zaštita informacionih sistema, milan milosavljević 1 kontrola pristupa

Zaštita informacionih sistema, Milan Milosavljević 1

Kontrola pristupa


Kontrola pristupa Kontrola pristupa se sastoji iz dva dela Autentifikacija: Ko pristupa?

o Odredjuje se kome je dopušten pristupo Autentifikacija čoveka od strane mašineo Authentifikacija mašine od strane mašine

Autorizacija: Da vam li je dozvoljeno da nešto uradite?o Kad vam je dozvoljen pristup, šta možete da uradite?o Obezbedjuje ogrsničenja na mguće akcije

Primedba: Kontrola pristupa se često koristi kao sinonim za autorizaciju


Autentifikacija


Ko pristupa?

Kako mašina može da autentifikuje čoveka? Autentifikacija može biti zasnovana…

o Na nečemu što znate Npr., lozinke

o Na nečemu što imate Npr., smartkartica

o Na osnovu nečega što jeste Npr., otisak prsta


Nešto što znate Lozinke Što šta može da bude lozinka!

o PINo Matični brojo Majčino devojačko prezimeo Datum rodjenjao Ime vašeg kućnog ljubimca, i td.


Problemi sa lozinkama “Lozinke su jedan od najvećih praktičnih

problema sa kojima se susreću inženjeri bezbednosti danas.”

“Ljudi ne poseduju sposobnost bezbednog memorisanja kriptografskih ljučeva visokog kvalitete, i imaju neprihvatljivu brzinu i tačnost u obavljanju kriptogrfskih operacija.”


Zašto lozinke?

Zašto je “nešto što znam” popularnije od “nečeg što imam” i “nečeg što jesam”?

Cena: lozinke su besplatne Pogodnost: jednostavnije je resetovati

lozinke nego izdati korisniku novi otisak prsta


Ključevi i lozinke

Kripto ključevi Neka je ključ 64 bita Tada imamo 264

različitih ključeva Izabrati ključ slučajno Tada napadač mora

da isproba oko 263 ključeva

Lozinke Neka je lozinka od 8

karaktera, i neka ima 256 različitih karaktera

Tada je 2568 = 264 lozinki Korisnici ne biraju lozinke

slučajno Napadač mora da isproba

daleko manje lozinki od 263 (napad pomoću rečnika)


Dobre i loše lozinke Loše lozinke

o franko Fidoo passwordo 4444o Pikachuo 102560o AustinStamp

Dobre lozinke?o jfIej,43j-EmmL+yo 09864376537263o P0kem0No FSa7Yagoo 0nceuP0nAt1m8o PokeGCTall150


Eksperiment sa lozinkama Postoje tri grupe korisnika svakoj grupi je

savetovano da izabere lozinke na sledeći način

o Grupa A: Najmanje 6 karaktera, 1 neslovnio Grupa B: Lozinka zasnovana na frazio Grupa C: 8 slučajnih karaktera

Rezultati:o Grupa A: Oko 30% lozinki je lako razbitio Grupa B: Oko 10% lozinki se razbija

Lozinke se lako pamteo Grupa C: Oko 10% se razbija

Lozinke se teško pamte

winner


Eksperiment sa lozinkama Saglasnost korisnika je teško ostvariti U svakom slučaju, 1/3 nije saglasna (i oko 1/3

ovih se lako razbija!) Ponekad je najbolje dodeliti lozinke Ako lozinke nisu unapred dodeljene, najbolji

saveti pri izboru suo Izaberite lozinku zasnovanu na frazamao Koristiti posebne alate za krekovanje slabih lozinkio Zahteva se periodična zamena lozinki?


Napadi na lozinke Napadač može…

o Ciljati smo jedan poseban nalogo Ciljati bilo koji nalogo Ciljati bilo koji nalog na bilo kom sistemuo Pokušaj napada odbijanja servisa ( denial of

service - DoS) Uobičajeni redosled napada

o Spoljašnji korisnik normalni korisnik administrator

o Potrebna je možda samo jedna slaba lozinka!


Uzastopno probanje lozinki Pretpostavimo da se sistem zaključa

nakon 3 pogrešne lozinke. Koliko dugo treba da bude zaključan?o 5 sekundio 5 minutao Dok sistem administrator ne obnovi servise

Šta su pozitivne, a šta negativne strane?


Fajl lozinki Memorisanje lozinki u fajlovima je loše

rešenje Potreban nam je mehanizam za verifikovanje

lozinki Kriptografsko rešenje: heširanje lozinke

o Zapisati y = h(lozinka)o Možemo verifikovati lozinku hešngomo Ukoliko napadač ima fajl lozinki, time nije dobio i

same lozinkeo Napadač koji ima fajl lozinki, može da pokuša da

pogodi x za koje je y = h(x)o Ako uspe, napadač je pronašao lozinku!


Napad pomomoću rečnika

Napadač unapred izračuna h(x) za sve x u rečniku uobičajenih lozinki

Neka napadač ima pristup fajlu hešovanih lozinkio Napadač treba samo da poredi hešove sa

češovima već izračunatim na osnovu rečnikao Svi naredni napadi se mogu obaviti na isti način

Da li se može osujetiti ovakav napad? Ili barem, posao napadača učiniti težim?


Fajl lozinki Sadrži heširane lozinke Bolje je memorisati hešove sa salt-om Za zadatu lozinku, izabrati slučajan s, i

izračunati y = h(lozinka, s)

i memorisati (s,y) u fajlu lozinki Primedba: salt s nije tajan Lozinka se lako verifikuje Napadač mora da izračuna hešove rečnika

lozinki za svakog korisnika-što je mnogo posla!


