Lecture 20Trusted Computing and Multilevel Security

modified from slides of Lawrie Brown and Hesham El-Rewini

Computer Security Models• two fundamental computer security facts:

• all complex software systems have eventually revealed flaws or bugs that need to be fixed

• it is extraordinarily difficult to build computer hardware/software not vulnerable to security attacks

• problems involved both design and implementation– led to development of formal security models

2

Trusted OS Functions

3

Multilevel Security (MLS)• RFC 2828 defines multilevel security as follows:“A class of system that has system resources (particularly stored information) at more than one security level (i.e., has different types of sensitive resources) and that permits concurrent access by users who differ in security clearance and need-to-know, but is able to prevent each user from accessing resources for which the user lacks authorization.”

4

Bell-LaPadula (BLP) Model• formal model for access control

– developed in 1970s• subjects and objects are assigned a security class

– a subject has a security clearance– an object has a security classification– form a hierarchy and are referred to as security

levels• top secret > secret > confidential > restricted >unclassified

– security classes control the manner by which a subject may access an object

5

BLP Model Access Modes• READ

– the subject is allowed only read access to the object• APPEND

– the subject is allowed only write access to the object• WRITE

– the subject is allowed both read and write access to the object

• EXECUTE– the subject is allowed neither read nor write access

to the object but may invoke the object for execution6

Multi-Level Security• no read up

– subject can only read an object of less or equal security level

– referred to as the simple security property • ss-property

• no write down– a subject can only write into an object of greater

or equal security level– referred to as the *-property

7

Multi-Level Security

8

Multi-Level Security• ds-property : An individual (or role) may

grant to another individual (or role) access to a document – based on the owner’s discretion,

• constrained by the MAC rules

• site policy overrides any discretionary access controls

9

BLP Formal Description• based on current state of system (b, M, f, H):

– current access set b: triples of (s, o, a)• subject s has current access to object o in access mode a

– access matrix M: matrix of Mij

• access modes of subject Si to access object Oj

– level function f: security level of subjects and objects• fo ( Oj ) is the classification level of object Oj

• fs ( Si ) is the security clearance of subject Si

• fc ( Si ) is the current security level of subject Si

– hierarchy H: a directed rooted tree of objects10

BLP Formal Description• three BLP properties:

– ss-property: (Si, Oj, read) has fc (Si) ≥ fo (Oj)

– *-property: (Si, Oj, append) has fc (Si) ≤ fo (Oj) and

(Si, Oj, write) has fc (Si) = fo (Oj)

– ds-property:current (Si, Oj, Ax) implies Ax M[Si Oj]

• BLP gives formal theorems– theoretically possible to prove system is secure– in practice usually not possible

11

Illustration

o1

s1 o2

o3

s2 o4

o5

Low

High

o5

s2

Object

Subject

12

Append

Append

ReadRead

Write

Read

Write

Read

BLP Rules1 •get access

2 •release access

3 •change object level

4 •change current level

5 •give access permission

6 •rescind access permission

7 •create an object

8 •delete a group of objects

13

BLP Example

*-property

BLP Example

ss-property and *-property

16

BLP Example

“downgrade” in a controlled and monitored manner

“classification creep”: information from a range of sources and levels

“covert channels”: (untrusted) low classified executable dataallowed to be executed by a high clearance (trusted) subject

Integrity ?

Limitations to the BLP Model• Incompatibility of confidentiality and integrity within a

single MLS system– MLS can work either for powers or for secrets but not for both– This mutual exclusion excludes some interesting power and

integrity centered technologies from being used effectively in BLP style MLS environments

• Cooperating conspirator problem in the presence of covert channels– In the presence of shared resources the *-property may

become unenforceable• In essence, the BLP model effectively breaks down when (untrusted)

low classified executable data are allowed to be executed by a high clearance (trusted) subject

19

Biba Integrity Model• strict integrity policy:

– Modify: To write or update information in an object– Observe: To read information in an object– Execute: To execute an object– Invoke: Communication from one subject to another– simple integrity: I(S) ≥ I(O)– integrity confinement: I(S) ≤ I(O)– invocation property: I(S1) ≥ I(S2)

20

Clark-Wilson Integrity Model• closely models commercial operations

– Well-formed transactions• A user should not manipulate data arbitrarily

– Separation of duty among users• A person who create or certify a well-formed transaction

may not execute it

21

Clark-Wilson Integrity Model• Principal components of the model

– Constrained data items (CDIs)• Subject to strict integrity controls

– Unconstrained data items (UDIs)• Unchecked data items

– Integrity verification procedures (IVPs): • Intended to assure that all CDIs conform to some

application-specific model of integrity and consistency– Transformation procedures (TPs):

• System transactions that change the set of CDIs from one consistent state to another

22

Clark-Wilson Integrity Model

23

Chinese Wall Model• integrity and confidentiality• use both discretionary and mandatory access

