile submarines might be utilized. It is not the intent of this thesis to examine all
scenarios, but proposes the following four as possible for the Canadian Forces.
A "Peacemaking scenario" involves a peacemaking operation, against the %ill of a

nation which possesses a small number of modem diesel-electric submarines. The nation is
aligned politically with a larger nation which operates a number of nuclear powered attack
submarines. Canada provides a naval force to enforce a blockade.
A "Home defense scenario" involves a nation which is denied access to fisheries

within the 200 mile economic zone around Canada's coast, decides to force the issue, and
sorties a surface force to protect their national interest. The nation also operates a modem
submarine force, consisting of a small number of conventionally powered submarines,
which were sighted leaving port, have not been located, and are presumed to be enroute to

the Western Atlantic or Eastern Pacific.

A "Rogue submarine scenario" occurs when a nation possessing a large number of

submarines reports that a number of submariners recently forced to retire have seized a
submarine, left port, and are threatening to sink any vessel they encounter (or possibly,
strike inland targets with submarine-launched cruise missiles) unless a ransom is paid.
A "Covert delivery scenario" occurs when a drug cartel, disturbed by high losses in
other means of transporting drugs, buys modern submarines to off-load narcotics to North

American shores.
In the near term, i.e., the next ten to fifteen years, two submarine threats will be used

as examples of the range of targets ASW units could expect to encounter. Though they are

current technology, they are very capable, and challenge current ASW systems.

14
 I. Conventional Threat
The near term conventional submarine is a small diesel-electnc submarine. The

Thyssen Nordseewerke TR-1700 (built in Emden, West Germany) has been chosen as
typical of current capability. This submarine has the following characteristics [Ref. 5]:
Displacement 2116/2264 tons (surfaced/dived)
Speed 15/15/25 knots (surfaced/snorkeling/dived)
Dimensions 66x7.3x6.5 meters (length/width/height)

Diving limit 270 meters

Weapons: 6 torpedo tubes

mine capable
Range: 12,000 miles C 8 knots surfaced

20 miles C 25 knots dived (batter')

460 miles C 6 knots dived (batter')

Crew: 26
The indiscretion rate of the TR-1700 on patrol is assumed to be half of that
claimed on the voyage from Germany to Argentina cited above, one hour a day. This
means, on average, eight percent of the time, the submarine will have to snorkel in order to
charge batteries. During a rapid transit, the indiscretion rate would necessarily increase.
2. Nuclear Threat
The near term nuclear submarine is a relatively new SSN. Since it is unlikely that
a threat would develop from a western power, or that a western power would export an
SSN or SSN technology, a Russian Sierra 11 Class SSN has been selected (The second of

five was in service in 1993). [Ref. 51:

Displacement 7200/8200 tons (surfaced/dived)

Speed 18132 knots (surfaced/dived)

15
 Dimensions II Ix 14.2x8.8 meters (length/width/height)

Diving limit 650 meters

Weapons: 8 torpedo tubes (also missiles, tube launched)

mine capable (50 or 60 in lieu of torpedoes)

Range: unlimited

Crew: 100

The indiscretion rate of the SSN is near zero, as it need not ever Iea%e patrol
depth. Only tactical (final stages of attack) and National Command (communications)

considerations would require the Sierra to be indiscreet.

Calculations for the effectiveness of nonacoustic sensors later in this thesis will

be done using the characteristics of the TR-1700. It is assumed that larger submarines will

have larger nonacoustic signatures due to their larger volume and mass.

16
 Ili. SOME NON-ACOUSTIC DETECTION SYSTEMS

A. MAGNETIC ANOMALY DETECTION (MAD)

For distances greater than two submarine lengths, a submarine's magnetic field can be
considered to be a magnetic dipole field superimposed on the earth's magnetic field. This
field can be considered to be constant over the length of an encounter between a short-range

sensor and a submarine. The magnetic field is made up of two major components: A
permanent magnetic field and an induced magnetic field [Ref. 15]. The total field can be
considered to be the vector sum of the permanent and induced fields.
For a fixed spatial orientation a dipole field decreases by a factor of where r is the
-. ,
r
distance from the sensor. The characteristics of a submarine's magnetic signal depend

on:
* the size of the submarine.
" speed of the sensor relative to the submarine.
* the magnetic course of the sensor as it passes through the submarine's field.
* the magnetic course of the submarine.
* the effect of the natural magnetic background. [Ref. 9:p.6061

Magnetic fields are measured in Tesla (T) in the metric system. A Newton of force is
caused by 1 Coulomb of charge moving at 1 meter per second in a one Tesla field. One
Tesla equals 10,000 Gauss in the CGS system of measurement. One Gamma (commonly
quoted for military magnetic anomaly detection sensitivities) equals one nanoTesla.

For some feeling for the size of the magnetic signature of the submarine, some relevant
magnetic fields are given in Table I [Ref. 16]:

17
-- .I| I

500 T Pulsed electromagnetic fields (EMP).

100 T strongest DC fields.

IT laboratory magnet.

Ix I - 2 T refrigerator ceramic magnet.

(30- 60)x10 6 T earth's field (equator-pole).

"few" 10-9 T submarine magnetic signal.

