PKM: a pairwise key management scheme for wireless sensor networks

  • Authors:
  • F. An;X. Cheng;J. M. Rivera;J. Li;Z. Cheng

  • Affiliations:
  • Institute of Computing Technology, Chinese Academy of Sciences, Beijing, China;Department of Computer Science, The George Washington University, Washington, DC;Department of Computer Science, The George Washington University, Washington, DC;Department of Systems and Computer Science, Howard University, Washington, DC;National Taxation Bureau of Rizhao City, Shandong Province, China

  • Venue:
  • ICCNMC'05 Proceedings of the Third international conference on Networking and Mobile Computing
  • Year:
  • 2005

Quantified Score

Hi-index 0.00

Visualization

Abstract

Sensor networks are characterized by strict resource limitations and large scalability. Many sensor network applications require secure communication, a crucial component, especially in harsh environments. Symmetric key cryptography is very attractive in sensor networks due to its efficiency, but establishing a shared key for communicating parties is very challenging. The low computational capability and small storage budget within sensors render many popular public-key based key distribution and management mechanisms impractical. In this paper, we propose and analyze a truly in-situ key management scheme for large scale sensor networks, called: Public Key Management (PKM). In this scheme, we deploy service and worker sensors. The service sensors contain a key space, while worker sensors are deployed blind, with no pre-deployment knowledge. Worker sensors obtain security information from service sensors through a secure channel after deployment. After obtaining security information, worker sensors compute shared keys with their neighbors. For security reasons, service sensors erase stored key space information after deployment. During this procedure, PKM shifts a large amount of computational overhead from worker sensors to service sensors, thus conserving worker sensors' resources. PKM's performance, in terms of storage, computational overhead and resiliency, is very good.