DESIRE6G

A joint webinar with DESIRE6G, PREDICT6G and DETERMENISTIC6G on 22 October: 'Enhancements towards management of time-sensitive networks' More information here: https://lnkd.in/dVSnYCbS

Find our latest newsflash here! https://lnkd.in/eT39f6AT

We invite you to subscribe to our YouTube channel dedicated to the DESIRE6G project! Stay updated with the latest advancements, insights, and innovations in 6G technology. #6G #Innovation #DESIRE6G #YouTube https://lnkd.in/eKt8y5gv

"5GDAD: A Deep Learning Approach for DDoS Attack Detection in 5G P4-based UPF" Our latest paper was presented last week at #HSPR2024 by Francesco Paolucci, soon available to download from our website. The team introduced 5GDAD, a novel framework designed to enhance the detection and mitigation of DDoS attacks in 5G networks. Their approach integrates P4-based User-Plane Function (UPF) telemetry data with Convolutional Neural Networks (CNNs) for superior performance in attack detection. Key Highlights: 🔺 Framework Architecture: Figure 1 shows the Proposed Framework Network Topology, which outlines how different network components work together. It uses a P4-based UPF to gather real-time data, which is analyzed by a CNN-based model to spot anomalies related to DDoS attacks. 🔺 Deep Learning Model: Figure 2 illustrates the 5GDAD Architecture for DDoS Detection. The model uses CNNs to analyze network traffic and detect attacks. It's designed to adapt to new attack patterns, ensuring high accuracy. 🔺 Performance: The method achieved 98.6% accuracy and a 98% F1-score in detecting DDoS attacks, demonstrating its effectiveness in managing 5G network traffic. 🔺 Real-Time Detection: By leveraging P4 telemetry data, the framework provides real-time monitoring and alerts, crucial for protecting network services. As 5G networks expand, they face new DDoS threats that require advanced detection techniques. 5GDAD offers a powerful solution by integrating P4 technology with deep learning, enhancing network security. #NetworkSecurity #DeepLearning #P4 #SNSJU #horizoneurope

"In-network Control for Flow Steering" Demo presented at #ICTON2024 by Angelos Dimoglis and Anestis Dalgkitsis. 🔗 Paper available on our website: https://bit.ly/40NHW06 Key Highlights: 🔺 In-band Network Telemetry (INT): Our demo showcases the use of INT for real-time, in-depth network visibility by embedding network state information directly into data packet headers. This enables precise, granular monitoring of network performance metrics. 🔺 Ultra-Fast Control Loops: By implementing decision-making processes at the data plane, we achieve rapid network reactions to dynamic conditions. This significantly reduces reaction times and improves measurement accuracy, allowing the network to adapt swiftly to changing demands and conditions. 🔺 Dynamic Flow Steering: We demonstrate flexible and dynamic flow steering by detecting congestion or trust degradation and steering traffic via alternative routes without CPU involvement. This enhances network reliability and performance by dynamically optimizing traffic paths. 🔺 Comprehensive Telemetry Collection: Utilizing a P4-programmable switch as a telemetry collector, we facilitate ultra-fast identification of bottlenecks and trust level changes in flow paths. This allows for real-time adjustments and maintains optimal network performance. See Figure 1 below for the Network topology and components setup. 🔍 Our innovative approach to in-network control and flow steering is crucial for modern networks, which require high levels of automation and real-time adaptability. By embedding control mechanisms directly into the network fabric, we pave the way for more resilient, efficient, and secure network infrastructures. This work is particularly relevant for applications demanding ultra-low latency, high reliability, and dynamic adaptability, such as IoT, edge computing, and critical communication services. Authors: Angelos Dimoglis, Faris Alhamed, Anestis Dalgkitsis, Andrea Sgambelluri, Francesco Paolucci, Chrysa Papagianni, Paola Grosso 🤝 Collaboration by UvA, Scuola Superiore Sant'Anna and CNIT - Consorzio Nazionale Interuniversitario per le Telecomunicazioni. #6G #smartnetworks #demo #dissemination #SNSJU #horizoneurope

2nd 6G-PDN workshop; Call for Papers! Our joint "6G deterministic programmable networks with AI" Workshop will take place during MobiCom 2024, on November 18 2024. 📅 New submission deadline: 15 August https://lnkd.in/eBvhu48q