Table I - Typical Magnetic fields

The ranges of current Magnetic Anomaly Detectors (MAD) are from 0.2 nanoTcsla for
the AN/ASQ-IOA to 0.05 nanoTesla (nT) for the AN/ASQ-81 (MAD systems deployed on
current U.S. ASW aircraft). This increase in sensitivity gives only a 59% increase in range,
which is on the order of a few thousand feet [Ref. 17]. The AN/ASQ-208 digital MAD
system has a sensitivity of 0.003 nT. It can be carried internally within an aircraft (other

systems are towed or moved as far away as possible from any source ol electrical or

magnetic noise). The latest system is the Superconducting Quantum Interference Device
(SQUID), which has been reported to have a sensitivity from ixI -5 to Ixi0- 6 nT. At
sensitivities this low, noise is a problem. IRef. 16]

The main source of noise is geomagnetic, due to the interaction bet-ween the sun and
the earth's magnetic field. The noise follows the sunspot cycle, with peaks every eleven
years. Terrestrial effects of solar flares can be split into three types:
* Sudden Ionospheric Disturbances (SIDS) occur immediately (eight minutes) after
a flare due to the arrival of x-rays. They affect the entire daylight portion of the
earth and last as long as the flare.
* Polar Cap Absorption (PCA) starts fifteen minutes after the flare due to the arrival
of protons from the flare. This lasts from one to ten days (normally around three).
There are about seven to eight PCA events per year during the solar peak, and
less otherwise.

18
 * Ionospheric storms are caused by the arnival of the plasma clouds. Electron
density increases and decreases in waves. The storms start twenty-four to forty-
eight hours after the start of the flare, and last from two to five days jRef. 18].

All three phenomena have the same effect on a MAD sensor. It is swamped with noise.
Storms occur, on average, about fifteen percent of the time, and no accurate prediction

system exists to forecast them. The effect of the storm increases with magnetic latitude due
to convergence of the magnetic meridians. There are also daily variations in the earth's field

due to solar heating that are latitude dependent. Figure 9 shows a typical MAD trace taken
in a moderate magnetic storm [Ref. 17:p. 187]:

Figure 9 - Magnetic noise in an ASO-10 MAD system due to a moderate

magnetic storm, compared with a submarine-generated signal.
Geologic formations can cause great local disruptions to the magnetic field, causing
false magnetometer readings. Maneuvers of the ASW platform (such as aircraft banking)
cause the sensor itself to change orientation within the magnetic field, causing a false signal
to appear on the sensor display, and use of certain transmitters on the aircraft, or in the

loc~al arca, also cause false signals. Positioning the sensor close to the center of gravity of

the aircraft minimizes this effect, as does the use of internald MAD rather than tow~cd

19
systems. High powered radio systems also create a large magnetic field locally, causing

man-made noise. This can be minimized by not transmitting during MAD prosecutions.
To summarize, any vessel made from magnetic materials has a MAD signature. It
decreases as the cube root of distance, and does not propagate like a wave. There are
significant naturl and man-made noise sources. It is a good sensor for close-in detections.

There has been interest in the use of non-magnetic steels (German Type 205 and 206
submarines) and building non-steel submarines (The Russian Alfa is widely reponed to

have a titanium hull). Non-magnetic steel has not been popular in more recent submarines
for unknown reasons. Titanium is half as dense as steel and non-magnetic, but has major
disadvantages:
cost ($33,000 per ton).
* inadequate resistance to brittle fracture (especially at low temps found in the deep
ocean).
special gas welding techniques are required.
* mainly foreign reserves. (US has only 3.5% of world reserves). [Ref. 17:p. 1371
Recent Russian submarines have not been constructed with titanium hulls, presumably

for reasons of cost.

1. MAD encounter simulation
An encounter between a MAD sensor and a submarine can be simulated using a
computer program. One program computes estimates that a system such as the AN/ASQ-81
will detect a submarine dipole field during an encounter between an aircraft (moving with
constant course, speed and altitude) and a submarine (moving with constant course, speed
and depth). Using values for World War II submarine magnetic fields, and assuming they

are valid for a modern diesel electric submarine, calculations were made to show the sweep
width of the AN/ASQ-81 against a TR- 1700 (2264 tons) in the waters off Sable Island,
Nova Scotia (approximately 44"N 60*W, variation 22"W). A magnctic noise of
0.35 gamma, equivalent to moderate geomagnetic storm noise, was used. Figure 10 is a

20
trace of lateral range versus probability of detection at closest point of approach (CPA) in a

straight line encounter [Ref. 15].

o ooooooeeo
,o* IPd=1

-600 0

ao

Lateral Range (meters)

Figure 10 - Lateral Range versus probability of detection at CPA for a

straight line encounter between a MAD sensor and a submarine
The upper curve is based on a crosscorrelation detector (the target signal Is
known) and it represents an upper bound for a detector's performance. The lower curv'e is
based on an energy detector, which represents a lower bound lor a detector's perf'ormance.
The drop "afterzero lateral range and lack of symmetry in the lateral range curve is due to
the nature of the dipole fields, and is target aspect dependent (here, MAD course _90*,
submarine course 02_0* ) [Ref. 151. From the figure, it can be conjectured that the s%%-ccp
wvidth for an energy detection MAD is approximately 500 meters against a small dicsl
submarine. The lateral range curves approximate a "cookie-cutter" detection system, that is,

21
one that will always detect a target within the lateral range curve for a certain range and will

always miss a target beyond that range. [Ref. 191

2. Random Search Using MAD

Random search was selected because it is mathematically easy to calculate and

understand, and because it provides a rough lower bound on search performance when a

searcher attempts uniform coverage of an entire area but is thwarted by randomness in the
target motion, navigation system errors and environmental uncertainties. The expected time
required to first detect a target is:

A
E[T] . WV(1
Wv

where A is the area to be covered, W is the sweep width (twice the detection

range for a "cookie cutter" detection system), and v is the speed of the searcher.
The probability of detection with a search time equal to or less than t time units is
given by the function:
Wvt

Pd() -l - e A (2)
where P,(t) is the probability of detection by that time [Ref. 19.p.2-4 to 2-71.