– Subjects: Active entities that may wish to access protected objects

– Information: Information organized into a hierarchy • Objects: Individual items of information, each concerning a

single corporation• Dataset (DS): All objects that concern the same corporation• Conflict of interest (CI) class: All datasets whose

corporations are in competition– Access rules: Rules for read and write access

24

Chinese Wall Model

Simple security rule: A S can read O only if• O is in the same DS as an object already accessed by S, OR• O belongs to a CI from which S has not yet accessed any information

*-property rule: A S can write an O only if• S can read O according to the simple security rule, AND• All objects that S can read are in the same DS as O. sanitized data

25

Terminology Related to Trust• Trust: The extent to which someone who relies on a system can have confidence that

the system meets its specifications – i.e., that the system does what it claims to do and does not perform unwanted functions

• Trusted system: A system believed to enforce a given set of attributes to a stated degree of assurance.

• Trustworthiness: Assurance that a system deserves to be trusted, such that the trust can be guaranteed in some convincing way, – such as through formal analysis or code review.

• Trusted computer system: A system that employs sufficient hardware and software assurance measures to allow its use for simultaneous processing of a range of sensitive or classified information.

• Trusted computing base (TCB): A portion of a system that enforces a particular policy– The TCB must be resistant to tampering and circumvention. – The TCB should be small enough to be analyzed systematically.

• Assurance: A process that ensures a system is developed and operated as intended by the system's security policy.

• Evaluation: Assessing whether the product has the security properties claimed for it.• Functionality: The security features provided by a product.

26

Reference Monitors

Properties• Complete mediation• Isolation• Verifiability

o Trustworthy system

27

Trojan Horse Defense

Trusted Platform Module (TPM)

• concept from Trusted Computing Group • hardware module at heart of hardware/

software approach to trusted computing (TC)• uses a TPM chip

– motherboard, smart card, processor– working with approved hardware/software– generating and using crypto keys

• has three basic services: • authenticated boot, certification, encryption

28

29

TPM Functions

Authenticated Boot Service• responsible for booting entire OS in stages and ensuring

each is valid and approved for use– at each stage digital signature associated with code is verified– TPM keeps a tamper-evident log of the loading process

• log records versions of all code running– can then expand trust boundary to include additional

hardware and application and utility software– confirms component is on the approved list, is digitally

signed, and that serial number hasn’t been revoked• result is a configuration that is well-defined with

approved components

30

Certification Service• once a configuration is achieved and logged the TPM

can certify configuration to others– can produce a digital certificate– confidence that configuration is unaltered because:

• TPM is considered trustworthy• only the TPM possesses this TPM’s private key

• include challenge value in certificate to also ensure it is timely

• provides a hierarchical certification approach– hardware/OS configuration– OS certifies application programs– user has confidence in application configuration

31

Encryption Service• encrypts data so that it can only be decrypted by a

machine with a certain configuration• TPM maintains a master secret key unique to machine

– used to generate secret encryption key for every possible configuration of that machine

• can extend scheme upward– provide encryption key to application so that decryption can

only be done by desired version of application running on desired version of the desired OS

– encrypted data can be stored locally or transmitted to a peer application on a remote machine

32

33

Protected Storage Function

Security Assurance• “…degree of confidence that the security controls operate correctly and protect the system as intended. • Assurance is not, however, an absolute guarantee

that the measures work as intended.”

34

Assurance and Evaluation• evaluators

– use the assurance requirements as criteria when evaluating security features and controls

– may be in the same organization as consumers or a third-party evaluation team

• assurance– deals with security features of IT

products– applies to:

• requirements• security policy• product design• product implementation• system operation

target audiences:

• consumers – select security features and

functions – determine the required levels of

security assurance

• developers– respond to security

requirements– interpret statements of

assurance requirements– determine assurance

approaches and level of effort

36

Scope of Assurance• system architecture

– addresses both the system development phase and the system operations phase

• system integrity– addresses the correct operation of the system hardware and firmware

• system testing– ensures security features have been tested thoroughly

• design specification and verification– addresses the correctness of the system design and implementation

with respect to the system security policy• covert channel analysis

– attempts to identify any potential means for bypassing security policy

37

Scope of Assurance• trusted facility management

– deals with system administration• trusted recovery

– provides for correct operation of security features after a system recovers from failures, crashes, or security incidents

• trusted distribution– ensures that protected hardware, firmware, and software do not go

through unauthorized modification during transit from the vendor to the customer

• configuration management– requirements are included for configuration control, audit,

management, and accounting

Assurance• Verification

– each of the system’s functions works correctly

• Validation– developer is building the right product

• according to the specification

• Testing– based on the actual product being evaluated,

• not on abstraction38

Testing• Observable effects versus internal structure• Can demonstrate existence of a problem,

