Authors: Jin Yang, Jiannong Cao, Weigang Wu, Cheng-Zhong Xu
Addresses: Internet and Mobile Computing Lab, Department of Computing, Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong Kong. ' Internet and Mobile Computing Lab, Department of Computing, Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong Kong. ' Department of Computer Science, Sun Yat-sen University, Guangzhou 510275, China. ' Department of Electrical and Computer Engg., Wayne State University, Detroit, Michigan 48202, USA
Abstract: Redundancy is necessary for fault tolerance, but the overhead introduced by redundancy may degrade system| performance. In this paper, we propose efficient replication-based algorithms for fault-tolerant mobile agent execution, which enable parallel processing in the agent execution to reduce the overhead caused by redundancy. We also investigate failure detection mechanisms and identify the problems of the heartbeat style failure detection approach and modify it for use in our proposed algorithms. Performance evaluation has been performed to compare the proposed algorithms with the existing algorithm. Both analytic and simulation results show that our new algorithms can significantly improve the system performance.
Keywords: fault tolerance; mobile agents; parallel processing; failure detection; replication-based algorithms.
International Journal of High Performance Computing and Networking, 2009 Vol.6 No.2, pp.106 - 118
Available online: 26 Jul 2009Full-text access for editors Access for subscribers Purchase this article Comment on this article