As an example, suppose a submarine is in an area 40 nautical miles by

40 nautical miles. Assume an ASW unit searches the area with a speed of 240 knots and a
detection system with a sweep width of 500 meters. Then, E[T] would be 24.7 hours,
and the probability of detection of the submarine 0.63 by that time. This example will be

used as a basis for comparison to other sensors.

B. HYDRODYNAMICS
There are many types of hydrodynamic disturbance caused by submarines, for
example, the Bernoulli Hump, Kelvin Wake and Vortex shedding.
 1. Bernoulli Hump
The near-field surface wave associated with a submerged submarine, sometimes
called a Bernoulli hump, can be considered the near-field of a source-sink pair ('the rankine

ovoid'), which is geometrically a good approximation of a modern submarine hull

|Ref. 17:p. 1931. The shape of the Bernoulli hump is independent of speed and insensitive
to depth. Water is displaced upward above the submarine, causing the surface of the ocean
to rise. The water comes from an area behind the submarine, causing a depression in the
ocean surface. Both these effects occur very close to the location of the submarine. The
height, however, depends on both speed and depth. Table 2 gives some values for an
Ohio-class submarine (from Stefanick page 195):

Depth Speed Height

(meters) (knots) (centimeters)
30 5 1.1
30 12 6.5
30 20 19.0
100 5 0.10
100 12 0.59
100 20 1.60
300 5 0.01
300 12 0.07
300 20 0.18

Table 2 - Height of near-field wave (Bernoulli hump)

It can be seen from the table that a submarine can avoid producing a sizable

hump by operating slowly or by increasing depth.

23
 2. Kelvin Wake
The far-field solution for the surface waves are illustrated in Figure II

(Ref. 17 :p. 19 6 1:

TranseverseWae

DOVegenl Waves

Centvq t ne-- -- - - --- ---

-Li -. Umst Of WGvO Pattern

Figure I I - Shape of far-field surface disturbance (Kelvin wake) over a

moving submarine (Dimensions of the Ohio-class)

Two types of waves, transverse and divergent, may be present. Both will be
contained within an angle of 19.5" of the direction of motion. Maximum wave heights for
the far-field solution for various speeds and depths are given in Table 3 [Ref. 17:p. 1951.

Centerline Wave Height (cm)

@ Downstream Distance (m)
Depth (m) Speed (kt) 100 S00 10,000
30 5 0 0 0
12 0.13 0.1 0.02
20 12.0 8.3 2.0
50 5 0 0 0
12 0.001 0.5= 0.0001
20 1.9 1.3 0.32
100 5 0 0 0
12 0 0 0
20 0.02 0.01 0.003

Table 3 - Maximum far-field wave height due to a moving submarine

24
 It should also be obvious that the smaller the submarine, the smaller the effect on

the surface, but even a small submarine, say one accelerating away from the surface after
an attack (torpedo or missile launch) may cause a considerable disturbance on the surface.
In normal patrol, the wake would be much more difficult to detect. However in shallow

water, submariners try to keep speeds low, as the size of depth excursions increase with
speed, and in shallow water a submarine at high speed would risk either broaching (coming

to the surface) or running into the bottom [Ref. 17:p. 136).

3. Vortex Shedding
Submarines are generally operated with positive buoyancy. That is to say, in the
absence of an) other force, they will rise towards the surface. In practice this force is
balanced with lift from control surfaces such as the diving planes. These are similar to
airfoils, except they work to keep the vehicle down, not up. An area of high pressure is on
the top, low pressure on the bottom. Water tends to flow from high to low pressure, so a
vortex is shed off the end of the hydrofoil. The vortex rises towards the surface. An
equation for the velocity of the vortex v, is given as equation (3):

Vo - 2xpudw (3)

where AB is the buoyant force, p is the density of the fluid, u is the

speed of the submarine, w is the width of the control surface, and d is the depth of the

submarine.
Assuming the positive buoyant force is 1x1O5 Newtons (approximately 21 tons;
while this seems large, it represents only one tenth of one percent of the displacement of an

Ohio-class submarine), a control surface width of 10 meters, a depth of 50 meters, and

submarine speed of 5 meters per second (10 knots), and a water density of 1000 kilograms

per meter cubed, a vortex speed of 0.006 meters per second is calculated. This is verV

small compared to wind-driven currents at the surface.

25
 The vertical speed of the vortex is given by equation (4):
AB
v6 (vertical) A
pu2jrw' (4)

For the same example as above, the vertical speed can be calculated as
0.03 meters per second. It would take 1.7xO 3 seconds to rise the 50 meters to the

surface, and would be 8.3 kilometers behind the submarine when it did.

The end result of one of Kelvin's other theorems is that vortices never end.
Thus, a starting vortex occurs, and stays in place until friction dissipates its energy. This
can be seen when paddling a canoe. Should there be a sharp turn, or a rapid control

movement, a starting vortex, known as a knuckle, will be formed, which may persist. This
phenomena would likely only be exploitable by another submarine. [Ref. 161
Whether or not hydrodynamic phenomena are exploitable is open to question.

