Geo-Encryption Using Loran

33
Geo-Encryption Using Loran Geo-Encryption Using Loran Stanford University Sponsored by FAA Loran Program Di Qiu [email protected]

Transcript of Geo-Encryption Using Loran

Page 1: Geo-Encryption Using Loran

Geo-Encryption Using LoranGeo-Encryption Using Loran

Stanford UniversitySponsored by FAA Loran Program

Di [email protected]

Page 2: Geo-Encryption Using Loran

Digital Cinema DistributionDigital Cinema Distribution

“Today, the film studios spend over $1 billion each year to duplicate, distribute, rejuvenate, redistribute and ultimately destroy the thousands of film reels required to bring the close to 500 films released each year to audiences across the U.S.”

• “rise against the reel”– 35-mm print cost $ 1,200– limited showings– heavy, weighs 50 lbs– platter setup takes an hour

• Digital cinema– cost less, < $100 per screen load– unlimited showing– automatic setup

Page 3: Geo-Encryption Using Loran

Digital Distribution DisadvantageDigital Distribution Disadvantage

• Napsterization concern• Music sales are down 8%• Music company valuations are down 40%

• Satellite distribution• ~3 million unauthorized users

Page 4: Geo-Encryption Using Loran

Encryption & DecryptionEncryption & DecryptionSymmetric Cipher: Encryption key = Decryption key

Asymmetric Cipher: Encryption key = Decryption key

Plaintext

Key

Ciphertext

Plaintext

Key

Plaintext

Encryption Key

Ciphertext

Plaintext

Decryption Key

Page 5: Geo-Encryption Using Loran

Hybrid Systems-- a combination of asymmetric and symmetric

ciphers

Hybrid Systems-- a combination of asymmetric and symmetric

ciphers

Generate Random Key

Generate Random Key

EncryptEncrypt

DecryptDecrypt

Plaintext

Plaintext

Ciphertext

Key

Key

EncryptEncrypt

DecryptDecrypt

Encrypted Key

Key_E

Key_D

Page 6: Geo-Encryption Using Loran

What is Geo-encryption?What is Geo-encryption?

• Proposed by Dorothy Denning and Logan Scott

• Add another layer of security

• Not a replacement of the conventional crypto algorithms

Page 7: Geo-Encryption Using Loran

Geo-encryption Algorithm-- Enhance the security

Geo-encryption Algorithm-- Enhance the security

Generate Random Key

Generate Random Key

EncryptEncrypt

DecryptDecrypt

Plaintext

Plaintext

Ciphertext

Key

Key

EncryptEncrypt

DecryptDecrypt

GeoEncrypted Key

Key_E

Key_D

Recipient Location, Signature

Recipient Location, Signature

Location Signature →Geolock Mapping

Location Signature →Geolock Mapping ⊕⊕

AntiSpoof Enhanced GPS or Loran ReceiverAntiSpoof Enhanced

GPS or Loran Receiver

Location Signature →Geolock Mapping

Location Signature →Geolock Mapping

GeolockGeolock

GeoEncryptionGeoEncryptionGeoEncryption

GeoDecryptionGeoDecryptionGeoDecryption

⊕⊕GeolockGeolock

Page 8: Geo-Encryption Using Loran

Comparison of the SignalsGPS v.s. LORAN

Comparison of the SignalsGPS v.s. LORAN

GPS

Pros• Stable clock• High absolute accuracy• High repeatable accuracy • Global coverage

Cons• Low SNR• Position accuracy depends on SV geometry• LOS dependent• Easy to spoof• Indoor not capable

GPS

Pros• Stable clock• High absolute accuracy• High repeatable accuracy • Global coverage

Cons• Low SNR• Position accuracy depends on SV geometry• LOS dependent• Easy to spoof• Indoor not capable

LORAN

Pros• Stationary transmitters• Know signal shapes and SNR• Groundwave propagation• High signal power • Jamming Loran is hard• Indoors capable

Cons• Skywave contaminations• Signal quality depends on the transmitter distance

LORAN

Pros• Stationary transmitters• Know signal shapes and SNR• Groundwave propagation• High signal power • Jamming Loran is hard• Indoors capable

