Modeling and Simulation of Network-on-Chip Routing Algorithm Based on OPNET

Abstract

When a large amount of data floods into the network, the traditional routing algorithm of the Network-onChip will cause network congestion due to the selection of the same path, which will seriously affect the transmission and realtime performance of the data. Based on that, this paper proposes an efficient adaptive routing algorithm to avoid the situation. In this paper, the binary tree topology was improved. Then this paper uses OPNET simulation software to model the improved routing algorithm in three layers, and performs simulation experiments on the model. The evaluations show that the simulation results of the algorithm meet the expectations, and the algorithm can select the shortest path for the packets flooding into the network, which avoiding network congestion. The endto-end delay is 48s less than the traditional routing algorithm, and the real-time performance is higher. This research can effectively avoid network congestion when there is a mass of data in the network.