Cryptography and Security - discipline.elcom.pub.rodiscipline.elcom.pub.ro/cscn/slides/cscn_01_intr_2018.pdf · ⚫ Security policies are sets of rules that regulate how the resources

Cryptography and Security

in Communication Networks

Introduction

ETTI - Master - Advanced Wireless Communications

© Octavian Catrina 2

Security concepts

value

countermeasures

vulnerabilities

IT resources

Threat agents

(Attackers) risk

wish to minimize

to

wish to abuse or/and damage

give rise to

to

imposeto reduce

threats

that exploit

that increase

leading to

may be aware of

that may be reduced by

Owners

Based on Common Criteria for IT Security Evaluation, 1999

⚫ The purpose of information

system security is to protect

IT resources against threats

related to malicious or other

human activities.

⚫ Security policies are sets of

rules that regulate how the

resources are managed and

protected.

⚫ A threat is a potential abuse

of protected resources (i.e.,

a breach of a security policy).

⚫ Threats are enabled by

system vulnerabilities.

⚫ Countermeasures are used to

reduce vulnerabilities and to

meet security policies.


Security services and mechanisms

⚫ Security services

⚫ A security service is a measure used to address a threat:

⚫ Data confidentiality: prevents unauthorized data disclosure.

⚫ Data integrity/authentication: detects unauthorized data

modification (and falsification of data origin).

⚫ Access control: prevents unauthorized use of resources.

⚫ Entity authentication: prevents impersonation (user/device).

⚫ Security mechanisms

⚫ A security mechanism is a means to provide a service:

⚫ Encryption schemes.

⚫ Message authentication (and digital signature) schemes.

⚫ Authorization and access control protocols and systems.

⚫ Authentication protocols.

⚫ Example: a data confidentiality service can be implemented

using an access control system or an encryption scheme.


Access control (AC)

⚫ Access control service

⚫ Provides protection against unauthorized use of resources:

⚫ intercepts user requests,

⚫ determines if they are authorized or not, and

⚫ allows or denies them, based on access control policies.

⚫ The AC system needs a verifiable proof that the user is

authorized to perform the requested action.

⚫ Example: The AC system authenticates Alice and then uses

her (verified) identity to look up her permissions in an access

control list stored in a local repository.

Access

Control

Access RequestAuthorized

Access Request

Allows or denies a requestaccording to AC policies

ResourcesAlice


Data confidentiality

⚫ Data confidentiality service

⚫ Prevents unauthorized disclosure of the information

(i.e., unauthorized information reading).

⚫ Methods

⚫ Access control: Physical protection to deny access to the data

storage system or to the communication channel.

⚫ Cryptography - Encryption: Apply a transformation (encryption)

such that: (1) protected data becomes indistinguishable from

random bits; (2) the inverse transformation (decryption) can be

computed only by an authorized reader.

Adversary

(Charlie)

Intercepts messages

and wants to read

them

Data confidentiality in a communication system.

Alice and Bob make sure that Charlie cannot

understand intercepted messages.

Alice Bob


Data integrity

⚫ Data integrity service

⚫ Prevents or detects unauthorized modification of the

information (i.e., unauthorized information writing).

⚫ Modification can be: insertion, deletion, and substitution.

⚫ Methods

⚫ Access control: Physical protection to deny access to the data

storage system or to the communication channel.

⚫ Cryptography - message integrity (authentication) code:

Append a tag to the message, computed as a function of the

data and a secret known only by the authorized users.

Adversary

(Charlie)

Intercepts messages

and tries to alter or

delete them

Bob can detect any modification of the

messages sent by Alice.

Alice Bob


Data-origin authentication

⚫ Data-origin authentication (message authentication)

⚫ Enables verification of the identity of the message originator,

based on evidence associated with the message.

Adversary

(Charlie)

Tries to insert

messages and claim

they come from Alice

Bob can make sure that a received message

was written by Alice, not fabricated by Charlie.

⚫ Methods

⚫ Cryptography - message authentication

code or digital signatures.

Authentication

Integrity

⚫ Implies data integrity (when a message

is modified its origin changes).

Data authentication in a communication system

Alice Bob


Entity authentication

⚫ Entity authentication service

⚫ Enable one party to verify the identity of another party, through

interactive acquisition of evidence.

Also proves that the other party is currently present (active).

⚫ Unilateral or mutual authentication.

⚫ An entity can be authenticated based on ...... what it knows (password, PIN),

what it has (magnetic card, smart card),

what it is (biometrics: fingerprints, iris recognition, ...),

or a combination of the above (multi-factor authentication).