Did you miss our newly published deliverable about DESIRE6G's Functional Architecture? Download it from our website! https://lnkd.in/eVuyANfD D2.2 outlines the functional components, their interactions, and the processes within the DESIRE6G system, providing a roadmap for deploying an AI-powered, zero-touch, cloud-native 6G network that supports ultra-reliable low-latency communications (URLLC) and other critical 6G services. Key innovations are: 💥 Dynamic Service Framework: This flexible framework allows applications to discover, instantiate, and utilize services on a per-request basis, ensuring efficient resource utilization and streamlined service management. 💥 Multi-Timescale Control Loops: By incorporating various inputs such as network topology, resource utilization, and key performance indicators (KPIs), these control loops enable infrastructure adjustments and multi-parameter optimizations, supported by AI to enhance performance and efficiency. 💥 Infrastructure Management Layer: This layer integrates the Virtualized Infrastructure Manager (VIM) with a hardware abstraction layer (HAL), enabling the seamless use of hybrid hardware systems and cloud-native automation for managing physical resources without complicating upper layers, particularly business logic. #6G #smartnetworks #research #SNSJU #dissemination

"Augmented Reality App with AI-based Pervasive Latency Monitoring of RAN and Programmable Metro Packet-Optical Networks" Last week, Andrea Sgambelluri presented this Demo at #ICTON2024, and the paper is now available on our website: https://bit.ly/40NHW06 Key Highlights: 🔸 Innovative AI Integration: For the first time, this demo showcases how data analytics from RAN controllers and metro network collectors are processed by a Multi-Agent System running AI algorithms to detect and mitigate latency anomalies. 🔸 Real-Time Performance: Nearly real-time end-to-end latency telemetry, crucial for applications like Augmented Reality that demand ultra-low latency. 🔸 Pervasive In-band Network Telemetry (INT): Leveraging P4-based components across different networks to ensure dynamic network resource adaptation based on traffic patterns. 🔸 Federated Testbed: Comprehensive testing and validation conducted on a federated testbed encompassing facilities in Pisa, Italy, and Barcelona, Spain (see Figure 1 attached). 🔍 This approach ensures optimal Quality of Service (QoS) performance through intelligent, real-time network re-optimization. Authors: F. Alhamed, M. Guaitolini, P. González, R. Berozashvili, L. Ismail, H. Shakespear-Miles, S. Barzegar, L. Velasco, M. Ruiz, J.J. Vegas Olmos, A. Sgambelluri and F. Paolucci 🤝 Collaboration by five DESIRE6G partners: Scuola Superiore Sant'Anna, NVIDIA, CNIT - Consorzio Nazionale Interuniversitario per le Telecomunicazioni, Universitat Politecnica de Catalunya and Accelleran! #6G #AI #AugmentedReality #NetworkOptimization #LatencyMonitoring #SNSJU #horizoneurope

⚡ DEADLINE EXTENDED!! There is time until August 15 to submit your contributions to the 2nd 6G-PDN workshop! 👉 Visit https://lnkd.in/eBvhu48q Co-organized with PREDICT-6G Project and DETERMINISTIC6G, during #MobiCom2024

📣 PAPER SUBMISSION DEADLINE EXTENDED❗ The 2nd workshop on 6G deterministic programmable networks with AI #6GPDN -which will take place during MobiCom 2024- is 🌟looking for papers🌟 that contribute to the advancement of research on 6G networks. 📅 Key dates: → NEW Submission deadline: 15 August. → Workshop: 18 November. Submitted papers may cover any of the following topics: 🔹 Programmable data planes for TSN. 🔹Enhancements towards 6G TSN and DetNet integration. 🔹Network softwarization for 6G. 🔹Programmable networking protocols. 🔹Programmable SDN and NFV: languages and architectures (P4 and others). 🔹Hardware acceleration for programmable network functions. 🔹Multitenant data planes. 🔹Orchestration and management of software-defined deterministic networks. 🔹Control and management of data plane programmable devices. 🔹Artificial intelligence for deterministic networks. 🔹In network machine learning. 🔹In-network service level tuning and optimization; QoS. 🔹High precision traffic monitoring/telemetry. 🔹Service assurance and fulfilment programmability. 🔹Slicing for 6G. 🔹Intent-based systems and digital twinning applied in 6G. 🔹Routing and scheduling algorithms for reliable and real-time IoT. 🔹Trustworthy edge and cloud computing architectures and services. 👥 Organised by DETERMINISTIC6G, DESIRE6G and #PREDICT6G More information: https://lnkd.in/dWNYSAUU #MobiCom #CallForPapers #6G #AI