Measures of Fault Tolerance in Distributed Simulated Annealing
Source: Cornell University
In this paper, the authors examine the different measures of Fault Tolerance in a Distributed Simulated Annealing process. Optimization by Simulated Annealing on a distributed system is prone to various sources of failure. They analyze simulated annealing algorithm, its architecture in distributed platform and potential sources of failures. They examine the behavior of tolerant distributed system for optimization task. They present possible methods to overcome the failures and achieve fault tolerance for the distributed simulated annealing process. They also examine the implementation of Simulated Annealing in Map Reduce system and possible ways to prevent failures in reaching the global optima. This paper will be beneficial to those who are interested in implementing a large scale distributed simulated annealing optimization problem of industrial or academic interest.