TY - GEN
T1 - Fault-tolerant simulation of message-passing algorithms by mobile agents
AU - Das, Shantanu
AU - Flocchini, Paola
AU - Santoro, Nicola
AU - Yamashita, Masafumi
N1 - Copyright:
Copyright 2020 Elsevier B.V., All rights reserved.
PY - 2007
Y1 - 2007
N2 - The recently established computational equivalence between the traditional message-passing model and the mobile-agents model is based on the existence of a mobile-agents algorithm that simulates the execution of message-passing algorithms. Like most existing protocols for mobile agents, this simulation protocol works correctly only if the agents are fault-free. We consider the problem of performing the simulation of message-passing algorithms when the simulating agents may crash unexpectedly. We show how to simulate any distributed algorithm for the messagepassing model in a mobile-agents system with k agents, tolerating up to f ≤ k - 1 crashes during the simulation. Two fault-tolerant simulation algorithms are presented, one for non-anonymous settings (i.e., where either the networks nodes or the agents or both have distinct identities), and one for anonymous systems (where both the network nodes and the agents are anonymous). In both cases, the simulation overhead is polynomial. Unlike the existing fault-free simulation algorithm, both our protocols are able to detect termination even if the simulated algorithm has no explicit termination detection.
AB - The recently established computational equivalence between the traditional message-passing model and the mobile-agents model is based on the existence of a mobile-agents algorithm that simulates the execution of message-passing algorithms. Like most existing protocols for mobile agents, this simulation protocol works correctly only if the agents are fault-free. We consider the problem of performing the simulation of message-passing algorithms when the simulating agents may crash unexpectedly. We show how to simulate any distributed algorithm for the messagepassing model in a mobile-agents system with k agents, tolerating up to f ≤ k - 1 crashes during the simulation. Two fault-tolerant simulation algorithms are presented, one for non-anonymous settings (i.e., where either the networks nodes or the agents or both have distinct identities), and one for anonymous systems (where both the network nodes and the agents are anonymous). In both cases, the simulation overhead is polynomial. Unlike the existing fault-free simulation algorithm, both our protocols are able to detect termination even if the simulated algorithm has no explicit termination detection.
UR - http://www.scopus.com/inward/record.url?scp=38049186805&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=38049186805&partnerID=8YFLogxK
U2 - 10.1007/978-3-540-72951-8_23
DO - 10.1007/978-3-540-72951-8_23
M3 - Conference contribution
AN - SCOPUS:38049186805
SN - 9783540729181
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 289
EP - 303
BT - Structural Information and Communication Complexity - 14th International Colloquium, SIROCCO 2007, Proceedings
PB - Springer Verlag
T2 - 14th International Colloquium on Structural Information and Communication Complexity, SIROCCO 2007
Y2 - 5 June 2007 through 8 June 2007
ER -