Razbijanje lozinki:proračun

Pretpostavke Lozinka ima 8 karaktera, 128 mogućnosti po

karakteruo Tada je 1288 = 256 mogućih lozinki

Neka postoji fajl lozinki sa 210 lozinki Napadač poseduje rečnik sa 220 najčešće

korišćenih lozinki Verovatnoća da je data lozinka u rečniku iznosi

1/4 Posao koji je potrebno obaviti se meri brojem

heširanja


Razbijanje lozinki Napad na 1 lozinku bez rečnika

o Mora se isprobati 256/2 = 255 u srednjemo Analogno potpunoj pretrazi ključeva

Napad na 1 lozinku sa rečnikomo Očekivani posao je oko

1/4 (219) + 3/4 (255) = 254.6

o Ali u praksi, isprbati ceo rečnik i završiti ako se ne nadje rešenje posao je najviše 220 i verovatnoća uspeha je 1/4


Razbijanje lozinki Napad na bilo kojih 1024 lozinki u fajlu Bez rečnika

o Neka je svih 210 lozinki različito o Potrebno je 255 poredjenja pre nego što se očekuje

da pronadjemo pravu lozinkuo Ako se ne koristi salt, svako računanje heša daje

210 poredjenja očekivani posao (broj heševa) je 255/210 = 245

o Ako se koristi salt, očekivani posao je 255 budući da svako poredjenje zahteva novo računanje heša


Razbijanje lozinki Ako napadamo bilo koju loyinku u fajlu Sa rečnikom

o Verovatnoća da je barem jedna lozinka u rečniku je 1 - (3/4)1024 = 1

o Ignorišemo slučaj da u rečniku uopšte nema tražene lozinke

o Ako nema salta, posao je oko 219/210 = 29

o Ako ima salta, očekivani posao je manji od 222

o Primetimo da ako nema salta, možemo da preračunamo sve hešove za dati rečnik smanjujući ovaj posao


Ostala pitanja oko lozinki Isuviše lozinki za pamćenje

o Rezultuje u ponovnoj upotrebi lozinkio Zašto je ovo problem?

Ko pati zbog loših lozinki? o Lozinka za logovanje vs ATM PIN

Neuspešna promena difolt lozinki Socialni inženjering Logovi pogrešnih lozinki mogu sadžati skoro

ispravnu lozinku Bagovi, keystroke logging, spyware, itd.


Lozinke Definitivni zaključci Razbijanje lozinki je isuviše jednostavno!

o Lozinka u trajanju od jedne nedelje može srušiti sistem bezbednosti

o Korisnici biraju loše lozinkeo Napad zasnovan na socijalnom inženjeringu itd.

Loši momci imaju sve prednosti na svojoj strani

Sva natematika favorizuje loše momke Lozinke su veliki problem sigurnosti


Alati za razbijanje lozinki Popularni alati za razbijanje lozinki

o Password Crackerso Password Portalo L0phtCrack and LC4 (Windows)o John the Ripper (Unix)

Adminstratori treba da testiraju ove alate, budući da će ih napadači sigurno upotrebiti!

Dobar članak o razbijanju lozinki jeo Passwords - Conerstone of Computer Security

http://www.passwordportal.net/

http://www.atstake.com/products/lc/

http://www.openwall.com/john/

http://www.securityfocus.com/infocus/1192




Biometrika


Nešto što jesmo Biometrika

o “Vi ste vaš ključ” Schneier

Are

Know Have

Primerio Otisak prstao Potpiso Prepoznavanje licao Prepoznavanje govorao Prepoznavanje načina hodao “Digitalni pas” (prepoznavanje

mirisa)o I još mnogo što šta!


Zašto biometrika? Biometrika je vidjena kao poželjna zamena za

lozinke Potrebna je jeftina i pouzdana biometrika Danas je to vrlo aktivna oblast istraživanja Biometrika se danas koristi u bezbednosnim

sistemimao Miš sa senzorom za otisak palcao Otisak dlana za kontrolu pristupao Otisak prsta za otključavanje kola, vrata i td.

Medjutim, biometrika nije toliko popularnao Još uvek nije dostigla očekivanja


Idealna biometrika Universalnost primenljiva je na (skoro) svakog

o U praksi, ne postoji biometrika koja se može primeniti na svakog

Razlikovanje razlikovanje sa sigurnošćuo U praksi se ne možemo nadati 100% sigurnosti

Permanentnost upotrebljene i izmerene fizičke karakteristike ne bi trebale da se ikada promeneo U praksi se ovaj zahtev odnosi na odredjeni dugački

vremenski period Sakupljivost lako se mere i sakupljaju

o Zavisi os stepena kooperativnosti subjekata Sigurna, jednostavna za upotrebu, i td.


