Author(s) :

Juliet Rose D. B, T. Jaya

Conference Name :

International Conference on Green Technology and Management for Environmental Sustainability (ICGMES-2024)

Abstract :

Conventional application mapping techniques for 3D NoCs frequently face difficulties in simultaneously optimizing power consumption, communication delay, and area efficiency. This paper presents the HaSOA-ADP-3D NoC methodology as a new approach to address these challenges. This paper presents a novel Harbor Seal Whiskers Optimization Algorithm (HaSOA) coupled with a Communication Profile Registry (CPR) for efficient application mapping in 3D NoC architectures (HaSOA-ADP-3D NoC). The proposed methodology aims to optimize communication area, reduce communication delay, and minimize power consumption (ADP). HaSOA inspired by seal whiskers’ sensory capabilities, effectively explores the search space, and avoids local optima. CPR captures essential communication characteristics of cores, guiding the optimization process. The proposed HaSOA-ADP-3D NoC method achieves a power consumption of 18.415W, a delay of 12.105 microseconds, and an execution time of 21.115 seconds. Experimental results on standard benchmarks like VOPD and MPEG-4, demonstrate significant improvements in power consumption, communication delay, and overall performance compared to existing methods. The proposed HaSOA-ADP-3D NoC method demonstrates its potential as a valuable tool for developing 3D NoC systems that excel in both performance and energy efficiency.

No. of Downloads :

16