Comparison of inter-area rekeying algorithms for secure wireless group communications

  • Authors:
  • Chun Zhang;Brian DeCleene;Jim Kurose;Don Towsley

  • Affiliations:
  • University of Massachusetts, Computer Science Building, Amherst, MA;ALPHATECH, Inc., Burlington, MA;University of Massachusetts, Computer Science Building, Amherst, MA;University of Massachusetts, Computer Science Building, Amherst, MA

  • Venue:
  • Performance Evaluation
  • Year:
  • 2002

Quantified Score

Hi-index 0.00

Visualization

Abstract

Many emerging mobile wireless applications depend upon secure group communications, in which data is encrypted and the group's data encryption key is changed whenever a member joins or leaves the group's session. Hierarchical approaches have recently been proposed to manage the distribution of the data encryption key in a scalable manner for fixed (non-mobile) networks. In this paper, we characterize the impact of mobility on secure rekeying of group communication in a hierarchical key-distribution framework. We propose several rekeying algorithms that preserve confidentiality as members move within the hierarchy. The algorithms differ in the locality of communication, the amount of messages needed to rekey the data key/key-encryption key, the key-encryption key rekey rate, and the number of key-encryption keys held by group members. We develop Markov models to quantify the performance of the proposed algorithms. Our results shows that the FEDRP and SR inter-area rekeying algorithms are superior under different circumstances. In the situation of lower arrival rate and higher mobility, SR has the lowest intra-AS message rate, rekey rate and a low inter-AS message rate. On the other hand, with higher arrival rate and lower mobility, FEDRP has a low rekey rate, inter-AS message rate and the lowest intra-AS message rate. This is achieved by allowing members to hold a small number of keys. In a wireless environment, where bandwidth is often a limiting resource, minimization of communication overhead is of critical importance. This goal could be achieved by using a dynamic strategy to combine the benefits of FEDRP and SR algorithms.