Optimizing Network Performance: A Comparative Analysis of EIGRP, OSPF, and BGP in IPv6-Based Load-Sharing and Link-Failover Systems

Abstract

The purpose of this study is to evaluate and compare how well different routing protocols perform in terms of load sharing, link failover, and overall network performance. Wireshark was used for packet-level analysis, VMWare was used for virtualization, GNS3 was used for network simulation, and Iperf3 was used to measure network performance parameters. Convergence time, packet losses, network jitter, and network delay are the parameters that were selected for assessment. To examine the behaviors of Open Shortest Path First (OSPF) and Enhanced Interior Gateway Routing Protocol (EIGRP) routing protocols in a variety of network settings, a simulated network environment incorporating both protocols along with Border Gateway Protocol (BGP) is created for the research. The setup for the experiment entails simulating different network conditions, such as fluctuating traffic loads and connection failures, to track how the protocols function in dynamic situations. The efficiency metrics for OSPF and EIGRP with BGP are measured and evaluated using the data generated by Wireshark and Iperf3. The results of this research show that EIGRP has a better failover convergence time and packet loss percentage as compared to the OSPF.

1. Introduction

Corporate networks require a stable connection with internet services and remote branches. Organizations strive to minimize network downtime through link-failover mechanisms. However, a significant limitation of these mechanisms is the idle state of backup links when the primary link remains operational for extended periods, leading to unnecessary costs for unused resources. Additionally, organizations managing both branch-to-branch communication and internet traffic face challenges in handling these traffic types simultaneously. For security reasons, some organizations mandate that data traffic must traverse an overlay network, while internet traffic uses an underlay network. Moreover, organizations seek the flexibility to choose optimal paths for network traffic to enhance performance.

Network engineers often struggle to select the most appropriate routing protocol without a thorough understanding of the strengths and limitations of available options. This knowledge is essential for improving routing infrastructure and ensuring effective traffic distribution. Therefore, a comprehensive solution is needed that efficiently utilizes backup links, manages dual traffic types, secures and optimizes routing paths, and equips network engineers with the necessary insights to choose the best routing protocols for superior network performance.

This study aims to provide valuable insights for researchers and network engineers on the performance of Enhanced Interior Gateway Routing Protocol (EIGRP) and Open Shortest Path First (OSPF) when integrated with Border Gateway Protocol (BGP), particularly in the contexts of load sharing and link-failover strategies. To conduct a thorough evaluation, this work employs simulation tools like Graphical Network Simulator-3 (GNS3) for network topology simulation, VMware for virtualization, and Wireshark and Iperf3 for detailed network analysis. These tools allow us to create realistic network scenarios and assess critical performance metrics such as packet loss, convergence time, network jitter, and delay.

As networks continue to evolve to support new technologies, a robust routing infrastructure becomes increasingly essential. This study contributes to the understanding of how routing protocols perform in modern network environments. By combining BGP, which plays a key role in connecting autonomous systems, with widely used interior gateway protocols like OSPF and EIGRP, our research offers a comprehensive view of routing protocol efficiency in practical settings. This work aims to identify the strengths and weaknesses of OSPF and EIGRP in scenarios involving load sharing and link failures, thereby evaluating their suitability for various networking contexts. Including Internet Protocol version 6 (IPv6) in the analysis ensures that the findings align with the industry’s transition to the latest version of the Internet Protocol.

Advanced simulation tools, such as GNS3, facilitate this investigation by enabling the construction of realistic network environments for in-depth analysis. By simulating real-world conditions, we can observe how different routing protocols perform in dynamic and changing network environments. The ultimate goal is to extract actionable insights from the collected data, providing researchers and network engineers with the knowledge needed to make informed decisions about selecting and implementing routing protocols in contemporary networking scenarios. This research aims to enhance network performance, flexibility, and scalability by understanding the intricate dynamics of OSPF and EIGRP in conjunction with BGP, particularly in the context of IPv6. The expected outcomes will serve as a valuable resource for making informed choices, ensuring that routing protocols effectively meet the evolving demands of modern network architectures.

This research makes the following key contributions: (1) Introduces a mechanism that utilizes both primary and secondary links simultaneously, addressing the inefficiency of idle backup links and ensuring cost-effective resource utilization; (2) Focuses on managing two distinct types of traffic: data traffic for branch-to-branch communication and internet traffic for external services; (3) Employs an overlay network for secure data communication and an underlay network for internet traffic, enhancing security and optimizing network resource use; (4) Develops a route-mapping mechanism to manage both data and internet traffic effectively. This mechanism supports load sharing and link failover on the preferred link according to specific needs, ensuring optimal network performance; (5) Compares EIGRP and OSPF based on key network parameters such as delay, jitter, packet loss, and failover convergence time in an IPv6 environment. This comparison provides network engineers and administrators with the information needed to select the most appropriate routing protocol to enhance their routing infrastructure for effective traffic distribution.

The remainder of this paper is structured as follows: Section 2 provides an overview of routing protocols; Section 3 presents a literature review; Section 4 discusses load sharing and link-failover systems; Section 5 highlights the importance of IPv6 in networking; Section 6 details the methodology and performance metrics; Section 7 presents the results and discussion; and Section 8 concludes the paper with a summary and suggestions for future work.

