Reducing Tasks Migration in LRE-TL Real-time Multiprocessor Scheduling Algorithm

Abstract In this paper we present a modification of Local Remaining Execution (LRE-TL) real-time multiprocessor scheduling algorithm to reduce tasks migration significantly. LRE-TL, which is based on the concept of fluid scheduling, makes scheduling decisions using two events. The Bottom (B) event occurs when a task consumes its local utilization thus; it has to be pre-empted. The Critical (C) event occurs when a task consumes its local laxity thus; it has to be scheduled for execution. Event C always results in a task migration. We have modified the initialization procedure of LRE-TL to make sure that tasks which have higher probability of firing a C event will always be scheduled for execution first. This will ensure that the number of C events will always be the minimum; thereby tasks migration will be reduced. To verify our work, an independent-samples t-test is used to compare tasks migration using the original LRE-TL algorithm and the modified one. The results showed that there was a significance reduction in tasks migrations when we apply our proposed solution.