A Loss-aware Continuous Hopfield Neural Network (CHNN)-based Mapping Algorithm in Optical Network-on-Chip (ONoC)

Optical Network-on-Chip (ONoC) has become promising due to its advantages in transmission bandwidth and delay, compared to the traditional electrical interconnects on chip. However, different mapping algorithms in ONoC have different impacts on the reliability and accuracy of the results. This work proposes a Continuous Hopfield Neural Network (CHNN)-based mapping algorithm in ONoC to reduce the loss of communications, thus improving power efficiency and reliability. The evaluation results, under a given core graph and network topology, show that the proposed mapping algorithm is able to meet the loss requirement and also demonstrates low computational complexity, fast running speed and high accuracy, compared to other existing mapping algorithms.