2. Routing Protocols Overview

The performance and stability of network communication heavily rely on the choice and implementation of routing protocols. In business networks, two widely used interior gateway protocols are OSPF (Open Shortest Path First) and EIGRP (Enhanced Interior Gateway Routing Protocol). OSPF uses a link-state routing algorithm, which enables it to make efficient routing decisions based on a comprehensive understanding of the entire network topology. EIGRP, developed by Cisco, is designed to achieve fast convergence and low resource usage by incorporating both distance vector and link-state characteristics.

For scenarios that involve internet access and the need to connect multiple autonomous systems, the exterior gateway protocol BGP (Border Gateway Protocol) plays a crucial role. BGP is essential for routing data between different networks, ensuring stable and reliable communication across diverse and geographically distributed systems.

The study aims to assess the performance of EIGRP and OSPF, particularly in an IPv6 environment, focusing on their effectiveness in handling load sharing and link-failover scenarios. This evaluation could provide insights into how these protocols contribute to network resilience and traffic distribution, offering recommendations for enhancing failover capabilities and load-sharing techniques in IPv6-based systems.

2.1. EIGRP Features and Characteristics

EIGRP (Enhanced Interior Gateway Routing Protocol) significantly enhances the efficiency and flexibility of network operations. Developed by Cisco, EIGRP is a highly effective and proprietary interior gateway protocol that combines elements of both link-state and distance vector protocols. One of its key features is the Diffusing Update Algorithm (DUAL), which enables rapid convergence and efficient use of network resources. EIGRP excels in load-sharing scenarios due to its ability to dynamically distribute traffic across multiple paths, optimizing network efficiency and ensuring a more balanced workload distribution. Additionally, EIGRP’s adaptability in quickly responding to changes in network topology makes it well-suited for environments where link failure is a critical concern. In the context of load-sharing and link failover, EIGRP plays a pivotal role by continuously monitoring the network for changes and selecting the best routes based on real-time metrics. When evaluated alongside OSPF and BGP in an IPv6 environment, EIGRP’s performance analysis provides a comprehensive understanding of its contribution to enhancing network resilience and efficiency.

Convergence Process of EIGRP

The process begins with the use of Hello packets, which are sent to establish neighbor relationships and share crucial information. Once routers recognize each other, they exchange Update packets containing information about reachable destinations and their associated metrics. These data are essential for constructing and updating topology tables. EIGRP routers then utilize the DUAL algorithm to calculate feasible distances and determine the optimal paths. The continuous exchange of routing information through Update packets ensures that network updates occur in real time, allowing routers to quickly adapt to changes in topology. This combined process enhances EIGRP’s responsiveness and reliability, especially in dynamic network environments where conditions are constantly changing. The convergence process of EIGRP is illustrated in Figure 1.

Figure 1 shows the step-by-step process of how EIGRP routers establish neighbor relationships, exchange routing information, calculate optimal paths using DUAL, and update the network topology to ensure efficient routing.

2.2. OSPF Features and Characteristics

OSPF (Open Shortest Path First) is a link-state routing protocol that enables network routers to maintain a synchronized view of the network topology. By having a comprehensive understanding of the network architecture, OSPF can determine the shortest path to a destination, facilitating efficient and optimal routing decisions. In the context of this study, OSPF is likely employed to facilitate communication between routers within the same autonomous system, particularly in an IPv6 environment. OSPF’s role includes load sharing, where it allows traffic to be distributed across multiple paths, thereby enhancing network performance and resource efficiency. Additionally, OSPF contributes to system resilience by supporting efficient link-failover mechanisms, ensuring a seamless transition in the event of network outages. The performance analysis of OSPF, in conjunction with EIGRP and BGP, provides valuable insights into load sharing and link-failover strategies within IPv6-based systems.

Convergence Process of OSPF

The convergence process in OSPF involves several critical phases and the exchange of specific OSPF protocol packets. The process begins with the exchange of Hello packets between neighboring routers, initiating neighbor detection and verification. These Hello packets are essential for establishing and maintaining adjacencies between routers. For proper adjacency formation, routers must match OSPF parameters such as the router ID, area ID, and authentication settings.

Once adjacencies are established, the routers exchange Database Description (DBD) packets, which provide an overview of the router’s link-state database. If there are missing entries, Link State Request (LSR) packets are sent to request specific Link State Advertisements (LSAs) that are not present in the recipient’s database. Upon receiving the necessary LSAs, the responding router sends out Link State Update (LSU) packets. Finally, to confirm the receipt of LSAs, Link State Acknowledgment (LSAck) packets are exchanged.

This series of OSPF protocol packets ensures that routers share and synchronize their understanding of the network’s link-state information, enabling precise and efficient routing decisions. Figure 2 would typically illustrate this convergence process, depicting the step-by-step exchange of packets that leads to the synchronization of network topology information across routers.

Figure 2 shows the stages of OSPF convergence, from the initial Hello packet exchange to the final acknowledgment of LSAs, highlighting how OSPF routers achieve a consistent and accurate view of the network.