Biometrijski modeli Identifikacija Ko je tamo?

o Poredjenje jedan prema mnogimao Primer: FBI-ova baza otisaka prstiju

Autentifikacija Da li si to zaista ti?o Poredjenje jedan prema jedano Primer: Miš sa ćitačem otiska palca

Problem identifikacije je znatno težio Više “slučajnih” poklapanja usled mnogih

poredjenja Pozabavićemo se autentifikacijom


Upisivanje vs prepoznavanje Faze upisivanja

o Biometrijski podaci subjekta se pamte u bazi podataka

o Porebno je pažljivo izmeriti tražene podatkeo Ponekad je ovaj posao spor i zahteva ponovljena

merenjao Merenja moraju biti vrlo precizna za dobro

prepoznavanjeo Ovo je slaba tačka mnogih biometrijskih sistema

Faza prepoznavanjao Biometrijska detekcija u praksio Mora biti brza i jednostavnao Mora biti dovoljno tačno


Cooperativni subjekt Pretpostavlja se kooperativnost subjekta U problemu identifikacije obično

imamonekooperativnog subjekta Na primer, prepoznavanje lica

o Predloženo za upotrebu u Las Vegas kazinima za detekciju poznatih prevaranata

o Takodje se koristi za detekciju terorista na aerodromima i td.

o Verovatno ne postoje idealni uslovi za prijavljivanjeo Subjekt će verovatno pokušati da zbuni sistem

prepoznavanja Kooperativni subjekt čini ovu fazu mnogo

lakšom!o U autentifikaciji, subjekt je kooperativan


Biometrijske greške Stepen prevare versus stepen uvrede

o Prevara korisnik A se pogrešno autentifikuje kao korisnik B

o Uvreda korisnik A se ne autentifikuje kao korisnik A Za bilo koju biometriku, možemo smanjiti prevaru

ili uvredu, ali će ona druga vrednos biti povećana Primer

o 99% poklapanje glasa niska prevar, visoka uvredao 30% poklapanje glasa visoka prevara, niska uvreda

Jednake greške: pri tome je prevara == uvredio Najbolja mera za poredjenje biometrika


Istorija otisaka prstiju 1823 Profesor Johannes Evangelist

Purkinje diskutuje 9 razlićitih oblika otisaka prstiju

1856 Sir William Hershel koristi otisak prsta za potpisivanje ugovora

1880 Dr. Henry Faulds piše rad u Nature otiscima prstiju za identifikaciju

1883 u delu Mark Twain-a Life on the Mississippi ubica je identifikovan preko otiska prstiju


Istorija otisaka prstiju

1888 Sir Francis Galton (Darvinov rodjak) je razvio klasifikacioni sistemo Njegov sistem “minutia” je još uvek u upotrebio Verifikuje da se otisci prstiju ne menjaju sa

starenjem Neke zemlje propisuju broj tačaka (tj. minutia)

za verodostojnost identifikacije u kriminalnim slučajevimao U Britaniji je to 15 tačakao U Americi nije propisan fiksan broj tačaka


Poredjenje otisaka prstiju

Petlja (dvostruka) Vrtlog Luk

Primeri petlji, vrtloga i lukova Minutia se ekstrahuju iz ovih obeležja


Biometrika otiska prstiju

Snimanje slike otiska Izoštravanje slike Identifikacija minutia


Biometrika otiska prstiju

Ekstrahovane minutia se porede sa minutia memorisanim u bazi podataka

Da li se radi o statističkom poklapanju?


Geometrija dlana Popularni oblik biometrije Meri se oblik dlana

o Širina dlana, prstijuo Dužina prstiju, itd.

Ljudski dlanovi nisu jedinstveni

Geometrija dlana je dovoljna za mnoge primene

Pogodna za autentifikaciju Nije pogodna za

problematiku identifikacije


Geometrija dlana Prednosti

o Brzinao 1 minuta je dovoljna za priavljivanjeo 5 sekundi za prepoznavanjeo Dlanovi su simetrični (korišćenje druge ruke)

Nedostacio Ne može se koristiti za vrlo mlade i vrlo stare

osobeo Relativno visoka greška jednakosti


Oblik Irisa

Šara irisa je prilično “haotična” Mali ili gotovo nikakav uticaj genetike Različita čak i za identične blizanve Šara je stabilna kroz celokupan životni vek


Prepoznavanje irisa: istorijat 1936 sugerisan od strane Frank

Burch-a 1980s filmovi James Bond-a 1986 pojava prvog patenta na ovu

temu 1994 John Daugman je patentirao

najbolji savremeni sistemo Vlasnik patenta je Iridian Technologies


Skeniranje irisa Skener locira iris Uzima se b/w fotografija Koriste se polarne

koordinate… Računa se 2-D wavelet

transformacija Dobija se 256 bajtova iris

koda


Merenje sličnosti irisa Zasniva se na Hamming-ovom rastojanju Definiše se d(x,y) na sledeći način

o # ne poklapajućoh bita/# brojem poredjenih bitao d(0010,0101) = 3/4 and d(101111,101001) = 1/3

Računa se d(x,y) na 2048-bitsko iris koduo Perfektno poklapanje daje rastojanje d(x,y) = 0o Za identičan iris, očekivano rastojanje je 0.08o Za slućajne nizove, očekivano rastojanje je 0.50o Poklapanje se prihvata, ako je rastojanje manje od 0.32


Greška Iris skenera

rastojanje

0.29 1 in 1.31010

0.30 1 in 1.5109

0.31 1 in 1.8108

0.32 1 in 2.6107

0.33 1 in 4.0106

0.34 1 in 6.9105

0.35 1 in 1.3105

rastojanje greška

: greška jednakosti


Napad na skeniranje irisa Dobra fotografija oka se može skenirati Napadač može upotrebiti fotografiju oka

Jedna avganistanska žena je autentifikovana na osnovu skeniranja stare fotografije okao Priča o tome se može naći na here

Da bi se osujetio foto napad, skener bi mogao da koristi svetko da bi biosiguran da je iris “živ”

http://news.bbc.co.uk/1/hi/world/south_asia/1870382.stm


Poredjenje po kriterijumu kednakih grešaka (Equal Error Rate)

Equal error rate (EER): verovatnoća prevare == verovatnoći uvrede

Biometrika na bazi otisaka prstiju EER oko 5% Geometrija dlana : EER oko 10-3

Teoretski, skeniranje irisa ima EER oko 10-6

o U praksi je ovo teško ostvaritio Faza prijave mora biti ekstremno tačna

Većina biometrika je znatno lošija od otisaka prstiju!