Cons• Skywave contaminations• Signal quality depends on the transmitter distance

Page 9: Geo-Encryption Using Loran

Research ObjectivesResearch Objectives

• Signal authentication– Allows the receivers to ascertain its origins– Allows the receivers to verify that it has not been

modified in transit• Loran location signature

– Study the consistency of Loran signal– Design Loran location signature– Map Loran location signature into geo-lock

• Build geo-encryption demonstration testbed

Page 10: Geo-Encryption Using Loran

Modified Geo-encryptionModified Geo-encryption

Geo-lock MappingGeo-lock Mapping

Geo-lock MappingGeo-lock Mapping

Recipient’sLocation Signature

Recipient’sLocation Signature

Loran ReceiverLoran Receiver

AESAES

plaintextRandom KeyRandom Key

Symmetric key

RSARSA

Encrypted key

RSARSA

Key_E

Key_D

AESAES

plaintext

ciphertext

Symmetric key

Signal AuthenticationSignal Authentication

YesNo

Page 11: Geo-Encryption Using Loran

Signal Authentication RequirementsSignal Authentication Requirements

• Low computation overhead for generation and verification of authentication

• Low communication overhead• Buffering requirement• Robust to packet loss• Scales to a large number of receivers

Why is Security for Broadcasts Hard?• Symmetric authentication - not secure• Asymmetric mechanism - not as efficient as symmetric

authentication.Timed Efficient Stream Loss-tolerant Authentication (TESLA)Timed Efficient Stream Loss-tolerant Authentication (TESLA)

Page 12: Geo-Encryption Using Loran

• Hash function : One way function– Collision resistant– Digest any message to a fixed hash value– MD5 (128 bits), SHA1 (160 bits), SHA256 (256 bits)

• Message Authentication Code (MAC)– Keyed hash function– Symmetric– Require to transmit the key– TESLA uses MAC

Crypto ReviewCrypto Review

plaintext

mac Verify

Page 13: Geo-Encryption Using Loran

TESLA – SenderTESLA – Sender

timeInterval i-1 Interval i+2Interval i+1Interval i

KiKi+2Ki+1Ki-1

F(Ki) F(Ki+1) F(Ki+2) F(Ki+3)

K’i-1 K’i K’i+1K’i+2

F’(Ki) F’(Ki+1) F’(Ki+2) F’(Ki+3)

Pj Pj+1 Pj+2 Pj+3MiKi-1

MAC(Mi, Ki’)

Pi-1

Mi+1Ki

MAC(Mi+1, Ki+1’)

Mi+2Ki+1

MAC(Mi+2, Ki+2’)

Mi-1Ki-2

MAC(Mi-1, Ki-1’)

Pi+2Pi+1Pi

• Pre-compute a sequence of key values using one-way hash functions or pseudo-random functions. Kn = F(Kn-1), …, K1 = F(K2)• Use another hash function to compute K’. Ki’ = F’(Ki)• Generate MAC using K’ and Message M• Send packet P. Pi = <Mi, Ki-d, MACi>

Page 14: Geo-Encryption Using Loran

TESLA – ReceiverTESLA – Receiver

• The receiver buffers the packet• Each receiver also checks that the disclosed key is correct using self-authentication and previously released keys• checks the correctness of the MAC of buffered packets that sent in the time interval of disclosed key• If the MAC is correct, the receiver accepts the packet•Message Sequence is arbitrary

Mi-1Ki-2

MAC(Mi-1, K’i-1)

Mi+1Ki

MAC(Mi+1, K’i+1)

MiKi-1

MAC(Mi, K’i)

authenticated authenticated afterreception of Pi+d

not yet authenticated

Pi-1 Pi+1Pi

Page 15: Geo-Encryption Using Loran

Loran TransmissionLoran Transmission

Time

Master MasterStation W Station YStation X

Station WGroup Repetition Interval (GRI)

Time Difference (TD)