2.3. BGP Features and Characteristics

BGP (Border Gateway Protocol) is essential for enabling communication between multiple autonomous systems (ASs) on the internet. As an exterior gateway protocol, BGP is responsible for exchanging routing information and determining the best routes for data and internet traffic across diverse network domains. It must handle complex routing decisions based on policies, such as path attributes and network policies, to optimize traffic flow and ensure efficient load sharing. In the context of the load sharing and link-failover system being studied, BGP plays a critical role by influencing routing decisions that guide traffic between autonomous systems, thereby affecting the overall performance of the network. BGP is divided into two primary types:

• Internal Border Gateway Protocol (iBGP): iBGP is used for communication between routers within the same autonomous system (AS). It facilitates the distribution of routing information within the boundaries of a single AS, ensuring consistent and effective routing decisions across the internal network. iBGP helps maintain a coherent view of the network’s topology, enabling routers within the AS to make informed decisions about traffic routing.
• External Border Gateway Protocol (eBGP): eBGP is utilized for communication between routers in different autonomous systems. It ensures that networks managed by separate administrative entities can exchange routing information, thereby enabling inter-AS communication. eBGP plays a pivotal role in the broader internet infrastructure by facilitating the exchange of routing data across the global network, allowing data to traverse multiple autonomous systems efficiently.

Both iBGP and eBGP contribute to BGP’s ability to manage and optimize the flow of traffic across the internet, making it a crucial protocol for global network connectivity and stability.

3. Literature Review

The research presented in [1] introduces an innovative technique that integrates a backup mechanism directly into the real-time network, enhancing downtime management and maintaining optimal quality at 100%. The simulations demonstrate that the backup process operates smoothly, ensuring network quality remains intact. This study highlights the benefits of a failover system using BGP routing on a single link connection, showing superior performance.

The study in [2] evaluates the effectiveness of merging internal and external routing protocols using GNS3, focusing on parameters such as throughput, jitter, packet loss, and network convergence. To achieve high Quality of Service (QoS), it is essential to minimize levels of throughput, jitter, and packet loss. The findings reveal that the OSPF routing protocol offers the fastest network convergence, while the combination of OSPF and BGP delivers the highest throughput, lowest packet loss, and lowest jitter.

In [3], the research aims to identify the best interior gateway routing protocol for various traffic conditions within a corporate network. The study seeks to determine the most effective protocol for managing and routing network traffic, ensuring efficient communication and performance. The ultimate goal is to provide enterprises with valuable insights into selecting the internal gateway protocol that best meets their specific network requirements.

The research in [4] focuses on comparing various networking characteristics using Cisco’s Packet Tracer simulation environment, specifically examining packet drops, routing convergence speed, and end-to-end latency. The study likely assesses the performance of different networking setups or protocols based on these criteria to gain insights into their efficiency and effectiveness in a simulated environment.

The study in [5] analyzes the performance of IPv6, concentrating on the header format, security concerns, and routing protocols, utilizing Packet Tracer for simulations. The emphasis is on evaluating the efficiency and functionality of these elements within the context of IPv6.

The research in [6] investigates the performance of well-known routers, focusing on three IPv6 protocols: EIGRPv6, OSPFv3, and Routing Information Protocol Next Generation (RIPng). Data transmission rates and convergence times are measured using a Cisco simulation platform. Real-time comparisons are made using the ping command to check network connections, with findings collected through Cisco Packet Tracer, offering insights into the efficiency of these protocols in an IPv6 environment.

In [7], the work examines the challenges faced by Internet Service Providers (ISPs) during the transition to IPv6. Issues such as packet traversal difficulties, routing scaling limitations, performance reliability concerns, and the need for enhanced security measures are explored. The study provides a comprehensive overview of the complexities involved in migrating to an IPv6 network.

The fundamental purpose of the research in [8] is to evaluate the performance of specific protocols within simulated network models. The assessment is based on factors such as throughput, jitter, convergence time, end-to-end latency, and packet depletion. The results indicate that the EIGRP routing protocol outperforms OSPF in terms of overall performance, suggesting it is more suitable for real-world applications.

The study in [9] explores multiple Internet Protocol Security (IPsec) transformations in Network Simulator Version 2 (NS2) simulations, using different authentication and encryption techniques. The findings reveal that IPsec overhead is minimal for smaller packet sizes but significantly increases for larger packets, especially when fragmentation is necessary. This information can help network managers make informed decisions about the trade-off between processing costs and the need for larger address spaces when transmitting large IP packets.

In [10], the research compares the performance of three dynamic routing methods and investigates their redistribution mechanisms within a simulated network topology. The performance evaluation focuses on latency, throughput, and packet loss, with the network model comprising three PCs, nine Cisco routers, and four switches.

The study in [11] assesses the performance of system failover connections within a network architecture that utilizes a combination of internal and external routing protocols and various Autonomous System (AS) numbers. Using the GNS3 simulator, the study measures key performance metrics such as throughput, jitter, packet loss, and latency, providing insights into the reliability and effectiveness of failover mechanisms in complex network scenarios.

The research in [12] explores the Intermediate System to Intermediate System (IS–IS) and Open Shortest Path First version 3 (OSPFv3) protocols, including their metric systems, with a focus on simulation outcomes. The study aims to draw conclusions about the efficiency of these protocols when implemented in an IPv6 corporate network.

In [13], the research delves into RIPng and OSPFv3, comparing these protocols within an IPv6 framework. The study discusses simulation findings and seeks to determine which protocol is most effective under various network conditions. The study in [14] evaluates several transition techniques in networking using Cisco Packet Tracer, highlighting that the dual-stack approach performs better than tunneling. The research provides insights into the effectiveness of these processes in managing network transitions and emphasizes the benefits of adopting a dual-stack approach.

The research in [15] compares the performance of an Interior Gateway Routing Protocol (IGRP) with an Exterior Gateway Protocol (EGP), aiming to identify the best protocol combination for complex environments. The study also simulates the Hot Standby Routing Protocol (HSRP) and the Gateway Load Balancing Protocol (GLBP) to evaluate load balancing and redundancy specific to BGP.

