Thread scheduling using ant colony optimization: An intelligentscheduling approach towards minimal information leakage

Abstract

In multithreaded programs, scheduler controls the execution of threads. As a result, the scheduler may arrange the executionsequence of threads in such a way that multithreaded programs may violate the non-interference confidentiality policy. Due to aviolation of non-interference, multithreaded programs may leak security sensitive information. In the proposed work, Ant ColonyOptimization (ACO) based intelligent scheduling policy has been proposed in the form of algorithms to schedule threads inmultithreading environment in such a way that the execution sequence leads to minimal information leakage. In the present work,the proposed scheduler also deals with conflicting scheduling parameters and provides the algorithmic solution which can handle allconflicting entities of scheduling like throughput, delay, security-privacy and fairness. In this work, dynamic creation and deletionof threads are also handled during the scheduling. Although the focus of this work is on the scheduling of threads, the proposedpolicy can be used as a general purpose scheduling policy in many computing fields.