The sizes of any of the effects on a submarine which is operated with any care are very
small with respect to wind-driven gravity waves. Should a submarine be used by

inexperienced personnel be operated at high speed at a shallow depth, however, it would be

easily detectable. The calmer the water surface in the area, the easier it would be.
With the upcoming revolution in computer processing power (Appendix A), it is

possible that the effects of wind-driven waves could be recognized, and a sensor set to

ignore them since wind velocities are readily available to the computer from a number of air
data sources. The presence of a wake, with the characteristic shape of the Kelvin wake

envelope (19.5T), could be searched for by a cr(sscorrelation detector to maximize the

detection probability.
Modulation of the wind-driven surface waves may cause radar backscatter to

measurably change. Whether a detector with an acceptable false alarm rate could be

constructed is an open question, and one which may bear investigation. If the surface
effects are discernible, this technique is suited to both deep-water (SSN) detection, and

26
shallow-water use, though tides and current effects would make shallow water detections

more difficult. Waves grow as they enter shallow water, and this exacerbates the problem
of seeing the smaller waves due to hydrodynamic effects.

The use of a space-based system as a search sensor is considered unlikely since

the effects are ier' limited in extent, the use of a space-based radar would require
considerable power and the problem of transferring the bearing and distance obtained by a
vchicle moving so quickly into a ground-based coordinate system (latitude and longitude)
should a detection occur are formidable. An airborne sensor would be more appropriate for

this technique.
It is assumed for the purposes of this thesis that a sensor exists which can detect
the wake of a submerged submarine.

4. Sweep Width Using Wake Detection

The wake detection model is based on the following assumptions:
* A target wake is a moving straight line segment in a horizontal plane.
* The wake is detected if and only if the horizontal range of a detection
system's sensor from at least one point on the wake is equal to or less than a
definite value.
* In a straight line encounter, the angle between the projection of the sensor's
track on the plane of the wake and the wake is a random variable that is
equally likely to have any value between 0* and 900.

The encounter geometry is shown in Figure 12 [Ref. 1513 . The angle betwecn

the wake and the sensor's projected track on the plane of the wake is a, and the horizontal
distance between the midpoint of the wake is x. The encounter geometry is at the time the

midpoint of the wake is at the closest point of approach (CPA):

3 The definite range circle in the Reference has been replaced by a ddinite range band for darit), and the
angle between the wake and sensor track has been repositioned correctly after discussion with the
Reference's author.

27
 S~no Track
Defiite ang

\ Wake

Figure 12 - Encounter geometry in the plane of the wake for a straight line
encounter between a wake and a detection system's sensor
The sweep width of the sensor can be shown to be [Ref. 201:

W - 2(R+A)
jr (5)
where R is the detection range of the sensor.

A is the length of the wake.

The formula shows that the sweep width is approximately twice the detection
range of the sensor, with a factor to take into account the length of the wake. Since constant
motion of the sensor would cut a rectangular siath, it is an appropriate model for a radar
sweep.

A number of examples were done using the MATLAB computer language

(Appendix B). A fixed detection range was assumed (1M) and 1000 meters respectively).
Two graphs of sweep width versus detection range are attached as Figures 13 and 14.

28
 45W-

4000-

3500-

30D

2000-

1500-
S50 10-30 40 50 60 0 0 (0
1000

0 i00 _00 300 400 500 600 700 800 900 1000
wake length imeters

Figure 13 - Sweep width versus wake length with a fixed detection range of
100 meters

5an

4500

4000-

3500

3000-

2500 .

1000 -

500

o0 100 200 300 400 500 600 700 800 900 1000

wake lengh fmeterp]

Figure 14 -Sweep width versus wake length with a fixed detection range of
1000 meters

29
 The graphs show the most important factor of equation (5) is the detection range.
An increase in wake length from 100 to 1000 meters with a fixed 100 meter detection range
only increases the sweep width 600 meters, while a similar increase in detection range for a

fixed wake of 100 meters increases sweep width 1800 meters. In other words, the formula
for sweep width is relatively insensitive to wake length. The most gains can be realized by

increasing the detection range of the sensor. The sweep width is approximately double the

detection range.
5. Random Search Using Wake Detection
Assuming a detectable wake length of 100 meters, and a detection range of I(00

meters, the sweep width is 2063.7 meters. Using the same area as in the MAD example,
the EITI would be 6.0 hours, with a detection probability of 0.63 by that time.

C. LASER DETECTION SYSTEM.

1. Overview of Laser Radar (Lidar)
The concept of laser detection of a submarine is very simple. A laser is pointed at

the ocean, a pulse is sent, and a receiver plots the returned energy versus time. There are
significant problems, however. The atmosphere and ocean attenuate the pulse through
absorption and scattering, there are scattering losses from the submarine upon reflection,
and returned energy must be discriminated from that reflected from the surface.
As early as 1967, a submerged submarine (USS Thread Fin) was detected as a
target of opportunity during a mine-detection experiment near Panama City, Florida

[Ref. 211 The laser used a 530.8 nanometer wavelength, 25 nanosecond pulse, a 5 inch
(2 centimeter) aperture, and 2 megawatts peak power.
The wavelength of the laser is very important. Coastal waters are most transparent

at a wavelength of light of approximately 500 to 550 nanometers, and the open ocean varies
depending on which ocean is being measured, though it is usually around 470 nanometers.