The study in [16] investigates the transition and migration processes of IPv6, followed by a performance comparison of IPv4 and IPv6 for both data and voice traffic using Riverbed. The analysis shows that dual-stack is the most effective strategy for transitioning from IPv4 to IPv6, with tunneling as a viable alternative.

In [17], the research builds a network to evaluate the performance of OSPF and OSPFv3 routing protocols in both IPv4 and IPv6 environments. The study compares response times for various applications, such as Database Query, Email, and FTP Download, under OSPF and OSPFv3, measuring traffic dropouts, hop counts, background traffic latency, and Ethernet delay.

The study in [18] reviews multiple research papers on IPv6 deployment and the performance analysis of the OSPFv3 routing protocol. The review provides significant insights for researchers, laying the foundation for future investigation and advancement in the field.

The research in [19] analyzes how different Generic Routing Encapsulation (GRE) tunnel parameters affect the efficiency of real-time application streaming in IP networks. The study examines tunnel depth and the routing methods used within GRE tunnels, offering insights into the limitations and optimal routing schemes for real-time traffic.

The purpose of the research in [20] is to assess the effectiveness of routing protocols within an IPv6 network, particularly RIPng, OSPFv3, and EIGRP. Performance measures such as throughput, jitter, and packet loss are evaluated using GNS3, with the goal of determining the suitability of each routing protocol in an IPv6 context.

This research distinguishes itself by utilizing advanced simulation and virtualization tools like GNS3, VMware, Wireshark, and Iperf3 to evaluate critical metrics such as convergence latency, jitter, and packet loss. Unlike previous studies that often relied on physical network setups with inherent limitations, this study leverages sophisticated tools to create and analyze complex, realistic network environments. This approach allows for a more comprehensive assessment of the effectiveness of IPv6 routing protocols under various conditions, including traffic load sharing and connection failures. By exploring a broader range of scenarios, this research provides deeper insights into the strengths and weaknesses of the selected routing protocols in a complex network environment.

4. Load Sharing and Link-Failover Systems

The study focuses on the load-sharing capabilities and link-failover performance of EIGRP and OSPF, two dynamic routing protocols essential for optimizing network efficiency and reliability.

4.1. Load Sharing Capability of EIGRP and OSPF

Load sharing refers to the distribution of network traffic across multiple pathways to enhance resource utilization and overall network performance. EIGRP and OSPF, both dynamic routing protocols, have distinct mechanisms for managing load sharing.

EIGRP is known for its advanced load-sharing capabilities, which involve a sophisticated system that considers various parameters such as bandwidth, latency, reliability, and line load. EIGRP dynamically calculates composite metrics for multiple routes, enabling it to make informed decisions about how to distribute traffic effectively. This research will delve into the complexities of EIGRP’s load-sharing algorithms, evaluating their adaptability to changing network conditions and the efficiency of their traffic distribution strategies.

OSPF, on the other hand, utilizes link costs to determine the shortest path to a destination. OSPF’s load-sharing capability is based on Equal Cost Multi-Path (ECMP) routing, which allows it to split traffic across multiple paths with the same cost. The study will explore OSPF’s load-sharing methods, examining how it handles traffic distribution when multiple routes have equal costs and the impact this has on overall network performance.

Through a series of experiments simulating various network architectures and traffic flow patterns, this research will compare the load-sharing efficiencies of EIGRP and OSPF. The results will provide valuable insights into the comparative advantages of each protocol, aiding network administrators and engineers in optimizing their routing infrastructure for efficient traffic distribution.

4.2. Link-Failover Performance of EIGRP and OSPF

Link-failover performance is crucial for maintaining network stability and reliability. EIGRP and OSPF both excel in their ability to reroute traffic swiftly and effectively in the event of a connection failure.

EIGRP utilizes the DUAL (Diffusing Update Algorithm) to determine the best route to a destination. This approach considers both primary and secondary paths, allowing for rapid convergence when a link fails. The feasibility condition in EIGRP, coupled with the DUAL process, enables routers to quickly identify alternative paths and redirect traffic, minimizing downtime and packet loss. This inherent design enhances EIGRP’s link-failover performance.

OSPF employs a different method for handling link failures. It uses the Hello protocol to establish neighbor relationships and detect link failures. When a failure occurs, OSPF routers utilize the Shortest Path First (SPF) algorithm to recalculate optimal routes to the affected destinations. Although OSPF may take slightly longer to converge compared to EIGRP, it still provides reliable link-failover capabilities. OSPF’s use of designated routers and backup designated routers further improves its ability to respond swiftly to changes in network topology.

Both protocols support route summarization, load balancing, and route redistribution, enhancing overall link-failover performance. The choice between EIGRP and OSPF often depends on factors such as network size, vendor preferences, and specific feature requirements. Ultimately, the effective link-failover performance of these routing protocols is critical for ensuring seamless and efficient communication in dynamic network environments.

5. IPv6 in Networking

IPv6 plays a crucial role in enhancing load-sharing and link-failure systems by providing advanced capabilities that improve network resilience and efficiency. The expanded address space of IPv6 allows for a more extensive allocation of IP addresses, facilitating the distribution of network traffic across multiple servers or routes. This scalability ensures balanced resource utilization, reduces congestion, and enhances overall network performance.