Adversary

(Charlie)

Tries to impersonate

Alice (interactively)

Bob is assured that he

communicates with Alice.

Example:

Unilateral authentication.

Alice Bob


Non-repudiation

⚫ Non-repudiation service

⚫ Extension of authentication, which provides protection against

denial of previous commitments or actions.

E.g., deny having sent a message.

⚫ Allows resolution of disputes using a trusted third party which

can check document authenticity.

Non-repudiation

Authentication

Integrity

⚫ Non-repudiation implies both

data-origin authentication and

data integrity.

⚫ The basic objectives are similar

to those of handwritten signature

Digital signature techniques.

A6E3891F2939E38C745B

25289896CA345BEF5349

245CBA653448E349EA47

Digital signature:Handwritten signature:


Digital signatures

⚫ Digital signature schemes provide message

authentication with non-repudiation

Adversary (Charlie)

Tries to insert or modify

messages forging Alice's

signature.

Alice:

Cannot deny having created

and signed the message.

Bob:

Can check the integrity of

the message and the fact

that was signed by Alice.

Cannot forge the signature

or modify the message.

Trent:

Trusted third party for resolution of disputes.

Can check the integrity of the message and

the fact that it was signed by Alice.


Threats and attacks

⚫ Security threat

⚫ Potential violation of a security policy. E.g., vulnerabilities due

to flaws in system design, implementation, and operation.

⚫ Security attack

⚫ A realization of a threat: an action that violates security.

⚫ Two main classes: Passive attacks. Active attacks.

⚫ Passive attack

⚫ The adversary can only monitor the communication channel

(read messages, but not modify or delete them).

⚫ Threatens data confidentiality.

⚫ Active attack

⚫ The adversary attempts to delete, add, or otherwise alter the

transmission on the communication channel.

⚫ Threatens all security services.


Attacks: Interception

⚫ Interception (eavesdropping): attack on confidentiality

⚫ Unauthorized data access. Passive attack.

⚫ Basic capabilities of the adversary: receive data, packet sniffer.

⚫ Easiest for wireless channels: a standard interface within

reception range is sufficient.

⚫ Otherwise, requires network access (links, switches, routers,

hosts). May be either physical access (connect to the network,

wiretapping), or remote access (e.g., via backdoor, etc.).

⚫ Easier for broadcast multi-access network. Otherwise, modify

the data path to include a device controlled by the adversary.

Interception

(eavesdropping)

Sender

Adversary

Receiver


Attacks: Fabrication

⚫ Fabrication: attack on authenticity

⚫ Unauthorized use of an identity, generation and distribution of

messages under that identity. Active attack.

⚫ Basic capabilities: send data, impersonate legitimate party.

⚫ MAC addresses and IP addresses can be spoofed.

⚫ Easiest for wireless network: a standard interface within

transmission range is sufficient.

⚫ Otherwise, requires access to the receiver's network, either

physical (connect to the network), or remote (e.g., via Internet

connection).

Sends messages

under false

identity

Sender

Adversary

Receiver


Attacks: Modification

⚫ Modification: attack on integrity and authenticity

⚫ Intercept and modify or delete messages. Active attack.

⚫ Basic capabilities: send and receive data, deny packet delivery

for legitimate parties, impersonate legitimate party.

⚫ MAC and IP addresses can be spoofed.

⚫ Install a device on the normal data path or acquire an existing

device (switch, router, HTTP/DNS proxy, etc.).

⚫ Use some network indirection trick to modify the data path and

include the attacker's device: e.g., ARP or DNS spoofing, etc.

Modification

(tampering with

messages)

Sender

Adversary

Receiver


Attacks: Interruption

⚫ Interruption: attack on availability

⚫ Stop message flow or prevent service access. Active attack.

Interruption

Sender

Adversary

Receiver

⚫ Easy to mount, hardest to thwart. Many variants.

⚫ Basic capabilities: depend on variant.

⚫ Denial of service (DoS): use up computation or communication

resources, corrupt/delete data in transit. Requires the ability to

send to receiver, or control a network device on the data path.

⚫ Damage hardware or software. Requires physical access.

⚫ Cryptography can help a bit both the attack and the defense (!).


Example: ARP spoofing 1/2

⚫ ARP spoofing: An adversary forges ARP packets in

order to control connectivity in LANs

⚫ Redirection using ARP spoofing⚫ An attacker M makes host A map IPB to a MAC address MACB:

⚫ Impersonation: map IPB to MACM.

⚫ Denial of service: map IPB to non-existing MAC address.

Host A Attacker M

M can now impersonate B (for A-B communications)

ARP Cache:

IPB at MACM

Host B

ARP request:

Who has IPA? Tell (IPB, MACM)

ARP reply: ...

