A MODIFIED TASK SCHEDULING ALGORITHM OF TASK GRAPH WITHOUT COMMUNICATION TIME

The scheduling of tasks on the processors is considered to be a very critical problem in the multiprocessor system. Here, the problem is divided into subtasks and it is assigned on the processors. The overall objective of task scheduling is to minimize the execution time. Considering that the task scheduling is represented by a directed acyclic graph (DAG). In this paper, we are proposing an algorithm for task scheduling without communication time while assigning of tasks on the processors. It is a modified algorithm of task scheduling based on breath first search (TSB) in our previous paper. The new algorithm gives better result than TSB. A comparative study amongst the developed algorithm, TSB and BNP class of scheduling algorithms. It has been done based on the followings matrices: scheduling length, speedup, efficiency, load balancing and normalized scheduling length. According to the comparative results, it is found that the developed algorithm gives better results.