IPv6 also includes features like Stateless Address Auto Configuration (SLAAC) and Duplicate Address Detection (DAD), which are essential for link-failover systems. These features enable devices to automatically configure their IPv6 addresses and detect address conflicts, ensuring smooth transitions between alternative paths during link failures. The flexibility and adaptability of IPv6 contribute to continuous connectivity and efficient resource usage in dynamic network environments.

5.1. Improved Header Format

IPv6 introduces a more streamlined header format compared to its predecessor, IPv4. IPv6’s fixed header size of 40 bytes, in contrast to the variable-length header of IPv4, simplifies packet processing and reduces the processing burden on routers and network devices. This streamlined header includes only essential information, omitting unnecessary or rarely used functionalities, which accelerates packet forwarding and enhances network scalability. The improved header format aligns with IPv6’s broader goal of providing a more efficient and scalable foundation for the growing network of internet-connected devices.

5.2. Address Space Exhaustion

IPv6 addresses the issue of address space exhaustion, a significant limitation of IPv4. The limited number of IPv4 addresses creates bottlenecks as more devices connect to the network. IPv6’s vast address space eliminates these scalability challenges, ensuring that every device participating in load-sharing or link-failure scenarios can be uniquely identified with its IP address. This research emphasizes the practical implications of adopting IPv6 in networks where dynamic and efficient address allocation is critical for seamless operation.

5.3. Support for Quality of Service (QoS)

IPv6’s support for Quality of Service (QoS) is vital in optimizing load-sharing and link-failure systems. QoS mechanisms in IPv6 allow for the prioritization of different types of network traffic based on factors like latency, jitter, and bandwidth requirements. In load-sharing scenarios, this ensures that critical applications or services receive priority, preventing congestion and improving overall network performance. Similarly, in link-failover systems, QoS maintains an uninterrupted user experience by prioritizing essential traffic even during link transitions. IPv6’s QoS capabilities provide enhanced control over network resource management, creating a more flexible, efficient, and reliable network infrastructure.

5.4. Enhanced Security Features

IPv6 includes built-in security features such as mandatory support for IPsec at the network level. IPsec provides authentication and encryption to secure the confidentiality and integrity of data transmitted across the network. These security mechanisms are particularly important in load-sharing and link-failure scenarios, where secure data transfer and smooth network transitions are critical to maintaining network reliability and confidentiality. IPv6’s emphasis on security enhances the overall dependability of load-sharing and link-failure systems, establishing a more secure foundation for modern networking environments.

6. Methodology and Performance Metrics Measurement

The network topology for this study was meticulously designed using GNS3, incorporating thirteen Cisco routers configured with OSPF, EIGRP, and BGP protocols to simulate a complex network environment. Two virtual machines were deployed on VMware to represent end-user devices, all configured with IPv6 to evaluate the latest iteration of the Internet Protocol. Detailed packet-level analysis was performed using Wireshark, which allowed for an in-depth examination of routing protocol behavior. IPerf3 was employed to generate traffic and measure critical performance metrics such as network delay, jitter, and packet loss across various scenarios.

Load-sharing and link-failover mechanisms were configured to assess how OSPF and EIGRP manage traffic distribution and handle link failures. The experiments systematically varied parameters such as network size, traffic load, and link conditions to provide a comprehensive understanding of the strengths and weaknesses of OSPF and EIGRP, particularly in comparison to BGP within an IPv6 framework. The results were meticulously analyzed to evaluate the performance and suitability of each protocol for specific network conditions, offering valuable insights for network administrators and designers in making informed decisions about protocol selection and network architecture.

Performance metrics were carefully measured as follows:

• Network delay:

The total network delay was calculated by summing propagation, transmission, queuing, and processing delays, as in Equation (1):

$$Total Delay=d_{prop}+d_{trans}+d_{queue}+d_{proc}$$

(1)

where $d_{prop}$ is the propagation delay, $d_{trans}$ is the transmission delay, $d_{queue}$ is the queuing delay, and $d_{proc}$ is the processing delay.

2.

Network jitter:

Measured in milliseconds, jitter is represented as the average deviation from the mean delay calculated in Equation (2).

$$J=\sqrt{{\displaystyle \frac{1}{N}}\sum _{i=1}^N{(D_i-\overline{D})}^2}$$

(2)

where J is jitter, $1/N$ is the total number of measured packets, $D_i$ is a delay of the i-th packet, and $\overline{D}$ is the average delay of all the measured packets. The packet loss ($P_{loss}$) percentage.

3.

Packet loss:

The packet loss percentage was calculated using Equation (3):

$$P_{loss}={\displaystyle \frac{NumberofLostPackets}{TotalNumberofSentPackets}}\times 100$$

(3)

4.

Convergence Time:

This metric refers to the time taken to detect and recover from a link failure, given by Equation (4):

$$T_{convergence}=Tim{e}_{recovery}-Tim{e}_{failure}$$

(4)

In evaluating these metrics, the study employed a structured experimental setup to simulate various network conditions, load-sharing configurations, and link-failure scenarios. Rigorous testing and statistical analysis were conducted to assess the efficiency of EIGRP and OSPF, providing a comprehensive understanding of their real-world application and performance in load-sharing and link-failover systems. The findings contribute to a better grasp of the strengths and limitations of each protocol, offering guidance for enhancing network efficiency and reliability.