GRI range40,000 ~ 99,990 μsec

GRI range40,000 ~ 99,990 μsec

Page 16: Geo-Encryption Using Loran

Loran Data Channel Communication-- Ninth-Pulse Modulation

Loran Data Channel Communication-- Ninth-Pulse Modulation

Master

SecondaryModulated 9th pulse

Loran Signal

Page 17: Geo-Encryption Using Loran

Loran Modulation Technique-- Pulse Position Modulation

Loran Modulation Technique-- Pulse Position Modulation

0 50 100 150 200 250 300 350 400-1

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

1Time Domain of 32-state PPM

μsec

Each GRI can carry 5 symbols

))(*2.0sin()()()

65)(2

(2i

dt

ii dtedttsi

−−=−−

πModulated Loran Pulse:

Page 18: Geo-Encryption Using Loran

Matched Filter ModelMatched Filter Model

matched to delay 1

matched to delay 2

matched to delay 32)()( tntsi +

)()( tntsi +

)()( tntsi + ∫ +−+= ττττ dtTsnsty i )()]()([)( 11

∫ +−+= ττττ dtTsnsty i )()]()([)( 22

∫ +−+= ττττ dtTsnsty i )()]()([)( 33

Comparator… …

Page 19: Geo-Encryption Using Loran

Loran MessagesLoran Messages

45…1194…440…3Bit assignment75414Length (bits)

ParityPayloadTypeSection

Time of day111115

Undefined0011 thru 11103-14

Message for government use only00102

Almanac00011

Differential Phase Correction00000

DescriptionType codeType

Page 20: Geo-Encryption Using Loran

How to Implement TESLA?How to Implement TESLA?

Certified Loran Receiver• embed K0 inside the receiver• capable to compute station

dependent keys and timedependent keys

• keys can’t be recovered• Synchronized with Loran

stations

Certified Loran Receiver• embed K0 inside the receiver• capable to compute station

dependent keys and timedependent keys

• keys can’t be recovered• Synchronized with Loran

stations

Loran Station Dependent

Time Dependent

K0

…KFallon KGeorge KSearchlight

…K0(ti-1) K0(ti) K0(ti+n)

…K1(ti) K2(ti) Km(ti)TESLA key sequence

K0(ti) K0(ti+n)

time

Page 21: Geo-Encryption Using Loran

Station Dependent Keys GenerationStation Dependent Keys Generation

Random Key K0MD5(‘E4DAE8F68387ABF329F1E183B4F38EF6’)

‘e4224c00d5a648ffe65f325f80bcad3f’16-bytes16-bytes

MD5(‘Fallon’)‘96ddabca419f7153f2c0ed0cba63a9e4’

MD5(‘Middletown’)‘e1edacbd3f1982411ec85566099fcc19’

‘98D9A6C04F977E55FAC3E50BB068A6E7’‘98D9A6C04F977E55FAC3E50BB068A6E7’ ‘EFE9A1B731118D4716CB5D610394C31A’‘EFE9A1B731118D4716CB5D610394C31A’

Page 22: Geo-Encryption Using Loran

Proposed Authentication SchemeProposed Authentication Scheme

• SHA-256• HMAC – 256-bit output, minimum key size 128 bits• 384/41 ~ 10 messages/TESLA packet

F F F

0 1 2 3 15 0 1 0 1 2 3 15 0 110 msg 10 msg

Ki+1 Ki+2 Ki+3

Page 23: Geo-Encryption Using Loran

How TESLA Enhances Security?How TESLA Enhances Security?

CertifiedLoran Receiver

CertifiedLoran Receiver

• Keys are not right• Can’t verify MACs

Repositionthe pulses

• Keys are delayed• Can’t verify MACs

Ki Ki+1 Ki+2

Ki

delay due to repositioning the pulses

Page 24: Geo-Encryption Using Loran

Time of AlarmTime of Alarm

30 35 40 45 50 55 60 65 70 7530

40

50

60

70

80

90

100

110Authentication Performance -- without Message Loss

Authenticated Bandwidth Percentage (%)

Aut

hent

icat

ion

Tim

e (s

ec)

TESLA d = 1DSA

• consider GRI 9990

• DSA size > MAC size

• DSA verification hashigh computationoverhead

Page 25: Geo-Encryption Using Loran

TESLA PerformanceTESLA PerformancePPM32 Probability of Symbol Error

-2 0 2 4 6 8 1010-7

10-6

10-5

10-4

10-3

10-2

10-1

100

SNR(dB)

