New edge architecture: semantic serverless and P4 load balancing for intelligent applications
0709, 2026

New edge architecture: semantic serverless and P4 load balancing for intelligent applications

Within the SMARTY project, a new architecture integrates serverless computing (FaaS), P4-programmable networks and semantic communication to improve efficiency and reduce latency in edge environments by jointly considering resource usage and data semantics. A P4-based Load Balancer performs real-time traffic steering using CPU, GPU and memory utilization, node capabilities and packet-level semantics (e.g., person or vehicle), ensuring requests are routed to compatible and less loaded edge nodes. The system is

Read More

Enhancing AI-Based Intrusion Detection with GPU Acceleration for Next-Generation Networks
2608, 2026

Enhancing AI-Based Intrusion Detection with GPU Acceleration for Next-Generation Networks

Within the SMARTY project, a Graph Neural Network (GNN)-based solution for real-time malicious traffic detection has been enhanced to provide faster and more efficient AI-driven cybersecurity capabilities. The activity addresses a key limitation of the previous approach: the graph construction phase, which represented the main bottleneck in the overall prediction process. By redesigning the graph generation mechanism and moving this process from the CPU to the GPU, the solution enables

Read More

Protecting Sensitive Data through Confidential Computing
0508, 2026

Protecting Sensitive Data through Confidential Computing

Confidential Computing is one of the solutions explored within SMARTY to enhance the protection of sensitive applications and data in next-generation cloud and network infrastructures. This technology introduces advanced mechanisms to protect data during processing and use. Through Trusted Execution Environments (TEEs), applications run in isolated and protected environments, ensuring information confidentiality and integrity even within shared infrastructures. The SMARTY activity investigates the integration of Confidential Computing with network security

Read More

Improving Performance Control of Edge Microservices with Proactive Autoscaling
2807, 2026

Improving Performance Control of Edge Microservices with Proactive Autoscaling

Managing resources in cloud-edge applications is a complex challenge. While placing services closer to end users reduces latency and improves performance, it also introduces strict resource constraints and highly variable workloads. In this context, traditional autoscaling mechanisms, such as those used in Kubernetes, react only after changes in traffic have already occurred. This reactive approach can lead to inefficient resource usage and potential violations of Service Level Agreements (SLAs). To

Read More

SMARTY: SWARMification approach for resilient 5G/B5G networks
2107, 2026

SMARTY: SWARMification approach for resilient 5G/B5G networks

The SMARTY approach for 5G/B5G mobile network workloads targets trusted and resilient environments, where network and processing nodes can operate in a self-managed manner according to available AI/ML-based capabilities. The concept of SWARMification extends the capabilities of disaggregated Radio Access Networks, enabling dynamic adaptation of network operations through resilience mechanisms and interaction between RAN Intelligent Controllers and in-network computing resources. The approach combines AI/ML pipelines, policy exchange, conflict management and

Read More

AI platform for real-time analysis of spatio-temporal data
1307, 2026

AI platform for real-time analysis of spatio-temporal data

The SMARTY project develops an innovative platform based on artificial intelligence, cloud and edge computing for the real-time analysis of spatio-temporal data. The goal is to enhance the ability of digital systems to understand and interpret the physical environment in a fast and efficient way. The proposed technology is designed for Mixed Reality applications, where it is essential to detect and track objects in dynamic environments. To ensure high performance,

Read More

SMARTY: A Real-Time Graph Neural Network Approach for Malicious Traffic Detection
0707, 2026

SMARTY: A Real-Time Graph Neural Network Approach for Malicious Traffic Detection

The increasing complexity of cyberattacks makes traditional approaches ineffective for the timely detection of malicious network traffic. Graph Neural Networks (GNNs), while showing strong potential in traffic analysis, still face limitations related to scalability, latency and applicability in real-time scenarios. The SMARTY project addresses these challenges by introducing a new graph construction methodology designed to efficiently represent relationships between network flows and enable fast, accurate threat detection. The goal is

Read More

Corrective unlearning in GNNs with Smarty for removing corrupted data
2906, 2026

Corrective unlearning in GNNs with Smarty for removing corrupted data

With the increasing vulnerability of Graph Neural Networks (GNNs) to data manipulation and adversarial attacks, there is a growing need for techniques capable of selectively removing the influence of corrupted information after training, while preserving model performance. An approach to corrective unlearning in GNNs has been proposed in Smarty, enabling intervention even when only part of the manipulated data is known. The method combines a phase of identifying influenced nodes—based