30
 -
[Ref. 22, 231. Figure 15 shows the relationship between absorption coefficient (in m ) and
wavelength in the ocean (in nanometers) [Ref. 161:

1.0- 1 9

Omoan

.01 I I I I I a

200 400 600 8W

waowngth (nm)

1
Figure 15 - Absorption coefficient (W ) versus wavelength (nanometers)
The major features of a laser radar (Lidar) trace are the transmitted pulse, the

return from the surface, and the return from the target (if any). In deep water, the return
will be attenuated to the point where the return is hidden in noise, and in shallow water, a
return from the bottom is also displayed. Figure 16 shows a simplified diagram of a typical

laser return in deep water versus time [Ref. 16J:

31
 ,. laser Ocean Surface

II
Figure 16 - laser return versus time
There are two separate cases which must be considered due to the geometry of
the problem. Th,' beam can be smaller than the submarine or larger. To explain which areas
are being described irthe urrnulae to follow, a diagram (Figure 17) is included [Ref. 161:

Ai,

/ A,

r, / !

I ,/A,

Figure 17 - Areas in Lidar formulae

If the beam is larger than the submarine (the case seen in Figure 17), the returned

power to the aircraft receiver is given by equation (6):

32
 Pe -2 r2 (l R) 2 UeA5.6
1r (6)
A2 A
4

where P, is the transmitted power in watts, a is the absorption coefficient of

water in meters -1, r. is the depth of the submarine in meters, (1 - R, ) is the fraction of the

intensity transmitted through a surface, a, is the cross sectional area of the target in

meters 2. A; is the area of the receiver (aperture) in meters 2 , A2 is the area of the Lidar beam
at the submarine depth in meters 2 and A4 is the area of the returned beam at the aircraft

altitude in meters 2 .
If the beam is smaller than the submarine, the normal case, the equation (7)
becomes:

2
P,e-2 2 (I - R,)RPA
Pr A4 (7)

where R is the reflectivity of the target.

For diffuse reflection from a flat surface (Lambertian) the formula for the small
beam becomes equation (8):

-2 ar,(I_ R )2 A5
Pt s
2 r rr+ r2 I
A 4 (r! + r 2 ) \r, w
hair + r, "

where n is the index of refraction.

The back scatter from the ocean can be calculated by equation (9):
(
2 ar, (I1_14v) 2 p(O)cr~ _____
Pr Pte-2 )2(
2
j rl+ r2 (9)
2n(r I + r) I lrater + r2
hair )

33
 where #'(6) is the dimensionless volume backscattering coefficient, c is the

speed of light in meters per second, and T is the laser pulse length in seconds.
Typically, in current systems, 100 to 1,000 photons are required for a detection.

Energy is quantized, and this can be described by equation (10):

E = nhf (10)
where n is the number of photons, h is Planck's constant, and f is the

frequency in seconds-.

The energy of 100 photons in a 500 nanometer beam, of pulse length

1 nanosecond, is 4.0x 0-1'7 Joules. Converted to power, this is 4.0x 10" watts.
Graphically, a transmitted beam of 10 megawatts with no absorptive losses in the
atmosphere, in very clear water (a of 0.02 meters-1 ) would give the following results
(Figure 18):

0.001
.. Pt. 1l07 W

0.0001 "V""o.n. ~ -,
0m
oaVm

IEOle

1E-07

1 E-0 10 PhM (ira)

IE-09
0 20 40 60 80 100 120 140 160 180 200
epth (m)

Figure 18 - Theoretical lidar return from a submarine, clear water

Sevcral interesting things can be seen in the picture. First, the large beam acts
like a small (non-spreading beam) for a small depth, then begins to spread (at about 40
meters). Second, the ocean back scatter for the longer pulse (10 ns) would totally obscure
the returned signal. The small beam remains effective to depths of around 200 meters.

34
 In more turbid water (a of 0.06 meters- 1 ), the light is less able to penetratc.

Figure 19 shows the results of repeating the calculations with all else being constant (note

the reduced depth scale):

0.0011

0.0W01''--

1E-0,5
I1E-06i ph / "...0

1E-07"-.

1E-05.
I1E-09 10 Pho (S)
(I

IE-10
10 20 30 40 50 60 70 80 90 100
Depth (m)

Figure 19 - Laser return from a submarine, turbid water

The reduced ranges (shallow water) can be offset by using a more powerful

laser.

2. Random Search Using Lidar

The sweep width of the Russian Amethyst system, a blue-green laser carried by

Bear-F Mod 4 aircraft, is reported to be 100 meters, with the aircraft at an altitude of

100 meters, and a speed of 200 knots [Ref. 241. Assuming a system with these

characteristics, and using the same area as in the MAD example, the EITI would be 123.5

hours, with a detection probability of 0.63 by that time.

D. ELECTROMAGNETIC SUPPORT MEASURES (ESM)

A submarine's use of radio or radar allows ASW units an opportunity to detect the

submarine. A distinctive emitter signature can give away the identity of a submarine, which

35
makes the searching unit's job easier (knowing what to look for makes it easier to find).

This is balanced by several factors in the submarine's favor.

First, the emitter probably operates in a band shared by a number of other radars, or

radio transmitters. Having an emitter in the band may not indicate the presence of a

submarine, just a fishing vessel. This problem will get worse as more navies adopt

commercial-off-the-shelf (COTS) equipment, and identical systems are operated on

combatants and non-combatants. Second, other electronic navigation and fire control
systems have improved greatly, obviating the need to emit radar for navigation or final

target solutions. Communication can be done almost undetectably by several means.