Biometrika je korisna za autentifikaciju… Ali je gotovo neupotrebljiva za identifikaciju


Biometrika: zaključak Biometrika teško podleže prevarama Ali i dalje postoje mogući napadi

o Ukrasti Alisin prsto Fptpkopirati Bobov otisak prsta, oko, i td.o Izvršiti subverziju softvera, baza odataka, “puta

poverenja”, … Kako povući “razbijenu” biometriku? Biometrika nije otporna na podvale! Biometrika se danas ograničeno koristi Očekuje se promena u ovom pogledu u bližoj

budućnosti…


Nešto što posedujemo Nešto u vašem posedu Primeri

o Ključevi od kolao Laptop računar

Ili specifična MAC adresa

o Generator lozinkio ATM kartice, smartkartice, itd.


Generator lozinki

Alisa dobija “challenge” R od Bob-a Alisa unosi R u generator lozinki Alisa “odgovara” Bob-u Alisa ima generator lozinki i zna PIN

Alisa Bob

1. “Ja sam Alisa”

2. R

5. F(R)

3. PIN, R

4. F(R)

Generator lozinki


2-faktorsa autentifikacija

Zahteva 2 od 3 stavke1. Nešto što znate2. Nešto što imate3. Nešto što ste vi po svojoj prirodi

Primerio ATM: Kartica i PINo Kreditna kartica: Kartica i potpiso Generator lozinki: Uredjaj i PINo Smartkartica sa lozinkom/PIN


Single Sign-on Velika je nepogodnost unositi lozinke često

o Korisnici žele autentifikaciju samo jedanputo “Credentials” stay with user wherever he goeso Subsequent authentication is transparent to user

Single sign-on for the Internet?o Microsoft: Passporto Everybody else: Liberty Allianceo Security Assertion Markup Language (SAML)


Web Cookies Cookie is provided by a Website and stored on

user’s machine Cookie indexes a database at Website Cookies maintain state across sessions Web uses a stateless protocol: HTTP Cookies also maintain state within a session Like a single sign-on for a website

o Though a very weak form of authentication Cookies and privacy concerns


Authorization


Authentication vs Authorization

Authentication Who goes there?o Restrictions on who (or what) can access system

Authorization Are you allowed to do that?o Restrictions on actions of authenticated users

Authorization is a form of access control Authorization enforced by

o Access Control Listso Capabilities


Lampson’s Access Control Matrix

rx rx r --- ---

rx rx r rw rw

rwx rwx r rw rw

rx rx rw rw rw

OSAccounting

programAccounting

dataInsurance

dataPayrolldata

Bob

Alice

Sam

Accountingprogram

Subjects (users) index the rows Objects (resources) index the columns


Are You Allowed to Do That?

Access control matrix has all relevant info But how to manage a large access control (AC)

matrix? Could be 1000’s of users, 1000’s of resources Then AC matrix with 1,000,000’s of entries Need to check this matrix before access to any

resource is allowed Hopelessly inefficient


Access Control Lists (ACLs) ACL: store access control matrix by column Example: ACL for insurance data is in blue

rx rx r --- ---

rx rx r rw rw

rwx rwx r rw rw

rx rx rw rw rw

OSAccounting

programAccounting

dataInsurance

dataPayrolldata

Bob

Alice

Sam

Accountingprogram


Capabilities (or C-Lists) Store access control matrix by row Example: Capability for Alice is in red

rx rx r --- ---

rx rx r rw rw

rwx rwx r rw rw

rx rx rw rw rw

OSAccounting

programAccounting

dataInsurance

dataPayrolldata

Bob

Alice

Sam

Accountingprogram


ACLs vs Capabilities

Access Control List Capability

Note that arrows point in opposite directions! With ACLs, still need to associate users to files

file1

file2

file3

file1

file2

file3

r---r

Alice

Bob

Fred

wr

---

rwrr

Alice

Bob

Fred

rwrw

---rr

r---r


Confused Deputy Two resources

o Compiler and BILL file (billing info)

Compiler can write file BILL

Alice can invoke compiler with a debug filename

Alice not allowed to write to BILL

Access control matrix

x ---

rx rw

Compiler BILL

Alice

Compiler


ACL’s and Confused Deputy

Compiler is deputy acting on behalf of Alice Compiler is confused

o Alice is not allowed to write BILL Compiler has confused its rights with Alice’s

Alice BILL

Compiler

debug

filename BILLBILL


Confused Deputy

Compiler acting for Alice is confused There has been a separation of authority from

the purpose for which it is used With ACLs, difficult to avoid this problem With Capabilities, easier to prevent problem

o Must maintain association between authority and intended purpose

o Capabilities make it easy to delegate authority


ACLs vs Capabilities ACLs

o Good when users manage their own fileso Protection is data-orientedo Easy to change rights to a resource

Capabilitieso Easy to delegateo Easy to add/delete userso Easier to avoid the confused deputyo More difficult to implemento The “Zen of information security”

