Open Access

ARTICLE

Enhanced Mechanism for Link Failure Rerouting in Software-Defined Exchange Point Networks

Abdijalil Abdullahi^1,2, Selvakumar Manickam^2,*

1 Department of Information Technology, SIMAD University, Mogadishu, 630, Somalia
2 National Advanced IPv6 Centre, Universiti Sains Malaysia, Pulua Pinang, 11800, Malaysia

* Corresponding Author: Selvakumar Manickam. Email:

Computers, Materials & Continua 2024, 80(3), 4361-4385. https://doi.org/10.32604/cmc.2024.054215

Received 22 May 2024; Accepted 09 August 2024; Issue published 12 September 2024

Abstract

Internet Exchange Point (IXP) is a system that increases network bandwidth performance. Internet exchange points facilitate interconnection among network providers, including Internet Service Providers (ISPs) and Content Delivery Providers (CDNs). To improve service management, Internet exchange point providers have adopted the Software Defined Network (SDN) paradigm. This implementation is known as a Software-Defined Exchange Point (SDX). It improves network providers’ operations and management. However, performance issues still exist, particularly with multi-hop topologies. These issues include switch memory costs, packet processing latency, and link failure recovery delays. The paper proposes Enhanced Link Failure Rerouting (ELFR), an improved mechanism for rerouting link failures in software-defined exchange point networks. The proposed mechanism aims to minimize packet processing time for fast link failure recovery and enhance path calculation efficiency while reducing switch storage overhead by exploiting the Programming Protocol-independent Packet Processors (P4) features. The paper presents the proposed mechanisms’ efficiency by utilizing advanced algorithms and demonstrating improved performance in packet processing speed, path calculation effectiveness, and switch storage management compared to current mechanisms. The proposed mechanism shows significant improvements, leading to a 37.5% decrease in Recovery Time (RT) and a 33.33% decrease in both Calculation Time (CT) and Computational Overhead (CO) when compared to current mechanisms. The study highlights the effectiveness and resource efficiency of the proposed mechanism in effectively resolving crucial issues in multi-hop software-defined exchange point networks.

---

Keywords

---