Saturday, April 7, 2012

DSGC – Decentralized Secure group communication

1. Project Title

DSGC – Decentralized Secure group communication

2. Purpose

In modern globally connected world, group communication is becoming a date-to-date requirement. Secrecy of the messages, support of backward and forward secrecy & dynamic nature of the groups are critical and important requirements of such a communication. So the purpose is to analyze the better ways of achieving these.

3. Background

With the enhancements of technologies, current personal computers are so powerful, and some hand-held devices are also capable of doing computations up to a good extend. However group communication is still relying on third party owned centrally managed username/password based authentication servers and message relaying servers. Most of the these services are free of charge where users neglect the security concerns. Even when content of the communication is confidential, there is no way to impose security restrictions. Service provider might be having the ability to read the whole content because user authentication is also done by the same service. Another problem is the inability to support forward and backward secrecy with existing services.

4. Scope

As the time duration available for the project is limited, the main focus is put on the analysis of the different subsections of the problem. Implementation of such a software is not considered, but a POC will be. Following are included in the scope.
  • Group Key Management
    • Member authentication will be handled using PK cryptography
    • Group key generation mechanisms analyzed will not be exposing the username & password through the network
    • Key generator responsibility assignment for members will be analyzed
  • Group Key Generation occasions
    • Member join, leave will be considered for key regeneration operations.
  • Group key distribution mechanisms will be analyzed
  • Message distribution
    • Messages will be categorized as system messages & user messages and distribution will be considered based on the message category.

5. Theoretical framework

Key generation and management is theoretically considered and concepts like Public-key cryptography, Diffie–Hellman key exchange and Elliptic curve cryptography will be considered in this project. Leader election algorithms will also be another area of interest.

6. Method

  • Existing group communication software and services will be analyzed
  • Effective and efficient key generation mechanisms will be analyzed
  • Key exchange approaches and the support for dynamic nature of the groups will be analyzed with the help of existing researches on the areas of dynamic group management
  • Message encryption will be analyzed with the consideration on supporting forward and backward secrecy support
  • Based on above analysis, a POC will be generated using freely available tools & libraries
  • An implementation guide for a decentralized secure group communication will be created

7. Tiimeframe

Analyze Research Areas March
Interim Report May
Solution Design Phase June
Solution POC August
Final Report September

8. Limitations

A fully fledged software is not an expectation even though a software will be implemented only to combine the sections to prove the concepts.

9. References

  • W. Diffie and M. Hellman. New directions in cryptography. IEEE Transactions on Information Theory, IT-22(6):644– 654, 1976.
  • Koblitz, N. (1987). Elliptic Curve Cryptosystems. Mathematics of Computation, 48(177), 203. doi:10.2307/2007884
  • C.Wong, M. Gouda, and S. Lam. Secure group communications using key graphs. Technical Report TR97-23, University of Texas at Austin, Department of Computer Sciences, August 1997.
  • Adusumilli, P., Zou, X., & Ramamurthy, B. (2005). DGKD : Distributed Group Key Distribution with Authentication Capability. Security, (June), 15-17.
  • Lee, P. P. C., Lui, J. C. S., Yau, D. K. Y., & Lafayette, W. (n.d.). Distributed Collaborative Key Agreement Protocols for Dynamic Peer Groups. Group.
  • Perrig, A., & Song, D. (n.d.). ELK , a New Protocol for Efficient Large-Group Key Distribution £. Group.

No comments:

Post a Comment