Dynamic clustering-based routing scheme for 2D-Mesh networks-on-chip

The efficiency of Networks on Chip (NoC) is affected by related routing algorithms. Moreover, clustering can be utilized to improve the performance of the related routing algorithm. Since traffic type of a system on chip may change with time, dynamic clustering is required to deal with this issue. In this paper, a dynamic clustering-based routing algorithm is proposed utilizing Evolved Fuzzy-Neural Networks (EFuNN) architecture. In the proposed approach, the nodes that have the highest communication among them are grouped into the same cluster. Cluster formation dynamically changes depending on the type of traffic applied to the network. The related reliable routing algorithm splits into two categories; while intra-cluster routing mechanism varies according to the applied traffic type, inter-cluster routing mechanism operates based on EFuNN. Experimental results show that the on-chip network utilizing the proposed method outperforms state-of-the-art research works in terms of reliability, latency, and throughput regarding different traffic patterns for various 2D-mesh topologies.