Read More

Secure Optical IPsec for AI Training: SMARTY’s Post-Quantum Approach
2406, 2026

Secure Optical IPsec for AI Training: SMARTY’s Post-Quantum Approach

With the advancement of quantum computing, conventional cryptographic systems are becoming increasingly exposed to potential quantum attacks, creating the need for new security mechanisms to protect future digital infrastructures. Within SMARTY, in collaboration with QUARC (G.A. 101073355), the feasibility of integrating Post-Quantum Cryptography (PQC) into optical IPsec networks has been demonstrated to support confidential AI workloads. The proposed implementation combines NIST-selected PQC algorithms—Kyber, Falcon and Dilithium—with AES-256-GCM encryption to establish

Read More

1812, 2025

SMARTY’s Plenary Meeting in Malta

On 1 and 2 December, we welcomed SMARTY partners to St Julian’s, Malta, for our 4th Plenary Meeting, held shortly before the EFECS conference, where we also participated. Over these two days, colleagues from across Europe met in person to review our progress, exchange ideas, and plan the next steps for the SMARTY Chips JU Project. The agenda included technical updates, discussions on project milestones, and focused sessions on topics

Read More

3010, 2025

Unlocking Collaboration: SMARTY’s Confidential Workflows & Secure Accelerator Interface!

In today’s complex digital world, tackling big problems often requires multiple partners. But how do you collaborate effectively when each party needs to protect its sensitive data and intellectual property (IP)? At SMARTY, we’re building the future of secure cooperation with our Confidential Workflows!  We’ve developed a groundbreaking framework that enables multiple stakeholders to work together on sensitive data without compromising confidentiality. Imagine hospitals training an AI model on patient

Read More

2710, 2025

Smarter Monitoring: SMARTY’s Integrated Data Sketches for Traffic and Threat Detection! 

In today’s fast-paced digital landscape, efficient traffic monitoring is absolutely vital for detecting anomalies and cyberattacks. However, processing network traffic at “wire speed” is a significant challenge. Traditional methods, while effective, often involve using multiple separate data structures, which isn’t always the most efficient approach.  At SMARTY, we’re changing the game with our innovative work on Integrated Data Sketches for Traffic Monitoring and Threat Detection! We’re developing a novel, integrated

Read More

2310, 2025

Unleashing AI at Wire Speed: SMARTY’s Defensive Wirespeed AI! 

The digital world is moving faster than ever, and with it, the challenges of detecting evolving cyber threats. Traditional security methods simply can’t keep up with the sheer volume, speed, and complexity of today’s network traffic. That’s why at SMARTY, we’re developing game-changing solutions, including our innovative Defensive Wirespeed AI!  Our new methodology leverages the power of computer vision techniques for real-time network intrusion detection directly at the 5G base

Read More

2010, 2025

Securing the Speed: SMARTY’s Approach to Confidentiality in Optical Networks! 

Optical networks are the backbone of our high-speed, high-capacity digital world, demanding unparalleled performance and quality. But with their increasing complexity, manual monitoring for issues like ‘soft failures’ is no longer viable. Automated network management is key, but how do we ensure data privacy when sharing network data with third parties for fault detection?  At SMARTY, we’re tackling this challenge head-on with our research into the Effectiveness of Confidentiality-Preserving Clustering

Read More

1610, 2025

Detecting the Undetectable: SMARTY’s Real-Time GNN for Malicious Traffic! 

In an era where cyber threats constantly evolve, traditional detection methods often fall short. At SMARTY, we’re pioneering a powerful new approach: a Real-Time Graph Neural Network (GNN) for Malicious Traffic Detection. This innovative framework is designed to identify sophisticated and unknown attacks by modeling complex relationships within network traffic data.  Our GNN-based anomaly detection framework operates through a sophisticated three-stage pipeline:  What makes this particularly exciting is our focus

Read More

1310, 2025

Seamless Security: Discover SMARTY’s Transparent Communication Protection! 