– but passing tests does not imply absence of any• Hard to achieve adequate test coverage within

reasonable time– inputs & internal states

• hard to keep track of all states

• Penetrating Testing– tiger team analysis, ethical hacking

• Team of experts in design of OS tries to crack system39

Formal verification• The most rigorous method• Rules of mathematical logic to demonstrate

that a system has certain security property

• Proving a Theorem– Time consuming– Complex process

40

Entry

min A[1]i 1

i i + 1

i > n

min < A[i]

min A[i]

Exityes

noyes

no

Example: find minimum

41

Finding the minimum valueAssertions

P: n > 0 Q: n > 0 and1 i n and

min A[1]

R: n > 0 and S: n > 0 and1 i n and i = n + 1 and

for all j 1 j i -1 for all j 1 j i -1

min A[j] min A[j] 42

Validation• Requirements checking

– system does things it should do• also, system does not do things it is not supposed to do

• Design and code reviews– traceability from each requirement to design and

code components• System testing

– data expected from reading the requirement document can be confirmed in the actual running of the system

43

Common Criteria (CC)• Common Criteria for Information Technology and

Security Evaluation– ISO standards for security requirements and defining

evaluation criteria• aim is to provide greater confidence in IT product security

– development using secure requirements– evaluation confirming meets requirements– operation in accordance with requirements

• following successful evaluation a product may be listed as CC certified– NIST/NSA publishes lists of evaluated products

44

45

CC Requirements

• functional requirements– define desired security behavior

• assurance requirements– basis for gaining confidence that

the claimed security measures are effective and implemented correctly

• component – describes a specific set of

security requirements– smallest selectable set

common set of potential security requirements for use in evaluation

• target of evaluation (TOE) – refers to the part of product or

system subject to evaluation• class

– collection of requirements that share a common focus or intent

49

CC Security Paradigm

51

CC Assurance LevelsEAL 1: functionally tested

EAL 2: structurally tested

EAL 3: methodically tested and checked

EAL 4: methodically designed, tested, and reviewed

EAL 5: semi-formally designed and tested

EAL 6: semi-formally verified design and tested

EAL 7: formally verified design and tested

Evaluation• ensures security features work correctly and effectively

and show no exploitable vulnerabilities• performed in parallel with or after the development of TOE• higher levels entail: greater rigor, more time, more cost• principle input: security target, evidence, actual TOE• result: confirm security target is satisfied for TOE• process relates security target to high-level design, low-

level design, functional specification, source code implementation, and object code and hardware realization of the TOE

• degree of rigor and depth of analysis are determined by assurance level desired

52

53

Evaluation Parties and Phases• evaluation parties:

– sponsor - customer or vendor– developer - provides evidence for evaluation– evaluator - confirms requirements are satisfied– certifier - agency monitoring evaluation process

• monitored and regulated by a government agency in each country

• Common Criteria Evaluation and Validation Scheme (CCEVS)– operated by NIST and the NSA

54

Evaluation Parties and Phases• preparation:

– initial contact between sponsor and developer– conduct of evaluation:

• confirms satisfaction of security target– conclusion:

• final report is given to the certifiers for acceptance

Evaluation• Review: requirements, design, implementation,

assurance• US “Orange Book” Evaluation

– Trusted Computer System Evaluation Criteria (TCSEC)• European ITSEC Evaluation

– Information Technology Security Evaluation Criteria• US Combined Federal Criteria

– 1992 jointly buy NIST and NSA

55

Summary

• The Bell-LaPadula model for computer security

– Computer security models– General description– Formal description of model– Abstract operations– Example of BLP use– Implementation example-multics– Limitations to the BLP model

• Other formal models for computer security

– Biba integrity model– Clark-Wilson integrity model– Chinese wall model

• The concept of trusted systems– Reference monitors– Trojan horse defense

• Trusted computing and the trusted platform moduleo Authenticated boot serviceo Certification serviceo Encryption serviceo TPM functionso Protected storage

• Common criteria for information technology security evaluation

– Requirements– Profiles and targets– Example of a protection profile

• Assurance and evaluationo Target audienceo Scope of assuranceo Common criteria evaluation assurance

levelso Evaluation

Role-Based Access Control

58

RBAC Elements

59

RoleHierarchy

UserAssignments

60

Database Classification

Table

Column

61

Database Classification

Row

Element

Database Security: Read Access• DBMS enforces simple security rule (no read up)• easy if granularity is entire database or at table

level• inference problems if have column granularity

– if can query on restricted data can infer its existence• SELECT Ename FROM Employee WHERE Salary > 50K

– solution is to check access to all query data• also have problems if have row granularity

– null response indicates restricted/empty result• no extra concerns if have element granularity

62

Database Security: Write Access• enforce *-security rule (no write down)• have problem if a low clearance user wants to

insert a row with a primary key that already exists in a higher level row:– can reject, but user knows row exists– can replace, compromises data integrity– polyinstantiation and insert multiple rows with

same key, creates conflicting entries• same alternatives occur on update• avoid problem if use database/table granularity

63

Example of Polyinstantiation

64