Submarine operations in shallow water used to mean that more emissions could be
expected, because the submarine wished to fix its position more frequently, to avoid
grounding. The use of the Global Positioning System satellite navigation systems has

meant that a unit can fix its location covertly to a higher precision than hitherto was possible
overtly with radar.
It is the author's experience that current electronic detection devices are limited by
computing power. A balance has been necessary between coverage of all frequencies
bands, and probability of detection of the most important signals due to the time-sharing of

the processor. This compromise has meant that detection opportunities have been missed,
according to post-exercise reports (personal experience). The advanced processing and
storage capability of the upcoming computers (Appendix A) will give electronic warfare
equipment the ability to monitor a larger range of the electro-magnetic spectrum

simultaneously, resulting in higher probabilities of detection. Yet, this must be balanced by

the fact the submarine may not have to transmit, due to the factors previously mentioned.

36
 APPENDIX A - COMPUTING ADVANCES

One of the major drawbacks of current (and future) sensor employment is limited

processing capability. In the past, specialized computers were developed specifically for
ASW use. Because of the long lead times required to design and produce these computers,
were obsolete on delivery. For example, many P-3 aircraft continue to fly today with
1970's vintage computers. The traditional computer research and development (R&D)
leader, the military, has been replaced by industry. The advances commercial industry has
made in processing, data storage and retrieval, reduced size, and increased reliability must
be taken advantage of.
The thesis is based on the assumptions that ASW systems will increasingly rely on
commercial-off-the-shelf (COTS) systems, also known as non-developmental items (NDI),
and it is assumed that computers in military systems now in the final stages of

development, and those on the draving board, will almost certainly closely resemble those
in the best contemporary technology. A recent article from The Journal of Electronic

Defense, "Military Computing in the Year 2001 and Beyond" (Ref. 25] gives an excellent
assessment of what Central Processing Unit (CPU) is likely to be found in a military
computer in the year 2001.
Six microprocessors have been introduced in the 1992-93 time frame, all of which
except one are super scalars, i.e., they are able to issue two or more instructions in a single
clock cycle. Digital Equipment Corporation's Alpha 21604 (a reduced-instruction-set-

computer (RISC) running at 200 MHz with 64-bit wide data registers), and Hewlett-
Packard's PA7100 (100 MHz, 64-bit, RISC) chips were released in February 1992. The
U.S. Air Force has announced that the Alpha chip would replace the Motorola 68040

processor in the E-8 Joint-STARS aircraft.

37
 In May 1992, Sun Microsystems and Texas Instruments brought out the SuperSparc

(60 MHz, 32-bi, RISC) which can issue three instructions per clock cycle. In November,

MIPS Technologies Incorporated brought out the R4400SC (150 MHz. 64-bit, RISC). In
1993, Intel introduced the Pentium, which uses an older complex-instruction-set-

computing (CISC) chip, and Motorola introduced the PowerPC 601 chip (80 MHz, 32-bit,

RISC), able to issue three instructions in a clock cycle. Both IBM and Apple computer will
produce home computers with performance heretofore only seen in (and sometimes
surpassing) mainframe workstation computers.
The next generation chips will be even faster. Digital's Alpha 21064A (275 MHz, 64-
bit, RISC) promises a 30 percent increase in integer (i.e., normal computing operations)
and 60 percent increase in floating point (i.e., mathematical calculations) performance. But
what does this mean?

As microprocessors speeds increase, the limit on what a single chip can do is being
approached. The future is between the single chip, and the "massively parallel"
megaprocessors used in supercomputers, in a niche known as "arrayed microprocessor."

The newest Silicon Graphics family of processors (R-4400, RISC, 4 instructions per clock
cycle, throughput of 300 MFLOPS (300 million floating point operations per second -
equivalent to the current Cray Y-MP supercomputer)) use such technology, are expected to
achieve between 2.7 and 5.4 GFLOPS (billion floating point operations per second).
Compare this with a common system with today's technology, the Intel-80386 running at

20 MHz, which achieves 232 KFLOPS (that is, 0.000232 GFLOPS). Capability for
mathematical operations will be 20,000 times what they are today.

Mass storage systems have kept pace with the increased power of the central
processing units. Storage of data used to be done on magnetic tapes, which were slow to
find any but sequential data, and required bulky readers. These were replaced with quickly
rotating magnetic drum storage, which allowed much quicker access to any stored data.

38
These systems wcre capable of storing a few megabits of data in the early 1970's, and arc
still used in some airborne systems such as the CP-140 Aurora Maritime Patrol Aircraft.

In the 1980's, systems made up of stacked rotating platters of magnetic media, hard

disks, became the commercial norm. Access speeds increased, and the abilit to pack more
data into the same area increased. By the end of the decade, systems able to store Gigabits
of data were common. These systems can be linked in arrays to provide inexpensiC.
effectively limitless mass storage. In the 1990's, systems still use magnetic media, but the
trend is toward the use of optical media. These use low power lasers to modify the surface
C , c into orientations which the systems can recognize as binary data. They do not
dct., e in the short term (ten years), are resistant to electro-magnetic pulse, and are very
light. Access speeds are slightly slower than magnetic media, but this problem will be
cured in the near term. Again, arrays of these disks represents unlimited storage, so that

libraries of threat submarine data can be carried, and the greater processing power of the

aircraft computer could be used for automatic detection.

Based on the above, the COTS military computer projected to enter senice in the year

2001 is expected to run at clock speeds of 2 GHz or faster, and be capable of speeds
equivalent to a Cray C90, that is, about I-GFLOPS operation.
Current ASW systems are held back by computing considerations. For example,
monochrome displays are common, because the computing power to display color is not
available (for 256 color operation, even" picture element requires an extra 8 bits of data).