Capabilities loved by academicso Capability Myths Demolished

http://www.cis.upenn.edu/~KeyKOS/ConfusedDeputy.html

http://www.erights.org/elib/capability/duals/myths.html


Multilevel Security (MLS) Models


Classifications and Clearances Classifications apply to objects Clearances apply to subjects US Department of Defense uses 4

levels of classifications/clearancesTOP SECRETSECRETCONFIDENTIALUNCLASSIFIED


Clearances and Classification

To obtain a SECRET clearance requires a routine background check

A TOP SECRET clearance requires extensive background check

Practical classification problemso Proper classification not always clearo Level of granularity to apply classificationso Aggregation flipside of granularity


Subjects and Objects

Let O be an object, S a subjecto O has a classificationo S has a clearanceo Security level denoted L(O) and L(S)

For DoD levels, we haveTOP SECRET > SECRET > CONFIDENTIAL > UNCLASSIFIED


Multilevel Security (MLS) MLS needed when subjects/objects at

different levels use same system MLS is a form of Access Control Military/government interest in MLS for many

decades o Lots of funded research into MLSo Strengths and weaknesses of MLS relatively well

understood (theoretical and practical)o Many possible uses of MLS outside military


MLS Applications Classified government/military information Business example: info restricted to

o Senior management onlyo All managemento Everyone in companyo General public

Network firewallo Keep intruders at low level to limit damage

Confidential medical info, databases, etc.


MLS Security Models

MLS models explain what needs to be done Models do not tell you how to implement Models are descriptive, not prescriptive

o High level description, not an algorithm There are many MLS models We’ll discuss simplest MLS model

o Other models are more realistico Other models also more complex, more difficult to

enforce, harder to verify, etc.


Bell-LaPadula BLP security model designed to express

essential requirements for MLS BLP deals with confidentiality

o To prevent unauthorized reading Recall that O is an object, S a subject

o Object O has a classificationo Subject S has a clearanceo Security level denoted L(O) and L(S)


Bell-LaPadula

BLP consists ofSimple Security Condition: S can read O if

and only if L(O) L(S)

*-Property (Star Property): S can write O if and only if L(S) L(O)

No read up, no write down


McLean’s Criticisms of BLP McLean: BLP is “so trivial that it is hard to

imagine a realistic security model for which it does not hold”

McLean’s “system Z” allowed administrator to reclassify object, then “write down”

Is this fair? Violates spirit of BLP, but not expressly

forbidden in statement of BLP Raises fundamental questions about the nature

of (and limits of) modeling


B and LP’s Response BLP enhanced with tranquility property

o Strong tranquility property: security labels never changeo Weak tranquility property: security label can only change if it

does not violate “established security policy” Strong tranquility impractical in real world

o Often want to enforce “least privilege”o Give users lowest privilege needed for current worko Then upgrade privilege as needed (and allowed by policy)o This is known as the high water mark principle

Weak tranquility allows for least privilege (high water mark), but the property is vague


BLP: The Bottom Line

BLP is simple, but probably too simple BLP is one of the few security models that

can be used to prove things about systems BLP has inspired other security models

o Most other models try to be more realistico Other security models are more complexo Other models difficult to analyze and/or apply in

practice


Biba’s Model BLP for confidentiality, Biba for integrity

o Biba is to prevent unauthorized writing Biba is (in a sense) the dual of BLP Integrity model

o Spse you trust the integrity of O but not Oo If object O includes O and O then you cannot trust

the integrity of O Integrity level of O is minimum of the integrity

of any object in O Low water mark principle for integrity


Biba Let I(O) denote the integrity of object O and

I(S) denote the integrity of subject S Biba can be stated as

Write Access Rule: S can write O if and only if I(O) I(S)(if S writes O, the integrity of O that of S)

Biba’s Model: S can read O if and only if I(S) I(O)(if S reads O, the integrity of S that of O)

Often, replace Biba’s Model withLow Water Mark Policy: If S reads O, then

I(S) = min(I(S), I(O))


BLP vs Biba

level

high

low

L(O)

L(O) L(O)

Confidentiality

BLP

I(O)

I(O)

I(O)

Biba

level

high

lowIntegrity


Multilateral Security (Compartments)


Multilateral Security Multilevel Security (MLS) enforces access

control up and down Simple hierarchy of security labels may not be

flexible enough Multilateral security enforces access control

across by creating compartments Suppose TOP SECRET divided into TOP

SECRET {CAT} and TOP SECRET {DOG} Both are TOP SECRET but information flow

restricted across the TOP SECRET level


Multilateral Security

Why compartments?o Why not create a new classification level?

May not want either ofo TOP SECRET {CAT} TOP SECRET {DOG}o TOP SECRET {DOG} TOP SECRET {CAT}

Compartments allow us to enforce the need to know principleo Regardless of your clearance, you only have access

to info that you need to know


Multilateral Security Arrows indicate “” relationship

Not all classifications are comparable, e.g.,TOP SECRET {CAT} vs SECRET {CAT, DOG}

TOP SECRET {CAT, DOG}

TOP SECRET {CAT}

TOP SECRET

SECRET {CAT, DOG}

SECRET {DOG}

SECRET

TOP SECRET {DOG}

SECRET {CAT}


MLS vs Multilateral Security MLS can be used without multilateral security or

vice-versa But, MLS almost always includes multilateral Example

o MLS mandated for protecting medical records of British Medical Association (BMA)

o AIDS was TOP SECRET, prescriptions SECRETo What is the classification of an AIDS drug?o Everything tends toward TOP SECRETo Defeats the purpose of the system!

