Overview

Authentication: The process of assuring that the communicating entity or the origin of a piece of info is the one that it claims to be

Types:

  • Entity authentication
    • For connection oriented communication
    • Communicating entity is an entity involved in communication
    • Mechanism: Password, challenge, response and biometrics
  • Data origin authentication
    • For connectionless communcation
    • Communicating entity is the origin of a piece of info
    • data origin authneticity implies data integrity
    • Mechanisms: MAC or digital signature

Threats

CS2107-3-5.PNG

Authenticity and integrity

  • Authentic: The claimed entity/origin is assured by supporting evidence
  • Authenticity: The condition of being authentic
  • Authenticity and integrity: Are they related? Yes

Data origin authenticity implies data integrity

  • Authenticity is a stronger requriement than integrity

But the reverse is not true: Data integrity does not imply data origin authenticity

  • Some documents use the term integrity to mean authenticity
  • Some claim that authenticity does not necessarity give integrity
  • Pay attention to the context and the application involved

Examples

Over different communication channels:

CS2107-3-6.PNG

  • Alice get call
  • Claim to be from police
  • “Police” ask for info regarding brother

  • Alice log into luminNUS
  • Wonder if laptop is actually interacting with the real lumiNUS
  • Why is lumiNUS server convinced that the user logged in is ALICE

  • ALice tries to connect to WIFI using phone at NUH
  • The was item “NUS” listed in wifi list
  • Was that WIFI access point real?

Password (Weak authentication)

Stage 1: Bootstrapping

  • Server and user establish common password
  • Server keeps track of a file recording the identity and corresponding password

  • The password is to be established during bootstrapping
  • This can be done
    • Server choose password
    • Default password

      Question: Describe some bootstrapping mechanism that you have encountered

Stage 2: Authentication

  • Server authenticates an entity
  • If entity gives the correct password corresponding to the claim identity, the entity is deemed authentic

CS2107-3-8.PNG

CS2107-3-9.PNG

The identity does not need to be kept secret:

  • It could be: User name in computer system, bank accound no, customer id

The password is a secret:

  • Only user and server knows

Question: Anaylyse a password system where no identity is involved

  • Same password

CS2107-3-7.PNG

Identification, Authentication, Authorisation

Password system is classified as weak authentication system

  • Weak authentication: Something subjected to simple replay attack, infomation can be sniffed from communication channel and can be used to impersonate the user
  • Strong authenticaton:
    • Info sniffed during process cant be used to impersonate the user
    • PKI

Understand the difference between sniff and spoof

Possible attacks:

  • Attacking bootstrapping
  • Search password:
    • Guess
    • Dictionary
    • Exhastive
  • Steal:
    • Eavesdropping: Sniff the network , use key logger
    • Phishing
    • Spoofing login screen
    • Password caching
    • Insder attacks

Intercepting password while bootstrapping

Attacker may intercept the password during bootstrapping

Attacker use default password

There are IOT devices which still has the default password

Question:

  • What are the implication of shipping all access with the same default password
    • With same password, attacker can access the device if the common default password
  • What are the implication of shipping each access point with its individual password
    • Expensive: Database and printing
    • Usability will decrease for users when they cannot find the password (Bad product review) Tutorial question

Searching password

The attacker gain social info about user and infer the password

  • Mobile phone number, spouse name

Password guessing:

  • Online guess: Directly interacts with authentication system
  • Offline guessing: An attacker can obtain the password file from authnetication system

Exhastive search and dictionary attacks

  • Attacker tries differnet password during login sessions
  • Attacker can employ exhastive search: tries all combo
  • Restrict the search space to large collection of probable passwords
    • Words from dictionary

CS2107-3-10.PNG

Dictionary attacks

Hybid attacks: It is possible to carry out exhastive search together with dictionary attack/

Stealing password

Shoulder surfing, sniffing

  • Shoulder surfing: Look over the shoulder attack
  • Sniffing: Listening/intercepting the communication channel:
    • Some system and protocol simply send the password over a public network in clear
    • FTP, Telnet, HTTP/

Key-Logger

Captures/records the keystrokes and sends the info back to attacker via covert channel

  • By software: Some computer virus are designed as keylogger
  • By hardware

Login spoofing

Fake loging screens

Prevention:

  • Some system has a secure attention key or secure attention sequence
  • When press, system starts the trusted login processing

Phising

  • Same as login spoofing
  • User is tricked to voluntarily send password to attacker over network
  • Phising attacks ask for password under some false pretense

Spear phishing

  • Phisihing can be targetted to a particular small group of users
  • Such attacks is known as spear phishing which is targetted
  • Social engineering attack

Prevent phishing:

  • User education
  • Phishing repository site

Password caching

  • Shared workstation, password can be stolen
  • Prevention: Clear browser cache

Insider attack

  • Malicious system admin who steals password file
  • System admin account is comporomised, leading to lost of password file

Preventive measures

  • Random
    • Automated password generator
    • High entrophy but difficult to remember
  • Limited login attemps
    • Add delay in login session
    • Add security questions
    • Lock account after few failed attempts
  • Password checker:
    • Check for weak password
  • Password metering: Indicate weak, avg, strong password
  • Password ageing
    • regularly change password
  • Password usage policy
    • Rule set by org to ensure that users use strong password
    • e.g password has to be 10 chara long

