Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

NetScaNDN: A Scalable and Flexible Testbed To Evaluate NDN on Multiple Infrastructures

Amir Esmaeili, Maryam Fazli

TL;DR

NetScaNDN introduces a scalable, flexible testbed for evaluating IP and Named Data Networking (NDN) across wired and wireless infrastructures. It implements automated node discovery and configuration, dual-network operation, and a central syslog-based logging system to collect cross-layer metrics. The architecture is modular, with dedicated modules for discovery, network management, application deployment, and log management, and supports both IP routing (via Dijkstra or Bird) and NDN routing (via static or NLSR), along with prefix management like $/testbed/P_i$. Evaluation demonstrates the ability to run complex topologies and extract actionable metrics, with visualization and the capacity to compare forwarding strategies under failure scenarios, highlighting NetScaNDN’s potential to advance IP/NDN research and deployment readiness.

Abstract

The evolution from traditional IP-based networking to Named Data Networking (NDN) represents a paradigm shift to address the inherent limitations of current network architectures, such as scalability, mobility, and efficient data distribution. NDN introduces an information-centric approach where data is identified and retrieved based on names rather than locations, offering more efficient data dissemination and enhanced security. However, the transition to NDN, alongside the need to integrate it with existing IP infrastructures, necessitates the development of flexible and scalable testbeds that support diverse experimental scenarios across various physical media and networking protocol stacks. In this paper, we present NetScaNDN, a scalable, flexible, and plug-and-play testbed designed to facilitate such experiments. NetScaNDNl employs an automated process for node discovery, configuration, and installation, enabling seamless setup and execution of experiments on both wired and wireless infrastructures simultaneously. Additionally, it incorporates a central log repository using the syslog protocol, allowing comprehensive measurement and evaluation of user-defined metrics across different network layers. NetScaNDN offers a robust platform for researchers to explore and validate various networking scenarios, advancing the study of IP and NDN-based applications.

NetScaNDN: A Scalable and Flexible Testbed To Evaluate NDN on Multiple Infrastructures

TL;DR

NetScaNDN introduces a scalable, flexible testbed for evaluating IP and Named Data Networking (NDN) across wired and wireless infrastructures. It implements automated node discovery and configuration, dual-network operation, and a central syslog-based logging system to collect cross-layer metrics. The architecture is modular, with dedicated modules for discovery, network management, application deployment, and log management, and supports both IP routing (via Dijkstra or Bird) and NDN routing (via static or NLSR), along with prefix management like . Evaluation demonstrates the ability to run complex topologies and extract actionable metrics, with visualization and the capacity to compare forwarding strategies under failure scenarios, highlighting NetScaNDN’s potential to advance IP/NDN research and deployment readiness.

Abstract

The evolution from traditional IP-based networking to Named Data Networking (NDN) represents a paradigm shift to address the inherent limitations of current network architectures, such as scalability, mobility, and efficient data distribution. NDN introduces an information-centric approach where data is identified and retrieved based on names rather than locations, offering more efficient data dissemination and enhanced security. However, the transition to NDN, alongside the need to integrate it with existing IP infrastructures, necessitates the development of flexible and scalable testbeds that support diverse experimental scenarios across various physical media and networking protocol stacks. In this paper, we present NetScaNDN, a scalable, flexible, and plug-and-play testbed designed to facilitate such experiments. NetScaNDNl employs an automated process for node discovery, configuration, and installation, enabling seamless setup and execution of experiments on both wired and wireless infrastructures simultaneously. Additionally, it incorporates a central log repository using the syslog protocol, allowing comprehensive measurement and evaluation of user-defined metrics across different network layers. NetScaNDN offers a robust platform for researchers to explore and validate various networking scenarios, advancing the study of IP and NDN-based applications.
Paper Structure (14 sections, 5 figures)

This paper contains 14 sections, 5 figures.

Table of Contents

  1. Introduction
  2. Related works
  3. Named Data Networking (NDN) Overview
  4. Testbed Overview
  5. Controller Architecture
  6. Network Discovery Module (NDM)
  7. Network Management Module (NETMG)
  8. Application Configuration (AC)
  9. Log Repository Manager (LRM)
  10. Evaluation
  11. General Metrics
  12. Node Related Metrics
  13. Visualization
  14. Conclusion, Future Works

Figures (5)

  • Figure 1: NetScaNDN physical topology includes a controller, and nodes, as well as two wired, and wireless network infrastructures.
  • Figure 2: NetScaNDN controller architecture.
  • Figure 3: User defined topology via NetScaNDN's GUI includes 15 nodes. Link delays are automatically measured and shown in GUI
  • Figure 4: In this experiment, the testbed uses the best path and multicast forwarding strategies and results show link degrading throughput of C0 on link failure when the best route is used, also multicast forwarding strategy is resilient against link failure.
  • Figure 5: The benchmarking tests for memory and CPU utilization (a) as well as run times (b) using a Raspberry Pi3, Raspberry Pi4, and a Laptop machine.