Operators' ability to recognize targets will improve, because more information will be
displayed. Studies have shown that the average human can recogni/e five to twenty-five

hues within a color (as might be seen on a "waterfall" acoustic display), but can recognize
256 colors IRef. 261 As well, automatic detections, which are based on only a small
amount of information and limited threat libraries (due to throughput problems), and have

been characterized as taking more time than they save, can be greatly improved. Also, the

39
number of sensors which can simultaneously be processed will increase. For example,
instead of processing sixteen sonobuoys on a monochrome display, a future processor will
be able to handle fifty or more, and multi-color displays will be the norm. With the advent

of smaller, cheaper, more robust electronics, some of the processing hitherto done on

board the ASW unit may be done by the sonobuoy instead.

The cost of the new hardware can also be justified because the ability to support
vinu',, technology is also r'ery costly. Chips which were inexpensive ten years ago due to
mass production must now be produced one at a time, at considerable expense and long
delivery times, when they fail. Current technology chips can be ordered off-the-shelf
cheaply.
But hardware is only half the solution. At least fifty percent of the cost of a new
system is due to software costs. It is assumed that spending will continue in developing
software to use with new computers, and that the old software will not just be modified to

run on a newer system.

The amount of information that can be linked to and from ASW units, and assets
ashore, will increase greatly, allowing the use of a smaller number of much more capable
units. In an era when there are fewer units, increased computing power will be required to
make them more capable.

40
 APPENDIX B - WAKE DETECTION MATLAB PROGRAM
%Matlab 3.5 program for sweep widths with various wake lengths, detection
% ranges. Daly. 23 March. Thesis chapter 3, wake detection.
clear
cig
axis(10 1000 0 5000]); %fix graph axis x 0 to 1000, y 0 to 5000
% first fix detection range R at 100 moters, vary wake length;
R-100; %dot range in meters
lambdamlOO:100:1000; %wake length in moter.
w-2*(R+(lambda/pi)); % forrala for sweep width
plot ( lanrbda,w) ; grid; xlabel (wake length [maters.I') ;
ylabel ( sweep width [naters] );title( 'sweep width for R-100 mater. ); pause

% change detection range R to 500 moters, vary wake length;

R-500; % dot range in meters
luuda100z100t1000; %wake length in maters
w-2*(R+ (lambda/pi)); %fozmla for sweep width
plot(lanbda,v) ;grid;xlabel( 'wake length (mater,]');
ylabel ('sweep width [meters] );title( 'sweep width for Rin500 meters'); pause

% change detection range R to 1000 meters, vary wake length;

R-1000; % dot range in mter.
lambdalOO0:100t1000; % wake length in meters
w-~* (R+( lambda/pi)); % formala for --OFpwidth
plot(lanbda,w) ;grid;xlabel( 'wake length [motors]*);
ylabel('sweep width [meters]');title('sweep width for R-1000 mator.'); pause

% change detection range R to 2000 maters, vary wake length;

R-2 000; % dot range in meters
lasbdainloo:100sl000; % wake length in meters
w-2*(R+(lmbdalpi)); % formula for swaep width
plot (lezrbda,w); grid; xlabel ('wake length [mater.]1);
ylabel ('sweep width [Eaters]') ;title( 'sweep width for R-2000 mater.' );pause

1
%~~****** new section ********
% now fix wake length and vary dot range;
clear
clg

% axisQO 1000 0 3500])

% make length of 100 meters
lambda-1OG; % wake length in meter.
R-100:100:1000; % dot ranget in metors, 100 m,- 10O0m in 100m, steps
w-2*(R+(lambda/pi)); % formala for sweep width
plot (R,v) ;grid;xlabel ('detection range [meters]');
ylabel ('sweep width [meters)');
title( 'sweep width for wake of 100 meters'); pause

% wake length of 500 meters;

lambda-5 00; % wake length in meoters
R-100:100t1000; % dot range in meters, 100 m - 1000m in 100m. steps
wm2*(R+(ambd/pi)); % formala, for sweep width
plovt(R,w) ;grid;xlabel( 'detection range [Eoars]');
ylabel ('sweep width (meters]');
title( 'sweep width for wake of 500 meters'); pause

41
 wake length of 1000 meters;
lusbda-l00; % wake length in meters
R-iOlOOslOOO; % dot range in meters, 100 m - lO00m in lOOm steps
w-2*(R+(lanbda/pi)); % formula for sweep width
plot(R,w) ;grid;xlabel( 'detection range (meters] '); pause
ylabel( 'sweep width [mters]' );
title( sweep width for wake of 1000 meters' );

%make length of 2000 meters;

lambda-2000; % wake length in meters
R-1001001000; % dot range in meters, 100 m - 1000m in 100m steps
w-2*(R+(lambda/pi)); % formula for sweep width
plot(R,w) ;grid;xlabel( 'detection range [meters]');
ylabel( 'sweep width [meters]');
title( 'sweep width for wake of 2000 meters'); pause

42
 LIST OF REFERENCES

1. Benedict, J.R., "Third World Submarine Developmcnts," MaritinePatrolAviation,

Vol. 2 No. 2, p. 5 7 , March 1992.

