Effective Task Scheduling based on Candidate Optimization Algorithm (COA) in Heterogeneous NoC-Based MPSoC

Efficiency of the network communication and performance of the Network of Chips (NoC) with respect to the Quality of Service is greatly affected by the contention of the shared resources. Here, in this paper the design of MPSoC (MultiProcessor System on Chip) is done which is based on Network of Chips involves two steps – scheduling the sub tasks with respect to the Processing elements (PE) and further mapping the PE to the switching nodes which is present in the NoC topology. The evolutionary computation technique is applied once the development of the task model is completed to complete the first step which in turn helps in achieving less transfer cost as well as the running and execution time. The algorithm used in the proposed approach is the Candidate Optimization Algorithm (COA). Based on the NoC topology minimal expected network transmission delay along with less resource and power consumption is used to achieve the second step. The results obtained out of the comparison shows the resources and power consumption which is preferred by the COA algorithm in designing the system.