An Evaluation of Routing Algorithms in Traffic Engineering and Quality of Service Provision of Network on Chips

Efthymios N Lallas

doi:10.4236/eng.2021.131001

Abstract

Nowadays the number of cores that are integrated into NoC (Network on Chip) systems is steadily increasing, and real application traffic, running in such multi-core environments requires more and more bandwidth. In that sense, NoC architectures should be properly designed so as to provide efficient traffic engineering, as well as QoS support. Routing algorithm choice in conjunction with other parameters, such as network size and topology, traffic features (time and spatial distribution), as well as packet injection rate, packet size, and buffering capability, are all equivalently critical for designing a robust NoC architecture, on the grounds of traffic engineering and QoS provision. In this paper, a thorough numerical investigation is achieved by taking into consideration the criticality of selecting the proper routing algorithm, in conjunction with all the other aforementioned parameters. This is done via implementation of four routing evaluation traffic scenarios varying each parameter either individually, or as a set, thus exhausting all possible combinations, and making compact decisions on proper routing algorithm selection in NoC architectures. It has been shown that the simplicity of a deterministic routing algorithm such as XY, seems to be a reasonable choice, not only for random traffic patterns but also for non-uniform distributed traffic patterns, in terms of delay and throughput for 2D mesh NoC systems.

Highlights

This paper proposes an evaluation of routing algorithm impact on traffic engineering and performance specified via QoS metrics, such as average delays and throughputs in 2D NoCs, under different values of critical importance parameters, as spatial and time distribution traffic profiles, or more realistic Hotspot traffic profiles, while varying buffer size, packet injection rate (PIR), packet size and network size, either individually or as a set
It has been shown that the simplicity of a deterministic nature routing algorithm such as XY, seems to be a primary and reasonable choice, for random traffic patterns and for more realistic, non-uniform distributed traffic patterns, in both terms of delay and throughput for 2D mesh NoC systems
As NoC system scalability increases, delay performance, increasing according to the proportional network size increase, would be the same for either deterministic or adaptive algorithms, as the probability of a congestion event would be eliminated for any applied routing scheme

Summary

Introduction

Routing algorithm choice in conjunction with other parameters such as network size and topology, traffic features (time and spatial distribution), as well as packet injection rate (PIR), packet size, and buffering capability, are all equivalently critical for designing a robust NoC architecture, on the grounds of traffic engineering and QoS provision. Routing algorithms in NoC systems are divided into two major classes, adaptive and deterministic [2]. Deterministic routing main advantage is simplicity while for the cognition of adaptive routing, extra computation logic and overheads and more complicated designs are required, in order to decide for an optimized path. There are many routing algorithms representative on either of the two routing classes, such as the XY deterministic algorithm [3], the odd even (OE) adaptive algorithm, the West First, Norst Last και Negative First, as partially adaptive algorithms [4], and hybrid ones, such as DyAD (Dynamic Adaptive Deterministic) algorithm [5], which is switched on either deterministic or adaptive mode, depending on the network congestion status

Results

Discussion

Conclusion