SuperStabilizing protocols for dynamic distributed systems
Proceedings of the fourteenth annual ACM symposium on Principles of distributed computing
An exercise in fault-containment: self-stabilizing leader election
Information Processing Letters
Fault-containing self-stabilizing algorithms
PODC '96 Proceedings of the fifteenth annual ACM symposium on Principles of distributed computing
Optimal reactive k-stabilization: the case of mutual exclusion
Proceedings of the eighteenth annual ACM symposium on Principles of distributed computing
Fault-local distributed mending
Journal of Algorithms
Stabilizing time-adaptive protocols
Theoretical Computer Science
Self-stabilizing systems in spite of distributed control
Communications of the ACM
ISTCS '97 Proceedings of the Fifth Israel Symposium on the Theory of Computing Systems (ISTCS '97)
Superstabilizing mutual exclusion
Distributed Computing
Self-stabilization of dynamic systems assuming only read/write atomicity
Distributed Computing - Special issue: Self-stabilization
How to Improve Safety under Convergence Using Stable Storage
IEEE Transactions on Parallel and Distributed Systems
Analysis of an Intentional Fault Which Is Undetectable by Local Checks under an Unfair Scheduler
SSS '09 Proceedings of the 11th International Symposium on Stabilization, Safety, and Security of Distributed Systems
Algorithms and theory of computation handbook
Loop-free super-stabilizing spanning tree construction
SSS'10 Proceedings of the 12th international conference on Stabilization, safety, and security of distributed systems
Space-efficient fault-containment in dynamic networks
SSS'11 Proceedings of the 13th international conference on Stabilization, safety, and security of distributed systems
| Hi-index | 0.00 |
A self-stabilizing protocol is a protocol that achieves its intended behavior regardless of the initial configuration. Thus, a self-stabilizing protocol is resilient to any number and any type of transient faults. A self-stabilizing protocol is called a superstabilizing protocol if it can recover from any "almost legitimate" configuration while satisfying a certain safety property during the recovery. Superstabilizing protocols are attractive since the number of processes that experience transient faults at the same time is small in most cases, and the resultant configuration is almost legitimate. Herman [Distrib. Comput. 13(1) (2002), 1-17] investigates superstabilizing mutual exclusion protocols in unidirectional rings. Herman considers, as the almost legitimate configuration, any configuration that results from a transient fault of a single process, and proposes the following safety property that should be satisfied during the recovery: no two processes simultaneously have privileges with the single exception of a temporary spurious privilege that the faulty process has. Herman adopts the process-register model, and shows an impossibility result: there exists no 1-latent protocol using fewer than 2n registers, where n is the number of processes in the ring. Herman also presents an n-latent protocol using n registers and a 1-latent protocol using 2n registers.In this paper, we further consider the latency of superstabilizing mutual exclusion protocols on unidirectional rings. On the shared variable model, we show that there exists no ⌈n/2⌉-latent protocol, and present a (⌈n/2⌉+ 1)-latent protocol. The proposed protocol is latency-optimal for the case that n is even.