sesame a secure and convenient mobile solution for
play

Sesame: A Secure and Convenient Mobile Solution for Passwords Dr. - PowerPoint PPT Presentation

Sesame: A Secure and Convenient Mobile Solution for Passwords Dr. Mehrdad Aliasgari , Nick Sabol, and Ashutosh Sharma California State University, Long Beach MobiSecServ Feb. 2015 Passwords CSU Long Beach 2 Most Popular Passwords of 2014*


  1. Sesame: A Secure and Convenient Mobile Solution for Passwords Dr. Mehrdad Aliasgari , Nick Sabol, and Ashutosh Sharma California State University, Long Beach MobiSecServ Feb. 2015

  2. Passwords CSU Long Beach 2

  3. Most Popular Passwords of 2014* • 123456 • password • 12345 • 12345678 • qwerty • 123456789 * Compiled by SplashData • 1234 • baseball • dragon • …… CSU Long Beach 3

  4. Password Managers Cont. • Three types – Desktop: No mobility – Mobile : Trust third party – Device based: Have to carry them • Have to set a master password – All passwords are encrypted using one single key phrase. • If master password is compromised…. CSU Long Beach 4

  5. Our Work • Biometric and Phone-based, online password manager • Data distributed in parts. All parts need to come together to read data • Our choice of biometric: Voice (Speech and Speaker recognition) CSU Long Beach 5

  6. Sesame • Idea: – Encrypt each password with a fresh key – Backup the encrypted passwords in the cloud – Encrypt the fresh keys and store them on Sesame server – If the user passes authentication then release the encrypted key • Neither the cloud nor Sesame knows anything about your passwords CSU Long Beach 6

  7. Sesame (Cont.) • User Authentication: – Voice (Speaker recognition) – Speech recognition to extract the requested entry • Master passwords are used as an alternative but users don’t have to type them every time. • If master password is compromised user is still safe (better change it soon) CSU Long Beach 7

  8. System Overview uid • P i : Password Enc(K 2 , K e ) Enc(K 2 , RSA-Public) • K i : Encrypting key Enc(K 3 , RSA-Private) Cloud storage Enc(K i , P i ) Sesame Enc(K e , K i ) and Phone server RSA(K i ) CSU Long Beach 8

  9. Adding a New Password Entry • S i : Service name • U i : Username Cloud • P i : Password storage • K i : Encrypting key (fresh) • m i : S i || U i ||P i Enc(K i , m i ) Sesame S i , Enc(K e , K i ) Phone server and RSA(K i ) CSU Long Beach 9

  10. Retrieving a Password Entry • S i : Service name • U i : Username Cloud • P i : Password storage • K i : Encrypting key • m i : S i || U i ||P i Enc(K i , m i ) Voice or S i Sesame Phone server S i and Enc(K e , K i ) or RSA(K i ) CSU Long Beach 10

  11. Cryptographic Tools • Master password is used to generate K 1 , K 2 and K 3 using KDF (Key Derivation Function) – 4096 iterations – uid is used as a salt • Symmetric Encryption: AES 256 bits with CBC mode • Asymmetric: RSA-OAEP 2048 bits CSU Long Beach 11

  12. Symmetric vs Asymmetric • Why we have both Enc() and RSA()? • It depends on what method of authentication the users chooses • When speaker recognition is used – Enc(K e , K i ) • When master password is used – RSA(K i ) CSU Long Beach 12

  13. Encryption and Distribution • All passwords are encrypted with a new key • Encrypted passwords are backed up • The keys encrypted and stored in Sesame server • To recover a password you need: – The backed up data in the cloud – The encrypted keys – The key to decrypt keys CSU Long Beach 13

  14. Security Analysis • No one party has all necessary pieces • Collusion attack: – Sesame serve and the cloud collide • Best they can do is to brute-force masterpassword • Exponential • No offline dictionary attack due to use of salt (uid) CSU Long Beach 14

  15. Other Capabilities • You can use the application on multiple devices – at the installation on second device: • Connect with your cloud • Enter the master password • Respond to the prompted speaker recognition challenge • Users can change their master password CSU Long Beach 15

  16. Android App CSU Long Beach 16

  17. CSU Long Beach 17

  18. CSU Long Beach 18

  19. CSU Long Beach 19

  20. Conclusion • Secure method of distributing sensitive data • Can be applied to secure cloud storage of any type of data • Other biometric modalities can be used – Handwriting CSU Long Beach 20

Download Presentation
Download Policy: The content available on the website is offered to you 'AS IS' for your personal information and use only. It cannot be commercialized, licensed, or distributed on other websites without prior consent from the author. To download a presentation, simply click this link. If you encounter any difficulties during the download process, it's possible that the publisher has removed the file from their server.

Recommend


More recommend