Distributed Denial of Service (DDoS) attacks are one of the most harmful threats in today’s Internet, disrupting the availability of essential services. The challenge of DDoS detection is the combination of attack approaches coupled with the volume of live traffic to be analysed. In this paper, we present a practical, lightweight deep learning DDoS detection system called LUCID, which exploits the properties of Convolutional Neural Networks (CNNs) to classify traffic flows as either malicious or benign. We make four main contributions; (1) an innovative application of a CNN to detect DDoS traffic with low processing overhead, (2) a dataset-agnostic preprocessing mechanism to produce traffic observations for online attack detection, (3) an activation analysis to explain LUCID’s DDoS classification, and (4) an empirical validation of the solution on a resource-constrained hardware platform. Using the latest datasets, LUCID matches existing stateof-the-art detection accuracy whilst presenting a 40x reduction in processing time, as compared to the state-of-the-art. With our evaluation results, we prove that the proposed approach is suitable for effective DDoS detection in resource-constrained operational environments.

Deep Learning is emerging as an effective technique to detect sophisticated cyber-attacks targeting Industrial Control Systems (ICSs). The conventional detection approach is to learn the normal behaviour of the system, to be then able to label noteworthy deviations from it as anomalies. However, the normal behaviour of ICSs continuously evolves over time for multiple reasons, such as update/replacement of devices, workflow modifications or others. As a consequence, the accuracy of the anomaly detection process may be dramatically affected with a considerable amount of false alarms being generated. This paper presents DAICS, a novel deep learning framework with a modular design to fit in large ICSs. The key component of the framework is a 2-branch neural network that learns the changes in the ICS behaviour with a small number of data samples and a few gradient updates. This is supported by an automatic tuning mechanism of the detection threshold that takes into account the changes in the prediction error under normal operating conditions. In this regard, no specialised human intervention is needed to update the other parameters of the system. DAICS has been evaluated using publicly available datasets and shows an increased detection rate and accuracy compared to state-of-the-art approaches, as well as higher robustness to additive noise.

Network Virtualization (NV) is one of the most promising technique to enable innovation in today's network. A recent approach toward NV has been proposed through FlowVisor, whose aim is to leverage on the specific features of an OpenFlow-controlled network to share the same hardware forwarding plane among multiple logical networks. However, FlowVisor lacks some features to enable a full implementation of a NV architecture: the virtual topologies that can be established are restricted to subsets of the physical topology and it has no way for two slices to share flowspace and simultaneously prevent them from interfering with each other's traffic. In this work, an innovative system called ADVisor (ADvanced FlowVisor) which enhances FlowVisor while overcoming its major constraints is presented and a set of experimental results discussed to demonstrate its capability to provide an effective support toward a Network Virtualization architecture.

In the recent years, the complexity of the network data plane and their requirements in terms of agility has increased significantly, with many network functions now implemented in software and executed directly in datacenter servers. To avoid bottlenecks and to keep up with the ever increasing network speeds, recent approaches propose to move the software packet processing in kernel space using technologies such as eBPF/XDP, or to offload (part of it) in specialized hardware, the so called SmartNICs. This paper aims at guiding the reader through the intricacies of the above mentioned technologies, leveraging SmartNICs to build a more efficient processing pipeline and providing concrete insights on their usage for a specific use case, namely, the mitigation of Distributed Denial of Service (DDoS) attacks. In particular, we enhance the mitigation capabilities of edge servers by transparently offloading a portion of DDoS mitigation rules in the SmartNIC, thus achieving a balanced combination of the XDP flexibility in operating traffic sampling and aggregation in the kernel, with the performance of hardware-based filtering. We evaluate the performance in different combinations of host and SmartNIC-based mitigation, showing that offloading part of the DDoS network function in the SmartNIC can indeed optimize the packet processing but only if combined with additional processing on the host kernel space.