Multilateral security was used instead


Covert Channel


Covert Channel MLS designed to restrict legitimate channels

of communication May be other ways for information to flow For example, resources shared at different

levels may signal information Covert channel: “communication path not

intended as such by system’s designers”


Covert Channel Example Alice has TOP SECRET clearance, Bob has

CONFIDENTIAL clearance Suppose the file space shared by all users Alice creates file FileXYzW to signal “1” to

Bob, and removes file to signal “0” Once each minute Bob lists the files

o If file FileXYzW does not exist, Alice sent 0o If file FileXYzW exists, Alice sent 1

Alice can leak TOP SECRET info to Bob!


Covert Channel Example

Alice:

Time:

Create file Delete file Create file Delete file

Bob: Check file Check file Check file Check fileCheck file

Data: 1 0 1 01


Covert Channel Other examples of covert channels

o Print queueo ACK messageso Network traffic, etc., etc., etc.

When does a covert channel exist?1. Sender and receiver have a shared resource2. Sender able to vary property of resource that

receiver can observe3. Communication between sender and receiver can

be synchronized


Covert Channel Covert channels exist almost everywhere Easy to eliminate covert channels…

o Provided you eliminate all shared resources and all communication

Virtually impossible to eliminate all covert channels in any useful systemo DoD guidelines: goal is to reduce covert channel

capacity to no more than 1 bit/secondo Implication is that DoD has given up trying to

eliminate covert channels!


Covert Channel Consider 100MB TOP SECRET file

o Plaintext version stored in TOP SECRET placeo Encrypted with AES using 256-bit key, ciphertext

stored in UNCLASSIFIED location Suppose we reduce covert channel capacity

to 1 bit per second It would take more than 25 years to leak

entire document thru a covert channel But it would take less than 5 minutes to leak

256-bit AES key thru covert channel!


Real-World Covert Channel

Hide data in TCP header “reserved” field Or use covert_TCP, tool to hide data in

o Sequence numbero ACK number


Real-World Covert Channel Hide data in TCP sequence numbers Tool: covert_TCP Sequence number X contains covert info

A. Covert_TCPsender

C. Covert_TCP receiver

B. Innocent server

SYNSpoofed source: CDestination: BSEQ: X

ACK (or RST)Source: BDestination: CACK: X


Inference Control


Inference Control Example

Suppose we query a databaseo Question: What is average salary of female CS

professors at SJSU?o Answer: $95,000o Question: How many female CS professors at

SJSU?o Answer: 1

Specific information has leaked from responses to general questions!


Inference Control and Research

For example, medical records are private but valuable for research

How to make info available for research and protect privacy?

How to allow access to such data without leaking specific information?


Naïve Inference Control

Remove names from medical records? Still may be easy to get specific info from

such “anonymous” data Removing names is not enough

o As seen in previous example What more can be done?


Less-naïve Inference Control Query set size control

o Don’t return an answer if set size is too small N-respondent, k% dominance rule

o Do not release statistic if k% or more contributed by N or fewer

o Example: Avg salary in Bill Gates’ neighborhoodo Used by the US Census Bureau

Randomizationo Add small amount of random noise to data

Many other methods none satisfactory


Inference Control: The Bottom Line

Robust inference control may be impossible Is weak inference control better than no

inference control?o Yes: Reduces amount of information that leaks and

thereby limits the damage

Is weak crypto better than no crypto?o Probably not: Encryption indicates important datao May be easier to filter encrypted data


CAPTCHA


Turing Test Proposed by Alan Turing in 1950 Human asks questions to one other human

and one computer (without seeing either) If human questioner cannot distinguish the

human from the computer responder, the computer passes the test

The gold standard in artificial intelligence No computer can pass this today


CAPTCHA CAPTCHA Completely Automated Public

Turing test to tell Computers and Humans Apart

Automated test is generated and scored by a computer program

Public program and data are public Turing test to tell… humans can pass the

test, but machines cannot pass the test Like an inverse Turing test (sort of…)


CAPTCHA Paradox “…CAPTCHA is a program that can generate

and grade tests that it itself cannot pass…” “…much like some professors…” Paradox computer creates and scores test

that it cannot pass! CAPTCHA used to restrict access to

resources to humans (no computers) CAPTCHA useful for access control


CAPTCHA Uses? Original motivation: automated “bots” stuffed

ballot box in vote for best CS school Free email services spammers used bots sign

up for 1000’s of email accountso CAPTCHA employed so only humans can get accts

Sites that do not want to be automatically indexed by search engineso HTML tag only says “please do not index me” o CAPTCHA would force human intervention


CAPTCHA: Rules of the Game Must be easy for most humans to pass Must be difficult or impossible for machines to

passo Even with access to CAPTCHA software

The only unknown is some random number Desirable to have different CAPTCHAs in

case some person cannot pass one typeo Blind person could not pass visual test, etc.


Do CAPTCHAs Exist? Test: Find 2 words in the following

Easy for most humans Difficult for computers (OCR problem)


CAPTCHAs Current types of CAPTCHAs

o Visual Like previous example Many others

o Audio Distorted words or music

No text-based CAPTCHAso Maybe this is not possible…


CAPTCHA’s and AI

Computer recognition of distorted text is a challenging AI problemo But humans can solve this problem

Same is true of distorted soundo Humans also good at solving this

Hackers who break such a CAPTCHA have solved a hard AI problem

Putting hacker’s effort to good use!


Firewalls


Firewalls

Firewall must determine what to let in to internal network and/or what to let out

Access control for the network

InternetInternalnetworkFirewall


Firewall as Secretary A firewall is like a secretary To meet with an executive

o First contact the secretaryo Secretary decides if meeting is reasonableo Secretary filters out many requests

You want to meet chair of CS department?o Secretary does some filtering

You want to meet President of US?o Secretary does lots of filtering!