In the complex landscape of digital operations, ensuring your data is secure at all times – whether in use, in transit, or at rest – is paramount. At SMARTY, we’re taking this a step further with our advancements in   Transparently Securing Communications, focusing on policy-controlled protection that works without you even noticing!   Our objective is to define a Software Defined Perimeter (SDP) in a declarative manner, allowing network communication to

Read More

0910, 2025

Quantum-Proof Your AI: SMARTY’s Next-Gen IPsec Tunnels! 

The future of computing is rapidly approaching, and with it, the challenge of quantum computers potentially breaking our current public key cryptography methods. At SMARTY, we’re not waiting; we’re actively developing solutions to withstand this “quantum threat” today, especially for confidential AI training!  We’re pioneering a Post-Quantum Secure Optical IPsec Tunnel for Confidential AI Training. This isn’t just theory; it’s a vital step towards securing the massive data flows inherent

Read More

0610, 2025

Unlocking Secure Data Processing: Dive into SMARTY’s Capability-Based Access Control! 

In today’s data-driven world, with the rise of hardware accelerators like GPUs, TPUs, and FPGAs transforming cloud platforms, ensuring stringent security and regulatory compliance (think GDPR and CCPA!) is absolutely critical. At SMARTY, we’re at the forefront of tackling these challenges with our innovative approach to Capability-Based Access Control.  Traditional trusted computing often relies on large Trusted Computing Bases (TCBs), offering limited fine-grained control over data processing and inter-device data

Read More

0210, 2025

The Brains Behind the Barrier: SMARTY’s SDP Architecture 

At SMARTY, we’re diving deep into the architecture that makes our Software Defined Perimeter (SDP)! Our SDP architecture is built upon three key entities: the SDP Controller, the client Initiating SDP Host (IH), and the endpoint Accepting SDP Host (AH).   The SDP Controller is the mastermind, determining who can communicate with whom. It can even integrate with external authentication services for attestation, geo-location, or identity verification. The IH requests connections,

Read More

2509, 2025

🛡️ Breaking New Ground in Network Security: Inside SMARTY’s Cutting-Edge Framework for Trustworthy Edge Computing 

The digital era demands unprecedented security, scalability, and intelligence in communication networks. In our latest milestone, the SMARTY project unveils the design and preliminary implementation of revolutionary secure networking solutions tailored for next-generation fibre-wireless ecosystems. This work lays the foundation for resilient, AI-driven infrastructures capable of thwarting evolving cyber threats while enabling seamless multi-stakeholder collaboration.  🔒 Core Innovations Highlighted:  As 5G/6G networks converge with edge computing, vulnerabilities multiply. SMARTY’s integrated

Read More

2708, 2025

Semantic-Driven Interoperability, Embedded AI, and Hardware/Software Integration in SMARTY 

Service Mediation via Semantic Web Mappings  The SMARTY Project addresses interoperability challenges through declarative semantic-web based service mediation, resolving both syntactic and semantic mismatches across systems. Traditional any-to-any mappings require pairwise translations between systems, leading to scalability issues. Instead, SMARTY adopts an any-to-one approach, where a central RDF-based reference model serves as a semantic pivot.  This approach ensures seamless integration across partners’ systems, critical for use cases like secure edge

Read More

2208, 2025

Semantic Management for Interoperable AI and Embedded Systems

The SMARTY Project addresses the challenge of interoperability across heterogeneous systems by leveraging semantic web technologies and structured metadata ontologies. This approach ensures seamless integration of hardware, software, datasets, and AI models in secure communication environments.  Interoperability Framework  Interoperability is achieved through a common metadata model that unambiguously describes digital assets—from hardware components to AI models. This model enables:  A central asset catalog (developed in WP5) acts as a command

Read More

1308, 2025

SAFE-SU: A Modular Performance Monitor Unit for Secure Hardware Acceleration 

The SMARTY Project has developed SafeSU, a Performance Monitor Unit (PMU) designed to enhance observability and controllability in secure hardware accelerators. SafeSU is AMBA AHB/AXI-compliant and integrates with multicore systems to monitor and manage contention, latency, and resource usage in real time.  Core Functionality  SafeSU consists of three key components:  Interfaces & Programmability   Key Features  Applications  SafeSU is particularly valuable in secure SoC designs, where real-time monitoring of multicore interference