2. "Aven Views Shift in Weapons Export Market, Increased Sales," Inteta.t. 30

November 1992 (Trans.: FBIS-SOV-92-234), p. 7,quoted in Zaloga, S.,
"Harpoonski," Naval Institute Proceedings, Februar%94, page 40.
3. Kraus, G.F. Jr., "World Submarine Proliferation and U.S. Contingencies." The
Submarine Review, page 61, April 1993.
4. The Enc.clopedia AmericanaInternationalEdition, quoted in Benedict, J. "Thi rd
World Submarine Developments," The Johns Hopkins University Applied Physics
Laboratory, presented to the 8th Annual Naval Submarine League Symposium, 13
June 1990.
5. 1993-94 Jane's FightingShips, Jane's Information Group Incorporated,
Alexandria, VA, 1993, page 11.
6. 10/31/88 and 4/3/89 Navy News & Undersea Technology, quoted in Benedict, J.
"Third World Submarine Developments," The Johns Hopkins University Applied
Physics Laboratory, presented to the 8th Annual Naval Submarine League
Symposium, 13 June 1990.
7. 1/30/90 CNO Intelligence Notes (Unclassified Portion), quoted in Benedict, J.
"Third World Submarine Developments," The Johns Hopkins University Applied
Physics Laboratory, presented to the 8th Annual Naval Submarine League
Symposium. 13 June 1990.
8. 89/90 Jane's Fighting Ships, quoted in Benedict, J. "Third World Submarine
Developments," The Johns Hopkins University Applied Physics Laboratory,
presented to the 8th Annual Naval Submarine League Symposium, 13 June 1990.
9. Gerken, L., ASW versus Submarine Technology Battle, American Scientific
Corporation, 1986, page 393.
10. NITA-DA30-006-89 quoted in Benedict, J. "Third World Submarine
Developments," The Johns Hopkins University Applied Physics Laboratory,
presented to the 8th Annual Naval Submarine League Symposium, 13 June 1990.
11. Fox, H., "Closed Recycling Powers Underwater Diesel Engine," Sea Technology,
April 1990, page 29.
12. 1993-94 Jane's FightingShips, Janes's Information Group Incorportated.
Alexandria, Virginia, page 642.

43
13. Breemer, Dr. J.S., "AlP: The Quiet Underwater Revolution," Maritime Pairol
Aviation, September 1991, page 53.
14. Hill, Rear Admiral J.R., Anti-Submarine Warfare, Naval Institute Press, 1985,
page 39.
15. Forrest, R.N., "A Program to compute Magnetic Anomaly Detection Probabilities,"
Technical Report NPS71-88-001 (2d rev), Monterey, March 1990, page 20.
16. Unpublished class notes, Physics 3002, Naval Postgraduate School, MontcreN,
California.

17. Stefanick, T., StrategicAntisubmarine Warfare and Naval Strategy , Institute for
Defense & Disarmament Studies, Lexington Books, 1987, page 186.

18. Hargreaves, J.K., The Upper Atmosphere and SolarTerrestrialRelations, Van

Nostrand Reinhold Company, 1965, quoted in unpublished class handouts,
Meteorology (for ASW) 2413, Naval Postgraduate School, Monterey, California.
19. Washburn, A.R., Search and Detection, 2d ed., Operations Research Society of
America, October 1989, page 4-1.

20. Forrest, R.N., Estimating Search Effectiveness with Limited Information,

Technical Report NPS-AW-93-001, September 1993 page 14.
21. Swarner, W.G., et al.. Report 2170-5, Subsurface Target Detection Techniques,
Technical Report AFAL-TR-67-78, May 1967, page 13.
22. Shifrin, K.S., Physical Optics of Ocean Water, American Institute of Physics,
translated by Oliver, D., 1983, page 173-175.

23. Hulbert, E.O., "Optics of Distilled and Natural Water," Benchmark Papers in
Optics Volume 3 - Light in the Sea, edited by Tyler J.E., Benchmark, 1977,
page 135.
24. Slade, S., "Hunting the Sound of Silence," Naval Forces, Volume XIV, Number
V, 1993, page 25.
25. Herskovitz, D., "Military Computing in the Year 2001 and Beyond," The Journal
of ElectronicDefense, Volume 17, No. 2, February 1994, pages 41-45.

26. Sanders, M.S., and McCormick, E.J., Human Factorsin Engineeringand Design,
7d, McGraw-Hill, 1993, page 100.

44
 INITIAL DISTRIBUTION LIST

Defense Technical Information Center 2

Cameron Station
Alexandria, VA 21304-6145
2. Library, Code 52 2
Naval Postgraduate School
Monterey, CA 93943-5002
3. Director, Submarine Warfare Di%ision
Chief of Naval Operations (N87)
Pentagon Room 4E453
Na\ y Department
Washington, DC 20350-2(XX)
4. Director, Exeditionary Warfare Division
Chief of Naval Operations (N85)
Pentagon Room 4A720
Navy Department
Washington, DC 20350)-2(XX)0

5. National Defence Headquarters

MGen George R. Pearkes Building
Ottawa, Ontario, Canada
KIA 0K2
Attention: DPED
6. National Defence Headquarters
MGen George R. Pearkcs Building
Ottawa, Ontario, Canada
K IA OK2
Attention: DMA 3-5
7. LCDR M.D. Kite, Code PH/Ki
Department of Physics and Chemistry
Naval Postgraduate School
Monterey, CA 93943-5002
8. Dr. R.N. Forrest, Code OR/Fo
Department of Operations Research
Naval Postgraduate School
Monterey, CA 93943-5002

45
9. Dr. I. V. Sanders, Code PH/SdI
Department of Phy'sics and Chemistry
Naval Postgraduate School
Monterey, CA 93943-5002
10. Maj. D.G. DaIN
14 Software Enginenng Squadron
14 Wing Greenwoodx
Greenwood, Nova Scotia, Canada
BOP INO

46