Firewall Terminology No standard terminology Types of firewalls

o Packet filter works at network layero Stateful packet filter transport layero Application proxy application layero Personal firewall for single user, home

network, etc.


Packet Filter Operates at network layer Can filters based on

o Source IP addresso Destination IP addresso Source Porto Destination Porto Flag bits (SYN, ACK, etc.)o Egress or ingress

application

transport

network

link

physical


Packet Filter Advantage

o Speed Disadvantages

o No stateo Cannot see TCP connectionso Blind to application data

application

transport

network

link

physical


Packet Filter Configured via Access Control Lists (ACLs)

o Different meaning of ACL than previously

Allow Inside Outside Any 80 HTTP

Allow Outside Inside 80 > 1023 HTTP

Deny All All All All All

Action

Source IP

Dest IP

Source

Port

Dest Port Protoco

l

Intention is to restrict incoming packets to Web responses

Any

ACK

All

FlagBits


TCP ACK Scan

Attacker sends packet with ACK bit set, without prior 3-way handshake

Violates TCP/IP protocol ACK packet pass thru packet filter firewall

o Appears to be part of an ongoing connection

RST sent by recipient of such packet Attacker scans for open ports thru firewall


TCP ACK Scan

PacketFilter

Trudy InternalNetwork

ACK dest port 1207

ACK dest port 1208

ACK dest port 1209

RST

Attacker knows port 1209 open thru firewall A stateful packet filter can prevent this (next)

o Since ACK scans not part of established connections


Stateful Packet Filter

Adds state to packet filter Operates at transport layer Remembers TCP connections

and flag bits Can even remember UDP

packets (e.g., DNS requests)

application

transport

network

link

physical


Stateful Packet Filter Advantages

o Can do everything a packet filter can do plus...

o Keep track of ongoing connections

Disadvantageso Cannot see application datao Slower than packet filtering

application

transport

network

link

physical


Application Proxy

A proxy is something that acts on your behalf

Application proxy looks at incoming application data

Verifies that data is safe before letting it in

application

transport

network

link

physical


Application Proxy Advantages

o Complete view of connections and applications data

o Filter bad data at application layer (viruses, Word macros)

Disadvantageo Speed

application

transport

network

link

physical


Application Proxy Creates a new packet before sending it thru

to internal network Attacker must talk to proxy and convince it to

forward message Proxy has complete view of connection Prevents some attacks stateful packet filter

cannot see next slides


Firewalk Tool to scan for open ports thru firewall Known: IP address of firewall and IP address

of one system inside firewallo TTL set to 1 more than number of hops to firewall

and set destination port to No If firewall does not let thru data on port N, no

responseo If firewall allows data on port N thru firewall, get

time exceeded error message


Firewalk and Proxy Firewall

Dest port 12345, TTL=4



Time exceeded

Trudy

Packetfilter

Router

This will not work thru an application proxy The proxy creates a new packet, destroys old TTL

RouterRouter


Personal Firewall

To protect one user or home network Can use any of the methods

o Packet filtero Stateful packet filtero Application proxy


Firewalls and Defense in Depth Example security architecture

Internet

Intranet withPersonalFirewalls

PacketFilter

ApplicationProxy

DMZ

FTP server

DNS server

WWW server


Intrusion Detection Systems


Intrusion Prevention

Want to keep bad guys out Intrusion prevention is a traditional focus of

computer securityo Authentication is to prevent intrusionso Firewalls a form of intrusion preventiono Virus defenses also intrusion prevention

Comparable to locking the door on your car


Intrusion Detection In spite of intrusion prevention, bad guys will

sometime get into system Intrusion detection systems (IDS)

o Detect attackso Look for “unusual” activity

IDS developed out of log file analysis IDS is currently a very hot research topic How to respond when intrusion detected?

o We don’t deal with this topic here


Intrusion Detection Systems Who is likely intruder?

o May be outsider who got thru firewallo May be evil insider

What do intruders do?o Launch well-known attackso Launch variations on well-known attackso Launch new or little-known attackso Use a system to attack other systemso Etc.


IDS

Intrusion detection approacheso Signature-based IDSo Anomaly-based IDS

Intrusion detection architectureso Host-based IDSo Network-based IDS

Most systems can be classified as aboveo In spite of marketing claims to the contrary!


Host-based IDS Monitor activities on hosts for

o Known attacks oro Suspicious behavior

Designed to detect attacks such aso Buffer overflowo Escalation of privilege

Little or no view of network activities


Network-based IDS Monitor activity on the network for

o Known attackso Suspicious network activity

Designed to detect attacks such aso Denial of serviceo Network probeso Malformed packets, etc.

Can be some overlap with firewall Little or no view of host-base attacks Can have both host and network IDS


Signature Detection Example Failed login attempts may indicate password

cracking attack IDS could use the rule “N failed login attempts

in M seconds” as signature If N or more failed login attempts in M

seconds, IDS warns of attack Note that the warning is specific

o Admin knows what attack is suspectedo Admin can verify attack (or false alarm)


Signature Detection Suppose IDS warns whenever N or more

failed logins in M seconds Must set N and M so that false alarms not

common Can do this based on normal behavior But if attacker knows the signature, he can try

N-1 logins every M seconds! In this case, signature detection slows the

attacker, but might not stop him


Signature Detection

Many techniques used to make signature detection more robust

Goal is usually to detect “almost signatures” For example, if “about” N login attempts in

“about” M secondso Warn of possible password cracking attempto What are reasonable values for “about”?o Can use statistical analysis, heuristics, othero Must take care not to increase false alarm rate


Signature Detection Advantages of signature detection

o Simpleo Detect known attackso Know which attack at time of detectiono Efficient (if reasonable number of signatures)

Disadvantages of signature detectiono Signature files must be kept up to dateo Number of signatures may become largeo Can only detect known attackso Variation on known attack may not be detected


Anomaly Detection Anomaly detection systems look for unusual

or abnormal behavior There are (at least) two challenges

o What is normal for this system?o How “far” from normal is abnormal?