Read More

1108, 2025

Post-Quantum Cryptography Acceleration for Edge Computing 

As quantum computing continues to evolve, existing public-key cryptographic schemes are becoming increasingly vulnerable. To address this, the SMARTY Project is advancing hardware acceleration for Post-Quantum Cryptography (PQC) within edge computing environments.  Led by BSC, the effort focuses on integrating PQC capabilities into the SELENE SoC, a RISC-V-based platform, using High-Level Synthesis (HLS) to accelerate the development of hardware modules that implement two NIST standardization candidates:  Design Highlights  System Integration 

Read More

0608, 2025

W-Band Wireless Prototype: Advancing Quantum-Resistant Communications  

The SMARTY Project is progressing toward a secure, high-speed, short-range wireless link through the development of a W-band prototype (75–110 GHz). This forms a key component of a quantum-resistant transceiver tailored for edge environments where secure communication is required but wired infrastructure is not available.  W-Band Prototype: Current Status  The initial laboratory setup has demonstrated a unidirectional uplink (UL) over a wireless W-band path, while the downlink (DL) remains fiber-based. 

Read More

3107, 2025

Building a Quantum-Resistant Transceiver for High-Speed, Secure Wireless Communication 

The SMARTY Project is advancing secure communications with the development of a quantum-resistant transceiver designed for short-range, high-speed wireless links. This innovation targets edge computing scenarios where fiber connections are unavailable or impractical, and secure, low-latency data exchange is critical.  Technical Overview:  Led by TUE, the prototype transceiver operates in the W-band (75–110 GHz) and is designed for chip-to-chip and short-range node communication. A J-band (220–330 GHz) version is also

Read More

2807, 2025

Power Meets Security: SMARTY’s Ultra-Low Power Processor for PQC & Edge AI  

What if your smart device could learn, protect, and process AI? And all without needing a bulky battery or cloud connection?  That’s exactly what we’re building in SMARTY with our Ultra-Low Power Processor for PQC and Edge-AI which is an innovation that combines next-gen security with AI capability in a compact, energy-efficient chip. As quantum computers advance, traditional encryption becomes more vulnerable. But PQC algorithms are computationally intensive. Because we

Read More

2707, 2025

From Maps to AR: SMARTY’s Blueprint for Smarter Vehicles 

Smarter navigation meets augmented reality! The SMARTY Project is developing cutting-edge software for connected vehicles, blending IGo Classic (precision navigation) with IGo AR (real-time augmented reality guidance). We designed for future V2X (vehicle-to-everything) integration.  This allows your car display to overlay real-time AR guidance onto the road! You will be able to see turns, hazards, and points of interest. That’s the vision behind SMARTY’s Connected Vehicle Software, developed with NNG. 

Read More

2407, 2025

How SMARTY Is Reinventing Secure Elements for the Quantum Era 

As quantum computing advances, traditional cryptographic systems face growing risks. In response, the SMARTY Project is developing a Quantum-Resistant Secure Element (QR-SE)—engineered to protect data and communication systems from quantum-enabled attacks.  This next-generation SE targets applications where long-term data security is critical, including autonomous vehicles, healthcare, and public sector systems.  Core Technical Focus Areas:  Current Achievements:  What’s Next?  SMARTY’s QR-SE reduces PQC key sizes by over 50% without compromising cryptographic

Read More

1707, 2025

Why Do Knowledge Graphs Need AI? 

As part of the SMARTY Project, we are exploring how AI enhances knowledge graph capabilities—not just for data storage, but for intelligent reasoning, search, and interaction.  To support this, SMARTY integrates Metaphactory, an advanced platform developed by metaphacts, designed to enable AI-driven knowledge management at scale.  Why Metaphactory?  What’s Next in SMARTY?  Our team is focused on:  We invite you to follow our progress and explore how AI and semantics

Read More

1407, 2025

How do we run real-time AI applications at the edge: fast, secure, and with minimal latency?

SMARTY’s Answer: An Enhanced Serverless Engine that’s deeply integrated with In-band Network Telemetry (INT) developed by our partner TKI.  In simple terms, this innovation helps us to deploy AI apps without waiting for a full server to spin up, monitor performance in real time, route traffic intelligently across edge and cloud, and scale instantly and heal automatically. Edge-based systems (like those in smart cars or smart cities) don’t have the

Read More

How can organizations securely share data and services without compromising control?
1007, 2025

How can organizations securely share data and services without compromising control?