There are three use cases of load-sharing and link-failover systems:

6.1. Normal Condition

In this condition, the initial configuration requires the intelligent distribution of data and Internet traffic between two ISPs called ISP-A and ISP-B, to maximize network resilience and performance. The data traffic is going from the overlay network (Tunnel) and Internet traffic is going from the underlay network. Because of its capacity and speed, ISP-A serves as the primary connection in this system for the majority of data traffic. In addition, ISP-B acts as a backup connection for data flow, guaranteeing load sharing to effectively disperse network traffic. Conversely, internet traffic is routed via ISP-B, which serves as a primary link, making use of its unique benefits, including reduced latency or cost. Automatic failover techniques ensure continuous connectivity and minimize service disruption in the case of a link breakdown or degradation on the major pathways by smoothly redirecting the impacted traffic to the secondary links. The quick and automatic reaction helps to create a more resilient and efficient network architecture by reducing downtime and guaranteeing constant connectivity. This dual-ISP strategy offers a strong failover mechanism, enhancing overall network efficiency and providing a dependable and resilient network architecture.

In this experiment for data traffic, ISP-A is working as a primary link and ISP-B is working as a secondary link. Similarly, for Internet traffic ISP-B is working as a primary link and ISP-A is working as a secondary link. In case of link failure, the traffic shifts to the secondary link accordingly.

6.1.1. EIGRP

In this scenario, EIGRP is configured to use Tunnel-1 as the primary connection and Tunnel-2 as the secondary connection within the overlay network for data traffic. This setup enables load sharing by distributing data flows across both tunnels, optimizing bandwidth utilization. If the primary link (Tunnel-1) experiences a failure, EIGRP dynamically reroutes data traffic to the secondary link (Tunnel-2) to maintain connectivity without interruption. EIGRP continuously monitors the health and status of these tunnels to ensure reliability. For internet traffic, EIGRP identifies ISP-A as the backup link and ISP-B as the primary link within the underlay network. This configuration facilitates both failover and load sharing for internet-bound traffic. EIGRP continuously assesses the condition of these links, and in the event of an issue with the primary link (ISP-B), it automatically redirects internet traffic to the secondary link (ISP-A). This dynamic routing capability of EIGRP ensures optimal use of network resources and enhances network resilience in the face of link failures. The overall network performance and reliability are significantly improved through this adaptive approach, which effectively balances data and internet traffic across multiple channels. The EIGRP topology with dual ISPs is illustrated in Figure 3.

The traceroute results for data and internet traffic originating from Branch-1 in the EIGRP-enabled network architecture are depicted in Figure 4. As shown in the diagram, internet traffic is routed through the underlay network via ISP-B, while data traffic from Branch-1 to Branch-2 is transmitted via the primary tunnel in the overlay network connected to ISP-A. This approach ensures an efficient distribution of both incoming and outgoing traffic, effectively balancing the load between the two ISP connections.

Similarly, the traceroute results for data and internet traffic from Branch-2 using EIGRP in the network architecture are presented in Figure 5. The findings indicate that internet traffic is directed through the underlay network via ISP-B, while data traffic between Branch-2 and Branch-1 travels through the primary tunnel in the overlay network connected to ISP-A. This method allows for load sharing by equally distributing both incoming and outgoing traffic across the two ISP connections.

6.1.2. OSPF

OSPF plays a critical role in enabling flexible and efficient routing decisions within the network. As a dynamic routing protocol, OSPF is particularly effective in disseminating routing information across an autonomous system involving multiple connections. In the current configuration, OSPF is responsible for directing data traffic along the most efficient paths. The data traffic traverses the overlay network (Tunnel), while internet traffic is managed by the underlay network. Tunnel-1 serves as the primary connection for data traffic, with Tunnel-2 acting as the backup. Additionally, OSPF manages internet traffic by designating ISP-B as the primary connection and ISP-A as the backup. This arrangement facilitates load sharing, provides link-failover capabilities, and optimizes network performance by distributing traffic across multiple links. In the event of a primary link failure, OSPF dynamically redirects traffic to the secondary link, thereby minimizing downtime and ensuring uninterrupted connectivity. The responsive and adaptive routing capabilities of OSPF enable the load-sharing and link-failover system to effectively utilize network resources while maintaining stable connectivity for both data and internet traffic. Figure 3 illustrates the EIGRP topology with dual ISPs, and Figure 6 presents the OSPF topology with dual ISPs under normal operating conditions.

The traceroute results for data and internet traffic originating from Branch-1 with OSPF enabled in the topology are shown in Figure 7. A closer inspection reveals that internet traffic is routed via ISP-B, while data traffic from Branch-1 to Branch-2 is transmitted through the primary tunnel assigned to ISP-A. This approach ensures efficient load sharing by distributing both incoming and outgoing traffic across the two ISP connections.

The traceroute results for data and internet traffic originating from Branch-2 while OSPF is active in the topology are depicted in Figure 8. The results show that data traffic from Branch-2 to Branch-1 is routed through the primary tunnel assigned to ISP-A, while internet traffic is directed via ISP-B. This configuration effectively distributes both incoming and outgoing traffic across the two ISP connections, optimizing network resource utilization.