Hash password

  • Passwords should be hased and stored in password files
  • During authentication, the password entered by entitty is hased and compared with the value stored in the password file
  • Desired tthat the same password will be hased into two diff values for two diff userid CS2107-3-11.PNG

  • salt

CS2107-3-12.PNG

Security questions

  • Fall back authentication or self services password reset
    • Enhancing usability: A user can still login even if password is lost
    • Reducing cost: Reduces operating cost of help desk
    • (BAD) Weakening security: Attackers have another mean to pbtain access
  • Common secret questioN?
    • Pet name etc

Choosing:

  • Memorable
  • Consistant: should not change over time
  • Nearly universal: Question apply to a wide audience of people
  • Safe: Answers should not be something easily guessed or researched

ATM attacks

To get authenticated, user has to present a card and the PIN. The card contains a magnetic stripe which stores the user account id. Essentially, the magnetic stripe simplidies the input of accound id into ATM system: Instead of keying in

  • The pin: password
  • Data: Encoded into magnetic stripe using well known standards, attacker can cipy the card by reading the info from the card and write on spoofed card

ATM skimmer

  • Reveals the victim account id (Username and password)
  • Consist of
    • Card reader attached on top of existing ATM reader
    • Camera overlooking the keypad or spoofed key pad
    • Some means to record and transmit the infomation back to the attacker

Attacker can 1) Spoof victim ATM card 2) Obtain the pin

Preventive measures

  • Instll anty skimmer device
  • Shield the keypad
  • User awareness

Biometrics

  • Uses unique physical characteristics of a person for authentication
  • Enrollment: reference template of an user biometric data is constructed and stored
  • Verfification: Biometric sample data of person in question is captured and compared with template using matching algo
  • Algo decides if to reject or accepts

Diagram: CS2107-3-13.PNG

Differences
Password Biometrics
Can be changed (revoked) Can’t
need to remember Dont have to
Zero non matched rate probability of error
User can pass the password to another person not possible

Matching algo: Similiarity

  • there might be noises in capturing data
  • leads to error making matchinf decision CS2107-3-15.png

Threshold value selection

  • make decison based on adjustable threshold
  • Lower: More relax in accepting
  • Higher: More stringent CS2107-3-14.PNG

Other erros

  • Equal error rate (EER): When FNMR = FNR -Failure to enroll rate (FER): Some user biometric data cannot be captured during enrollment
  • Failer to capture rate(FTC): Users biometric data cannot be captured during Authentication

CS2107-3-16.PNG

Multifactor authentication (2FA)

  • Require at least 2 different authentication factors
  • Common factors:
    • What you know: Password, pin
    • What you have: Smart card, atm, mobile phone, OTP,
    • Who you are: biometrics
  • Possible factors:
    • Where u are
    • What u do
  • It is called 2 factor authentication if 2 factors are employed
  • MAS expects all banks in sg to provide 2fa for ebanking

What you have

OTP Token

  • A hardware that generates one time password
  • Each token and the server shared some secret
  • Types:
    • Time base: Based on the shared secret and current time interval, a password k is generated
    • Sequence based: An even triggers the change the password (Like user pressing the button)

Password + OTP Token:

Registration:

  • The server issues a OTP token to the uer which contains a secret key that the server knows
  • User set a password

Authentication:

  1. User press the token which computes and display OTP
  2. User sends username and password and OTP to server
  3. Since the server has the secret key, the server can also compute the OTP
  4. Server verifies that both otp and password are correct

Password + Mobile phone (SMS)

Registration:

  • User gives the server his mobile phone number and password

Authentitcation:

  1. User sends password and username to server
  2. Server verifies that the password is correct, server sends OTP to the user via sms
  3. User receives the SMS and enters the OTP
  4. Server verifies that the OTP is correct

Examples:

  • Singpass, internet banking

SMS OTP is not secure

Security threats:

  • Interception of cellular network channel
  • SMS messages are stored as plaintext by the short message service center
  • Malware and trojan on smartphones

Smart card and Finger print (door access system)

Registration:

  • THe server issues a smartcard to the user (which contains a secret key k)
  • The user enrolls his fingerprint

Authentication:

  1. User insert smartcard to the reader.
    • The reader obtains the user identity and verifies whether the smart card is authentic. If so, continue
  2. User presents fingerprint to the reader
    • The reader performs matching to verify that it is authentic, if yes open door

Security requirements:

  • Very often info on user identity, the secret k, and fingerprint template are not stored in the reader
  • The reader has a secure communication channel to a server that stores these info

CS2107-3-20.PNG

In this case we also assume that the reader and server are secure. (Attackers are unable to access them)

Notes:

  1. A smart card has this security feature: Even if an attacker has a physical access to card, it is extremly difficult if not impossible to extract a secret stored in a card
  2. What are the actual two factors
  3. What is the role of the secret
  4. It is possible to eliminate the need of the server e.g By storing the fingerprint in the card and storing a small secret key in the reader

How to achievement

SLides