Statistics is obviously required here!o The mean defines normalo The variance indicates how far abnormal lives

from normal


What is Normal? Consider the scatterplot below

x

y

White dot is “normal”

Is red dot normal? Is green dot

normal? How abnormal is

the blue dot? Stats can be tricky!


How to Measure Normal?

How to measure normal?o Must measure during “representative”

behavioro Must not measure during an attack…o …or else attack will seem normal!o Normal is statistical meano Must also compute variance to have any

reasonable chance of success


How to Measure Abnormal? Abnormal is relative to some “normal”

o Abnormal indicates possible attack Statistical discrimination techniques:

o Bayesian statisticso Linear discriminant analysis (LDA)o Quadratic discriminant analysis (QDA)o Neural nets, hidden Markov models, etc.

Fancy modeling techniques also usedo Artificial intelligenceo Artificial immune system principleso Many others!


Anomaly Detection (1) Spse we monitor use of three commands:

open, read, close Under normal use we observe that Alice

open,read,close,open,open,read,close,… Of the six possible ordered pairs, four pairs are

“normal” for Alice:(open,read), (read,close), (close,open), (open,open)

Can we use this to identify unusual activity?


Anomaly Detection (1)

We monitor use of the three commands open, read, close

If the ratio of abnormal to normal pairs is “too high”, warn of possible attack

Could improve this approach by o Also using expected frequency of each pairo Use more than two consecutive commandso Include more commands/behavior in the modelo More sophisticated statistical discrimination


Anomaly Detection (2) Over time, Alice has

accessed file Fn at rate Hn

H0 H1 H2 H3

.10 .40 .40 .10

Is this “normal” use? We compute S = (H0A0)2+(H1A1)2+…+(H3A3)2 = .02 And consider S < 0.1 to be normal, so this is normal Problem: How to account for use that varies over

time?

Recently, Alice has accessed file Fn at rate An

A0 A1 A2 A3

.10 .40 .30 .20


Anomaly Detection (2) To allow “normal” to adapt to new use, we

update long-term averages as

Hn = 0.2An + 0.8Hn

Then H0 and H1 are unchanged, H2=.2.3+.8.4=.38 and H3=.2.2+.8.1=.12

And the long term averages are updated as

H0 H1 H2 H3

.10 .40 .38 .12


Anomaly Detection (2) The updated long

term average is

H0 H1 H2 H3

.10 .40 .38 .12

Is this normal use? Compute S = (H0A0)2+…+(H3A3)2 = .0488 Since S = .0488 < 0.1 we consider this

normal And we again update the long term

averages by Hn = 0.2An + 0.8Hn

New observed rates are…

A0 A1 A2 A3

.10 .30 .30 .30


Anomaly Detection (2) The starting

averages were

H0 H1 H2 H3

.10 .40 .40 .10

The stats slowly evolve to match behavior This reduces false alarms and work for admin But also opens an avenue for attack… Suppose Trudy always wants to access F3 She can convince IDS this is normal for Alice!

After 2 iterations, the averages are

H0 H1 H2 H3

.10 .38.364

.156


Anomaly Detection (2) To make this approach more robust, must

also incorporate the variance Can also combine N stats as, for example,

T = (S1 + S2 + S3 + … + SN) / Nto obtain a more complete view of “normal”

Similar (but more sophisticated) approach is used in IDS known as NIDES

NIDES includes anomaly and signature IDS


Anomaly Detection Issues System constantly evolves and so must IDS

o Static system would place huge burden on admin o But evolving IDS makes it possible for attacker to

(slowly) convince IDS that an attack is normal!o Attacker may win simply by “going slow”

What does “abnormal” really mean?o Only that there is possibly an attacko May not say anything specific about attack!o How to respond to such vague information?

Signature detection tells exactly which attack


Anomaly Detection Advantages

o Chance of detecting unknown attackso May be more efficient (since no signatures)

Disadvantageso Today, cannot be used aloneo Must be used with a signature detection systemo Reliability is unclearo May be subject to attacko Anomaly detection indicates something unusualo But lack of specific info on possible attack!


Anomaly Detection: The Bottom Line

Anomaly-based IDS is active research topic Many security professionals have very high

hopes for its ultimate success Often cited as key future security technology Hackers are not convinced!

o Title of a talk at Defcon 11: “Why Anomaly-based IDS is an Attacker’s Best Friend”

Anomaly detection is difficult and tricky Is anomaly detection as hard as AI?


Access Control Summary

Authentication and authorizationo Authentication who goes there?

Passwords something you know Biometrics something you are (or “you

are your key”)


Access Control Summary

Authorization are you allowed to do that?o Access control matrix/ACLs/Capabilitieso MLS/Multilateral securityo BLP/Bibao Covert channelo Inference controlo CAPTCHAo Firewallso IDS


Coming Attractions… Security protocols

o Generic authentication protocolso SSLo IPSeco Kerberoso GSM

We’ll see lots of crypto applications in the next chapter

zaštita informacionih sistema, milan milosavljević 1 kontrola pristupa

Documents