Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

FedAuxHMTL: Federated Auxiliary Hard-Parameter Sharing Multi-Task Learning for Network Edge Traffic Classification

Faisal Ahmed, Myungjin Lee, Suresh Subramaniam, Motoharu Matsuura, Hiroshi Hasegawa, Shih-Chun Lin

TL;DR

This work tackles privacy-preserving network edge traffic classification under data heterogeneity and limited labels by proposing FedAuxHMTL, a two-tier edge-server/base-station federated multi-task learning framework that uses auxiliary tasks with hard-parameter sharing. A 1D-CNN multi-task local model is trained with a random loss weighting scheme to balance auxiliary tasks and strengthen the main task, achieving faster convergence and lower global loss. Empirical evaluation on the QUIC dataset using Cisco Flame demonstrates significant reductions in communication cost (≈53%) and computing time (≈32%) and improved accuracy for the main task compared with FedAvg and MT-DNN-FL. The approach offers practical benefits for low-overhead, privacy-preserving network traffic classification in edge environments and sets the stage for further theoretical analyses of federated multi-task learning with auxiliary tasks.

Abstract

Federated Learning (FL) has garnered significant interest recently due to its potential as an effective solution for tackling many challenges in diverse application scenarios, for example, data privacy in network edge traffic classification. Despite its recognized advantages, FL encounters obstacles linked to statistical data heterogeneity and labeled data scarcity during the training of single-task models for machine learning-based traffic classification, leading to hindered learning performance. In response to these challenges, adopting a hard-parameter sharing multi-task learning model with auxiliary tasks proves to be a suitable approach. Such a model has the capability to reduce communication and computation costs, navigate statistical complexities inherent in FL contexts, and overcome labeled data scarcity by leveraging knowledge derived from interconnected auxiliary tasks. This paper introduces a new framework for federated auxiliary hard-parameter sharing multi-task learning, namely, FedAuxHMTL. The introduced framework incorporates model parameter exchanges between edge server and base stations, enabling base stations from distributed areas to participate in the FedAuxHMTL process and enhance the learning performance of the main task-network edge traffic classification. Empirical experiments are conducted to validate and demonstrate the FedAuxHMTL's effectiveness in terms of accuracy, total global loss, communication costs, computing time, and energy consumption compared to its counterparts.

FedAuxHMTL: Federated Auxiliary Hard-Parameter Sharing Multi-Task Learning for Network Edge Traffic Classification

TL;DR

This work tackles privacy-preserving network edge traffic classification under data heterogeneity and limited labels by proposing FedAuxHMTL, a two-tier edge-server/base-station federated multi-task learning framework that uses auxiliary tasks with hard-parameter sharing. A 1D-CNN multi-task local model is trained with a random loss weighting scheme to balance auxiliary tasks and strengthen the main task, achieving faster convergence and lower global loss. Empirical evaluation on the QUIC dataset using Cisco Flame demonstrates significant reductions in communication cost (≈53%) and computing time (≈32%) and improved accuracy for the main task compared with FedAvg and MT-DNN-FL. The approach offers practical benefits for low-overhead, privacy-preserving network traffic classification in edge environments and sets the stage for further theoretical analyses of federated multi-task learning with auxiliary tasks.

Abstract

Federated Learning (FL) has garnered significant interest recently due to its potential as an effective solution for tackling many challenges in diverse application scenarios, for example, data privacy in network edge traffic classification. Despite its recognized advantages, FL encounters obstacles linked to statistical data heterogeneity and labeled data scarcity during the training of single-task models for machine learning-based traffic classification, leading to hindered learning performance. In response to these challenges, adopting a hard-parameter sharing multi-task learning model with auxiliary tasks proves to be a suitable approach. Such a model has the capability to reduce communication and computation costs, navigate statistical complexities inherent in FL contexts, and overcome labeled data scarcity by leveraging knowledge derived from interconnected auxiliary tasks. This paper introduces a new framework for federated auxiliary hard-parameter sharing multi-task learning, namely, FedAuxHMTL. The introduced framework incorporates model parameter exchanges between edge server and base stations, enabling base stations from distributed areas to participate in the FedAuxHMTL process and enhance the learning performance of the main task-network edge traffic classification. Empirical experiments are conducted to validate and demonstrate the FedAuxHMTL's effectiveness in terms of accuracy, total global loss, communication costs, computing time, and energy consumption compared to its counterparts.
Paper Structure (16 sections, 10 equations, 2 figures, 3 tables)

This paper contains 16 sections, 10 equations, 2 figures, 3 tables.

Table of Contents

  1. Introduction
  2. The State-of-the-Arts
  3. Learning Problem Explanations
  4. System Model
  5. Local Computation & Energy Consumption Model
  6. Communication Costs Model
  7. FedAuxHMTL: Federated Auxiliary Hard-parameter sharing Multi-Task Learning
  8. Random Loss Weighting
  9. Performance Evaluation
  10. Simulation Setup
  11. Dataset
  12. Simulation Platform
  13. Performance Metrics
  14. Baseline Benchmarks
  15. Simulation Results
  16. ...and 1 more sections

Figures (2)

  • Figure 2: Hard-parameter sharing federated multi-task learning in ES-BS.
  • Figure 3: (a) Test Accuracy vs the Number of Communication Rounds for the Main Task and Baselines, (b) Test Accuracy vs the Number of Communication Rounds for the Auxiliary Task2 and Baselines, (c) Total Global Loss vs the Number of Communication Rounds.