A flow-based IDS using Machine Learning in eBPF

1 code implementation • 19 Feb 2021 • Maximilian Bachl, Joachim Fabini, Tanja Zseby

eBPF is a new technology which allows dynamically loading pieces of code into the Linux kernel.

BIG-bench Machine Learning Network Intrusion Detection

Detecting Fair Queuing for Better Congestion Control

1 code implementation • 16 Oct 2020 • Maximilian Bachl, Joachim Fabini, Tanja Zseby

Delay-based congestion control can achieve the same throughput but significantly smaller delay than loss-based one and is thus ideal for these applications.

Networking and Internet Architecture

EagerNet: Early Predictions of Neural Networks for Computationally Efficient Intrusion Detection

1 code implementation • 27 Jul 2020 • Fares Meghdouri, Maximilian Bachl, Tanja Zseby

Fully Connected Neural Networks (FCNNs) have been the core of most state-of-the-art Machine Learning (ML) applications in recent years and also have been widely used for Intrusion Detection Systems (IDSs).

Network Intrusion Detection

LFQ: Online Learning of Per-flow Queuing Policies using Deep Reinforcement Learning

1 code implementation • 6 Jul 2020 • Maximilian Bachl, Joachim Fabini, Tanja Zseby

The increasing number of different, incompatible congestion control algorithms has led to an increased deployment of fair queuing.

Deep Reinforcement Learning Fairness +3

SparseIDS: Learning Packet Sampling with Reinforcement Learning

1 code implementation • 10 Feb 2020 • Maximilian Bachl, Fares Meghdouri, Joachim Fabini, Tanja Zseby

To minimize the computational expenses of the RL-based sampling we show that a shared neural network can be used for both the classifier and the RL logic.

Computational Efficiency Edge-computing +5

Explainability and Adversarial Robustness for RNNs

1 code implementation • 20 Dec 2019 • Alexander Hartl, Maximilian Bachl, Joachim Fabini, Tanja Zseby

Recurrent Neural Networks (RNNs) yield attractive properties for constructing Intrusion Detection Systems (IDSs) for network data.

Adversarial Robustness Feature Importance +1

Cocoa: Congestion Control Aware Queuing

1 code implementation • 23 Oct 2019 • Maximilian Bachl, Joachim Fabini, Tanja Zseby

Recent model-based congestion control algorithms such as BBR use repeated measurements at the endpoint to build a model of the network connection and use it to achieve optimal throughput with low queuing delay.

Networking and Internet Architecture

Walling up Backdoors in Intrusion Detection Systems

1 code implementation • 17 Sep 2019 • Maximilian Bachl, Alexander Hartl, Joachim Fabini, Tanja Zseby

Interest in poisoning attacks and backdoors recently resurfaced for Deep Learning (DL) applications.

Autonomous Driving Intrusion Detection

City-GAN: Learning architectural styles using a custom Conditional GAN architecture

1 code implementation • 3 Jul 2019 • Maximilian Bachl, Daniel C. Ferreira

Generative Adversarial Networks (GANs) are a well-known technique that is trained on samples (e. g. pictures of fruits) and which after training is able to generate realistic new samples.