Performance evolution in Distributed Real-Time System under dynamic load sharing scheme

Distributed Real-Time System [DRTS] is a prodigious platform for high performance parallel applications. The Performance of the system can be amended by using the policies of dynamic load sharing. Allocation of parallel application´s tasks under load sharing schemes had increased the performance of the system. Task allocation is NP-hard or NP-complete problem. The new heuristic has been deployed in this paper. In the suggested model the limitation of memory and the number of tasks allowed over the processor has been considered. At the time of allocation, the above constraint has been checked and fulfilled. In the proposed model clustering has been done at two levels. For clustering task k-mean clustering has been used and MATLAB 7.11.0 has been used to simulate the proposed model.