Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Framework for Integrating Machine Learning Methods for Path-Aware Source Routing

Anees Al-Najjar, Domingos Paraiso, Mariam Kiran, Cristina Dominicini, Everson Borges, Rafael Guimaraes, Magnos Martinello, Harvey Newman

TL;DR

The paper addresses the challenge of scalable, congestion-aware TE by integrating an AI-driven optimizer (Hecate) with a path-aware routing engine (PolKA) in an SDN context. It formalizes TE as a constrained optimization problem and demonstrates a data-driven framework where telemetry informs QoS predictions to select optimal SR paths. The work validates the approach in an emulated P4/testbed setting using a wireless QoS dataset, revealing that Random Forest Regression yields strong bandwidth predictions that feed PolKA decisions, and showcasing agile path migration and load balancing. This framework advances practical self-driving networks by translating ML-driven insights into real-time SR-based routing decisions with a telemetry backbone and emulated testbed support.

Abstract

Since the advent of software-defined networking (SDN), Traffic Engineering (TE) has been highlighted as one of the key applications that can be achieved through software-controlled protocols (e.g. PCEP and MPLS). Being one of the most complex challenges in networking, TE problems involve difficult decisions such as allocating flows, either via splitting them among multiple paths or by using a reservation system, to minimize congestion. However, creating an optimized solution is cumbersome and difficult as traffic patterns vary and change with network scale, capacity, and demand. AI methods can help alleviate this by finding optimized TE solutions for the best network performance. SDN-based TE tools such as Teal, Hecate and more, use classification techniques or deep reinforcement learning to find optimal network TE solutions that are demonstrated in simulation. Routing control conducted via source routing tools, e.g., PolKA, can help dynamically divert network flows. In this paper, we propose a novel framework that leverages Hecate to practically demonstrate TE on a real network, collaborating with PolKA, a source routing tool. With real-time traffic statistics, Hecate uses this data to compute optimal paths that are then communicated to PolKA to allocate flows. Several contributions are made to show a practical implementation of how this framework is tested using an emulated ecosystem mimicking a real P4 testbed scenario. This work proves valuable for truly engineered self-driving networks helping translate theory to practice.

Framework for Integrating Machine Learning Methods for Path-Aware Source Routing

TL;DR

The paper addresses the challenge of scalable, congestion-aware TE by integrating an AI-driven optimizer (Hecate) with a path-aware routing engine (PolKA) in an SDN context. It formalizes TE as a constrained optimization problem and demonstrates a data-driven framework where telemetry informs QoS predictions to select optimal SR paths. The work validates the approach in an emulated P4/testbed setting using a wireless QoS dataset, revealing that Random Forest Regression yields strong bandwidth predictions that feed PolKA decisions, and showcasing agile path migration and load balancing. This framework advances practical self-driving networks by translating ML-driven insights into real-time SR-based routing decisions with a telemetry backbone and emulated testbed support.

Abstract

Since the advent of software-defined networking (SDN), Traffic Engineering (TE) has been highlighted as one of the key applications that can be achieved through software-controlled protocols (e.g. PCEP and MPLS). Being one of the most complex challenges in networking, TE problems involve difficult decisions such as allocating flows, either via splitting them among multiple paths or by using a reservation system, to minimize congestion. However, creating an optimized solution is cumbersome and difficult as traffic patterns vary and change with network scale, capacity, and demand. AI methods can help alleviate this by finding optimized TE solutions for the best network performance. SDN-based TE tools such as Teal, Hecate and more, use classification techniques or deep reinforcement learning to find optimal network TE solutions that are demonstrated in simulation. Routing control conducted via source routing tools, e.g., PolKA, can help dynamically divert network flows. In this paper, we propose a novel framework that leverages Hecate to practically demonstrate TE on a real network, collaborating with PolKA, a source routing tool. With real-time traffic statistics, Hecate uses this data to compute optimal paths that are then communicated to PolKA to allocate flows. Several contributions are made to show a practical implementation of how this framework is tested using an emulated ecosystem mimicking a real P4 testbed scenario. This work proves valuable for truly engineered self-driving networks helping translate theory to practice.
Paper Structure (17 sections, 3 equations, 12 figures)

This paper contains 17 sections, 3 equations, 12 figures.

Table of Contents

  1. Introduction
  2. Background and Motivation
  3. Need for Traffic Engineering and AI
  4. Source Routing
  5. Problem Formulation
  6. Optimizing min-max problem
  7. Framework Integration for Data-driven Learning
  8. Experiment Implementation
  9. Training Hecate: Data-driven Learning
  10. Dataset
  11. Supervised ML and Prediction Methods
  12. ML Experimental Evaluation
  13. PolKA enabling path-aware networks
  14. Virtual testbed with freeRtr
  15. Experiments
  16. ...and 2 more sections

Figures (12)

  • Figure 1: PolKA source routing tool uses polynomial identifier to guide the packet through the network.
  • Figure 2: Simple network setup with multiple routes having different QoS parameters.
  • Figure 3: PolKA-Hecate integration framework
  • Figure 4: Sequence diagram of Hecate-PolKA framework
  • Figure 5: Wireless bandwidth measurement of LTE and WiFi over a selected path at UQ
  • ...and 7 more figures