6.2. When Link Failure Occurs in EIGRP

When a network link failure occurs on the primary connection, such as Tunnel-1 from the overlay network established by ISP-A for data traffic, EIGRP quickly detects the issue using its rapid convergence mechanisms. EIGRP then reroutes the data traffic to the alternative connection, Tunnel-2, which is established by ISP-B. EIGRP routers are equipped with various timers and protocols that enable them to promptly identify network changes. Upon detecting a fault, the affected router initiates the Diffusing Update Algorithm (DUAL), which recalculates the routing table to find an alternate path. This transition is facilitated by the exchange of routing information among routers, ensuring that all routers within the EIGRP domain are updated on the changes in network topology. Additionally, precomputed backup routes, or viable successor routes, further reduce convergence time.

For Internet traffic, when ISP-B is the primary connection and ISP-A is the backup, EIGRP manages a link failure on the primary connection by redirecting traffic to the secondary connection from the underlay network. This seamless failover between primary and secondary connections ensures continuous connectivity and minimizes interruptions during link failures. Figure 9 illustrates the link failure through ISP-A while EIGRP is operational in the network topology.

Traceroute results for data and Internet traffic from Branch-1 with EIGRP enabled are shown in Figure 10. When a link failure occurs between Branch-1 and ISP-A, the system smoothly transitions data traffic to the backup connection, reducing network disruptions and ensuring uninterrupted data connectivity. Consequently, if the primary tunnel (Tunnel-1) via ISP-A fails, data traffic quickly shifts to the secondary tunnel (Tunnel-2) via ISP-B. Internet traffic, which is already routed through ISP-B, continues smoothly. This setup enables ISP-B to efficiently manage both data and Internet traffic, maintaining a stable network flow.

If the primary Internet link connecting Branch-1 to ISP-B fails, as shown in Figure 11, the secondary link, which connects Branch-1 to ISP-A, takes over to maintain Internet connectivity from the underlay network. This failover mechanism ensures that Internet traffic is seamlessly redirected through ISP-A, preserving network operation during a primary link failure. Thus, both data and Internet traffic are routed via ISP-A, ensuring minimal downtime for Internet services. For Branch-2, data traffic continues through Tunnel-1 via ISP-A, and Internet traffic is routed through ISP-B. This configuration enhances network resilience, ensuring minimal downtime in the event of a link failure.

Traceroute findings for data and Internet traffic from Branch-1, when ISP-B fails, are shown in Figure 12. The failover mechanism activates, efficiently redirecting Internet traffic to the backup link, ISP-A, to avoid network disturbances and maintain uninterrupted connectivity. As a result, if the primary connection for Internet traffic via ISP-B becomes unavailable, traffic is automatically rerouted through ISP-A. Data traffic, already routed through ISP-A, ensures that both data and Internet traffic are effectively managed, maintaining network stability.

A similar failover scenario occurs at Branch-2, where ISP-A is the primary link for data traffic and ISP-B is the backup, while ISP-B is the primary link for Internet traffic and ISP-A is the backup. In the event of a link failure, EIGRP’s failover mechanism automatically shifts traffic to the secondary link, ensuring continuous network connectivity. This transition is crucial for maintaining stable and uninterrupted network services for businesses and organizations.

6.3. When Link Failure Occurs in OSPF

As a link-state routing protocol, OSPF continually exchanges routing information between routers to maintain an up-to-date map of the network. In this setup, OSPF ensures that data traffic primarily uses Tunnel-1 via ISP-A when it is operational, as it is designated the primary connection. If Tunnel-1 fails, OSPF dynamically reroutes data traffic to Tunnel-2 via ISP-B, enhancing network resilience. For Internet traffic, when ISP-B is the primary link and ISP-A is the secondary link, OSPF adjusts the routing based on real-time network conditions and link statuses. This dynamic adaptability is a key advantage of OSPF, improving network performance and ensuring optimal use of available resources. Figure 13 illustrates the link failure through ISP-A while OSPF is functioning.

Traceroute results for data and Internet traffic from Branch-2 are shown in Figure 14. When connectivity between Branch-2 and ISP-A fails, the system smoothly reroutes data traffic to the backup link, avoiding network disturbances and ensuring continuous data connectivity. If Tunnel-1 via ISP-A fails, data traffic automatically shifts to Tunnel-2 via ISP-B, while Internet traffic continues through ISP-B from the underlay network. Consequently, both data and Internet traffic are routed through the backup link via ISP-B, maintaining reliable network performance.

When the primary Internet connection between Branch-2 and ISP-B fails, the network quickly reroutes Internet traffic to the secondary connection through ISP-A. This automatic transition allows for a swift response to connection failures, ensuring uninterrupted Internet access despite the outage of the primary link. This failover strategy enhances network reliability and stability by dynamically adapting to changing conditions and maintaining connectivity for critical services. For Branch-1, data traffic uses Tunnel-1 via ISP-A and Internet traffic is routed through ISP-B from the underlay network. Figure 15 shows the link failure through ISP-B with OSPF in operation.

Traceroute results for data and Internet traffic from Branch-2 are shown in Figure 16. When the primary Internet connection through ISP-B fails, the system automatically reroutes Internet traffic through the secondary connection via ISP-A. Data traffic continues through Tunnel-1 via ISP-A. This configuration ensures that both data and Internet traffic are routed through ISP-A, effectively mitigating the impact of the primary connection failure and supporting a stable network design.

A similar scenario applies to Branch-1, where ISP-A is the primary link for data traffic from the overlay network and ISP-B is the secondary link. Conversely, ISP-B is the primary link for Internet traffic from the underlay network, and ISP-A is the secondary link. In the event of a link failure, OSPF’s failover mechanism automatically shifts traffic to the secondary link, reducing downtime and maintaining continuous communication. This redundancy approach enhances network resilience, minimizes the impact of connection failures, and ensures efficient traffic rerouting, thereby improving overall network stability and efficiency.

