Publications

We demonstrate a scalable processing of OPM data using ML to detect anomalies in optical services at run time. A dashboard will show operational SDN controller metrics, raw OPM data, and the ML assessment result.

Moving forward from 5G to 6G, one critical open challenge is customer-driven and autonomous networks management. To enable autonomy in a large-scale, it is important to separate the customer-centric requirements from the detailed network-and resource-centric operations.

A novel framework for in-network P4 processing of distributed telemetry data is presented, enabling effective soft failure detection and recovery strategies enforced in just few microseconds.

We experimentally demonstrate an SDN architecture for WDM VNTs to offload pass-through high bit-rate optical channels from overloaded ROADMs by provisioning spatial channels between the source and destination ROADM’s add-drop stages, and rerouting the optical channels.

This paper discusses the advantages and challenges of multiple architectures that consider the negotiation of inter-domain transport network slices using blockchain technologies. To this end, we present results obtained using cloud-native ETSI TeraFlowSDN controller.

In addition to the already present heterogeneity in the network service and application ecosystem, new challenges come in the form of emerging applications which require deterministic delay performance characteristics that go beyond traditional SLAs.

The Internet has been constantly evolving beyond its original purpose to ensure the reachability between endpoints wherever they may be located.

In spite of their popularity, developing secure smart contracts remains a challenging task. Existing solutions are either impractical as they do not support many complex real-world contracts or leave the burden to developers for fixing bugs.

Proceedings of the IEEE Blockchain Conference (IEEE Blockchain), 2022 (to appear) Abstract: Permissionless blockchains such as Bitcoin have long been criticized for their high computational and storage overhead.

In this work, we address this problem and present MITOSIS, a novel approach to practically improve scalability of permissioned blockchains

Solving the convergence issues of Generative Adversarial Networks (GANs) is one of the most outstanding problems in generative models. In this work, we propose a novel activation function to be used as output of the generator agent.

Network Digital Twin (NDT) is a new technology that builds on the concept of Digital Twins (DT) to create a virtual representation of the physical objects of a telecommunications network.

In spite of their popularity, developing secure smart contracts remains a challenging task. Existing solutions are either impractical as they do not support many complex real-world contracts or leave the burden to developers for fixing bugs.

Due to the growing rise of cyber attacks in the Internet, the demand of accurate intrusion detection systems (IDS) to prevent these vulnerabilities is increasing. To this aim, Machine Learning (ML) components have been proposed as an efficient and effective solution.

This demonstration will showcase the end-to-end orchestration of virtual network functions in the full-fledged ADRENALINE Testbed Cloud Platform expanding from the edge to the cloud.

Abstract: Scenarios related to 5G and beyond give rise to a high degree of heterogeneity in terms of applications, services, and user expectations as well as more demanding QoS requirements with an end-to-end scope that can cover multiple operator domains.

Today's networks support a great variety of services with different bandwidth and latency requirements. To maintain high user satisfaction and efficient resource utilization, providers employ traffic shaping.

Data center (DC) interconnection allows us to have optical transmissions between DCs directly connected to optical networks, avoiding the use of a packet-based infrastructure.

We demonstrate zero-touch device bootstrapping, monitoring, and L3-VPN service management using the novel TeraFlow OS SDN controller prototype.

Unsupervised learning (UL) is a technique to detect previously unseen anomalies without needing labeled datasets. We propose the integration of a scalable UL-based inference component in the monitoring loop of an SDN-controlled optical network.

We experimentally demonstrate the dynamic reconfiguration of WDM VNTs in response to SDM spatial channel failures. We present an SDN control architecture with gRPC-telemetry and analytics to detect failures and restore failed virtual WDM links.

We provide an overview of SDN control architectures for WDM over SDM (WDMoSDM) optical networks. We address dynamic management of end-to-end connec- tivity services and WDM virtual network topologies (VNTs) for partially disaggregated multi-domain WDMoSDM networks.

The hierarchical token bucket (HTB) algorithm allows to specify per-flow bitrate guarantees and enables excess bandwidth sharing between flows of the same class. Additionally, it provides capabilities to prioritize the traffic of specific flows, potentially considering their delay demands.

We study the requirements that Control, Orchestration, and Management (COM) systems for optical networks should fulfil, present a minimal set of components they should implement, and propose our research point of view regarding the future needs for COM systems for optical networks.