MAC frame from MACA to MACM

containing IP packet from IPA to IPB

MAC frame from MACM to MACA

containing IP packet from IPB to IPA

M poisons the ARP cache of host A

This is an example that works with

typical ARP implementations. Some

ARP implementations are vulnerable

to simpler attacks.

E.g., a server

or a router


ARP spoofing 2/2

⚫ Man-In-The-Middle (MITM) attack using ARP spoofing⚫ Enables MITM attacks against authentication and key

exchange protocols. Enables sniffing in switched LANs.

Host A MITM Attacker M

M records & forwards the packets exchanged by A and B

ARP Cache:

IPB at MACM

Host B

ARP request:

Who has IPA? Tell (IPB, MACM) ARP Cache:

IPA at MACM

ARP request:

Who has IPB? Tell (IPA, MACM)

ARP reply: ... ARP reply: ...

MAC frame from MACA to MACM

containing IP packet from IPA to IPBMAC frame from MACM to MACB

containing IP packet from IPA to IPB

MAC frame from MACM to MACA


MAC frame from MACB to MACM


M poisons the ARP caches of hosts A, B

After the attack, M should un-poison the

ARP caches to avoid detection


Example: IP spoofing and DoS

⚫ IP (address) spoofing attacks

⚫ Send IP packets with forged source addresses.

⚫ To impersonate hosts (e.g., trusted hosts, to deceive a firewall).

⚫ To hide the origin of an attack.

⚫ To set up DoS (denial of service) attacks.

General scenario

Host y

Victim

Host z

(if any)

Spoofed packet

SA=z DA=y

Reply packet

(if any)

Intercept

or ignore

Attacker

Host x

In general the address z can take any value:

unicast or broadcast or multicast or invalid;

global or private; assigned or not.

Reflection ("smurf") DoS attack

Host y

Victim 1

Host z

Victim 2

Spoofed packet

SA=z DA=y

Attacker

Host x

Reply packets

Use host y to attack host or network z.

Amplification: the reply is much larger than the

original packet; or y is a broadcast address.

Other variants exist.


Case study: WLAN 1/2

⚫ Fact -1: Anybody within transmission range can

send/receive to/from WLAN radio channel.

AccessPoint (AP)

Wired

LAN

WLAN

⚫ Threat 1.2: Data integrity and origin violation.

⚫ Attack: Fabrication - Address spoofing. Modification - rogue AP (active).

⚫ Policy: Detect and drop frames from unauthorized senders.

⚫ Service: Data-origin authentication.

⚫ Mechanism (802.11): Integrity check mechanism in WEP.

⚫ Threat 1.1: Confidentiality violation

(unauthorized reception of WLAN traffic).

⚫ Attack: Eavesdropping (passive).

⚫ Policy: Prevent unauthorized reception.

⚫ Service: Data confidentiality.

⚫ Mechanism (802.11): Encryption using

WEP/RC4.


Case study: WLAN 2/2

AccessPoint (AP)

Wired

LAN

WLAN

⚫ Threat 1.3: Access control violation (unauthorized network access).

⚫ Attack: Connect to WLAN, access network resources (active).

⚫ Policy: Deny unauthorized access.

⚫ Service: Authentication, authorization, and access control.

⚫ Mechanism (802.11): Authentication protocol based on WEP and

authenticated encryption (WEP).

⚫ Fact -2: Vulnerabilities due to design flaws in 802.11

security services.

⚫ Threat 2.1: Flaw in authentication protocol.

⚫ Attack: Eavesdrop and replay (active).

Defeats authentication (traffic blocked by

authenticated encryption).

⚫ Threat 2.2: WEP/RC4 vulnerability.

⚫ Attack: Key recovery (cryptanalytic attack).

⚫ Breaks all services, violates all policies.

⚫ Fixed in 802.11i: New design of security services, upgraded mechanisms.

Etc.


Next steps

⚫ Understand cryptographic algorithms

⚫ Encryption, data authentication, signature, ...

⚫ What they can do. What they cannot do.

⚫ How do they work. How can they be attacked.

⚫ How to use them correctly!

⚫ Authentication and key management protocols

⚫ How are they designed?

⚫ How can they be attacked?

⚫ Then put them all together: Security protocols

⚫ Secure channels.

⚫ Virtual Private Networks (VPNs).

⚫ Etc.

Cryptography and Security - discipline.elcom.pub.rodiscipline.elcom.pub.ro/cscn/slides/cscn_01_intr_2018.pdf · ⚫ Security policies are sets of rules that regulate how the resources

Documents