7. Results and Discussion

The evaluation of network performance in this study focused on key metrics such as network latency, network jitter, packet loss, and convergence time. The findings indicate that both EIGRP and OSPF have distinct strengths and weaknesses in terms of load sharing and link failover. Specifically, EIGRP demonstrated effective load-sharing capabilities, which helped reduce network latency and packet loss. On the other hand, OSPF excelled in convergence time, especially under link-failure conditions. The integration of BGP further enhanced the system’s resilience, making it more robust.

The tools used, such as Wireshark and IPerf3, provided deep insights into packet-level dynamics, offering a thorough understanding of protocol behavior. This research serves as a valuable resource for network managers and engineers seeking to optimize IPv6 performance and resilience by making informed decisions about protocol selection and parameter adjustments.

7.1. Network Delay

Without authentication, EIGRP exhibited delays ranging from 30.21 to 30.39 s, while OSPF delays ranged from 30.24 to 30.35 s. With authentication, EIGRP delays varied between 30.40 and 30.48 s, and OSPF delays ranged from 30.31 to 30.46 s. These results indicate that authentication introduces some variability in the delay levels of both EIGRP and OSPF. Figure 17 provides a detailed graph illustrating the network delay differences with and without authentication.

7.2. Network Jitter

Without authentication, EIGRP jitter values ranged from 35.66 to 44.11 ms, while OSPF jitter values varied between 37.01 and 45.64 ms. With authentication enabled, EIGRP jitter ranged from 35.96 to 45.05 ms, and OSPF jitter ranged from 35.91 to 45.14 ms. The inclusion of authentication slightly increased jitter in both routing protocols. Figure 18 shows a comprehensive graph of network jitter for visual analysis. Network jitter, which can be calculated as the variance in network delay, is also depicted in the graph.

7.3. Packet Loss

Packet loss was minimal, which is an indicator of good network performance. The experiment was repeated multiple times, with packet losses observed using the IPerf3 tool. For EIGRP, the packet loss rate was 5%, while OSPF experienced a slightly higher packet loss rate of 5.4%. Though the difference is minor, it suggests that OSPF may face more challenges in maintaining data integrity and reliable communication. This difference in packet loss could lead to retransmissions and increased delay, potentially compromising the reliability of OSPF. Figure 19 presents the packet loss data for both protocols.

7.4. Convergence Time

Convergence time is critical for network resilience, stability, and performance. During a link failure between Branch-1 and Branch-2 via the primary provider (ISP-A), an alternative route through a secondary provider (ISP-B) was established. The goal was to measure how quickly the network responded to the link loss and rerouted traffic through the new connection.

Without authentication, EIGRP took between 125 and 140 s to establish an alternate path, while OSPF required 150 to 170 s. On average, EIGRP recovered from a link failover in 134.2 s, whereas OSPF needed 159.1 s. This demonstrates that EIGRP outperforms OSPF in backup path determination and link-failure recovery under these conditions. Figure 20 illustrates the convergence time without authentication.

With authentication, EIGRP required 13 to 15 s to generate an alternate path, while OSPF needed 20 to 27 s. On average, EIGRP recovered in 14.3 s, compared to 24.3 s for OSPF. These results further highlight EIGRP’s superiority in handling backup path determination and link-failure recovery with authentication enabled. Figure 21 presents the convergence time data with authentication.

Additionally, the study notes that a relatively short hold time in BGP, particularly less than 10 s, can lead to frequent disconnections, causing network instability. To mitigate this, a hold time of at least 10 s or more is recommended to ensure a stable and continuous BGP connection, thereby enhancing network efficiency.

8. Conclusions and Future Work

This research highlights the distinct advantages of EIGRP over OSPF, particularly in terms of packet loss and link-failure convergence time when operating in an IPv6 environment. While both protocols exhibited minimal differences in network latency and jitter, the absence of authentication led to lower network latency for both. However, the introduction of authentication introduced variability in delay levels. EIGRP consistently demonstrated superior performance in link-failover scenarios, whether authentication was enabled or not, recovering faster than OSPF. Additionally, the lower packet loss rate observed with EIGRP suggests its better capability in handling network challenges, contributing to more reliable data transmission.

The importance of selecting an appropriate routing protocol is underscored by the significant impact that even minor variations in performance metrics can have on overall network performance and reliability. This study provides a comprehensive approach, expanding the focus beyond traditional link-failure scenarios to include a more holistic strategy that addresses both connection failover and load sharing across various traffic types in an IPv6 context. This research fills a crucial gap in existing literature, offering valuable insights for network managers and engineers seeking to optimize their networks through intelligent protocol selection and parameter tuning.

The future work of this study aims to further enhance network resilience by incorporating hardware failover techniques. By integrating hardware failover options, the research seeks to bolster the network’s defenses against potential hardware failures, ensuring continuous connectivity with minimal disruption to data and internet traffic. This expanded module will conduct an in-depth analysis of how the combination of EIGRP, OSPF, and BGP, when paired with hardware failover mechanisms, impacts critical performance metrics such as network latency, jitter, packet loss, and link-failure convergence time within an IPv6 environment.

By pursuing this avenue of research, the goal is to develop a more robust and resilient network infrastructure that can withstand both software and hardware challenges, ultimately leading to more reliable and efficient communication systems.