Pro

babi

lity

Erro

r Rat

e

PPM 32 Level

AnalyticalSimulated

0 0.05 0.1 0.15 0.210-10

10-8

10-6

10-4

10-2

100

Analytical Message Loss vs. Packet Loss

Average Packet Loss

Mes

sage

Los

s

Message Loss

n

jt

j

k

RS

jnjn

tj

q

qjn

qerrorectedunP

ppjn

failuredecordererror

)1()1()_det(

)1()_/Pr(

0

1

−⎟⎟⎠

⎞⎜⎜⎝

⎛−

=

−⎟⎟⎠

⎞⎜⎜⎝

⎛=

=

+=

Page 26: Geo-Encryption Using Loran

Time of Alarm with Message LossTime of Alarm with Message Loss

35 40 45 50 55 60 65 70

50

100

150

200

250

300

350

TESLA Authentication Performance -- with Message Loss

Authenticated Bandwidth Percentage (%)

Aut

hent

icat

ion

Tim

e (s

ec)

SNR = 0SNR = 1SNR = 2SNR = 3SNR = 4SNR = 5SNR = 6SNR = 7SNR = 8SNR = 9SNR = 10w/o message loss

Page 27: Geo-Encryption Using Loran

Next Step: Geo-lock DesignNext Step: Geo-lock Design

plaintext

Geo-lock MappingGeo-lock Mapping

Geo-lock MappingGeo-lock Mapping

Recipient’sLocation Signature

Recipient’sLocation Signature

Loran ReceiverLoran Receiver

AESAES

plaintextRandom KeyRandom Key

Symmetric key

RSARSA

Encrypted key

RSARSA

Key_E

Key_D

AESAES

plaintext

ciphertext

Symmetric key

Signal AuthenticationSignal Authentication

YesNo

Page 28: Geo-Encryption Using Loran

Geolock Mapping functionGeolock Mapping functionLa

titu

de

B4124809 AF4534E7 281841BD 60AB7CFA

576F4595 C8F3262A 4E18CC0A 43653816

11AE2637 B8323B7F 952E3574 43D264E8

7C09A4D6 1482C152 124C1214 266B1F6D

E3D73F28 A4054068 93919767 6E76ED2A

EEAB8B2B FE8205A7 F82C9516 FC6D27DD

814CCF71 1DABFD91 85383231 F2F7218C

95CBDC2C 28DBB56E AAD8DF8E 78120469

Longitude

E61014C 955FC38 5DC67F29 BE15DD27

D58860CE 82DECE41 D3A8378E 127506C0

D6F02579 499D9599 588DA916 68A95323

F0E74523 5DF41C17 93F35661 14527F1D

Time

Possible Parameters

• ECD• TD• TDOA• Envelope shape

Page 29: Geo-Encryption Using Loran

Pseudo-random SequencePseudo-random Sequence

cryptologyLinear complexity

Spread spectrum communicationsNavigationSystem test and analysis

Good cross-correlation

Range and navigationSpread spectrum communicationsScrambling

Good auto-correlation

Application areaRequirement

Page 30: Geo-Encryption Using Loran

Keystream GeneratorKeystream Generator

Geffe generator: a keystream generator using three LFSRs, combined in a nonlinear manner.

LFSR 1, a1

LFSR 2, a2

LFSR 3, a3

Multiplexer

(a1^a2) (( a1)^a3)

Page 31: Geo-Encryption Using Loran

Steps to Build a TestbedSteps to Build a Testbed

Recipient’s location Mapping Algorithm Geo-lock

Geo-lock Generation (Matlab Simulation)

Loran Messages

Authenticating Message Generation• time-dependent key• TESLA key sequence• HMAC• complete one TESLA packet

Modulate Loran data and authenticating message

on 9th pulse using PPM32DAC

ground

Loran Front-end ADC

Demodulation

Verify authenticating messageyes

Loran location signature

Mapping Algorithm

Geo-lock

Page 32: Geo-Encryption Using Loran

Testbed SetupTestbed Setup

Modulated Loran Signal(MATLAB simulation) ICS-660

Loran H-field antenna

power

ICS-652/ICS-650

Page 33: Geo-Encryption Using Loran

Thank You!Thank You!