That’s not just the future. It’s what we’re building right now at SMARTY through our work on Decentralized Feature Extraction (DFE), it is led by CNIT. The SMARTY project introduces Declarative Cooperation is a policy-driven framework that: Traditional access controls are rigid and prone to breaches. Declarative Cooperation offers: The first release is expected soon  and will be tested in SMARTY’s Telco use case (UC2), where it can boost performance and efficiency in modern

Read More

Smarter Networks, Smarter AI: SMARTY’s Decentralized Feature Extraction 
0807, 2025

Smarter Networks, Smarter AI: SMARTY’s Decentralized Feature Extraction 

What if your network could automatically extract and feed critical data to AI systems without compromising performance?  The SMARTY project is developing Decentralized Feature Extraction (DFE)—a breakthrough technology that:  - Extracts network/device features in real-time using NVIDIA Bluefield DPUs.  - Supports distributed AI  by decentralizing data collection.  - Boosts efficiency by offloading feature extraction to programmable NICs/DPUs.  Traditional centralized data collection creates bottlenecks. DFE enables: 

0307, 2025

What if your sensitive data could stay encrypted even while being processed? 

❌ Myth: Once your data is being processed, it’s no longer protected.  ✅ Reality: With Confidential Computing, your data stays protected even while it’s being used.  Confidential Computing is a next-generation security solution that’s being developed, and SMARTY project leverages Confidential Computing powered by SCONE to:  Traditional security protects data at rest and in transit. Confidential Computing adds the missing piece:  SCONE is already available, and development is ongoing to support more platforms and make it

Read More

3006, 2025

What if tiny devices could run advanced AI and quantum-resistant security on minimal power? 

The SMARTY project is developing cutting-edge firmware for ultra-low power processors that:  As AI moves to the edge and quantum computing looms, this tech ensures small devices stay secure, efficient, and future-proof.  What SMARTY is focusing on to build:   Our partner EESY Innovation is leading the firmware development in collaboration with Infineon. We are aiming to ensure flawless integration, strong security, and AI capabilities, all in a compact, ultra-efficient package.  👉 Curious how it works?   Visit

Read More

Ever wondered how AI systems handle massive data flows without slowing down?  
2606, 2025

Ever wondered how AI systems handle massive data flows without slowing down?  

The SMARTY project is pioneering GPU Network Direct IO—a game-changer for AI-intensive workloads!  Here’s why it matters:  GPU Network Direct IO aims to create a high-speed highway between the GPU and the network card — the part that talks to the outside world — making data processing 3x faster! In AI ecosystems, GPUs are the powerhouse, but traditional setups strain network bandwidth by routing data through CPUs. This creates bottlenecks.   The SMARTY Solution

Read More

SMARTY’s Plenary Meeting in Copenhagen
1906, 2025

SMARTY’s Plenary Meeting in Copenhagen

Last week, we welcomed SMARTY partners to Copenhagen for our 3rd Plenary Meeting, hosted at the Danish Industry (DI) premises in the city center. Over two days, colleagues from across Europe met in person to review our progress, exchange ideas, and plan the next steps for the SMARTY Chips JU Project. The agenda included technical updates, discussions on project milestones, and focused sessions on topics such as PQC accelerators, Software-Defined

Read More

2502, 2025

UPM Advances Trustworthy and Efficient AI with Edge and LLM Research

The SMARTY Chips JU Project continues to push the boundaries of secure and efficient AI integration. Two recent research publications within the project, authored by researchers from UPM, demonstrate significant progress in two key areas: trustworthy AI reasoning and efficient edge AI deployment. Enhanced FIWARE-Based Architecture for Cyberphysical Systems With Tiny Machine Learning and Machine Learning Operations: A Case Study on Urban Mobility Systems One study focuses on integrating TinyML

Read More

2801, 2025

SMARTY Use Case Owners Series: Meet Cooperative ADAS  

Continuing our series on SMARTY use case owners, we focus on the transformative work in advancing cooperative ADAS, redefining automotive safety and efficiency in connected vehicles.  This SMARTY use case tackles key challenges in current ADAS systems, including limited onboard AI capabilities, sensor constraints, latency issues, and non-quantum resilient communication. By leveraging edge computing and quantum-resilient technologies, the project introduces a collaborative perception framework to enhance safety and operational efficiency

Read More

2001, 2025

SMARTY Use Case Owners Series: Meet Technische Universität Berlin

Continuing our series on SMARTY use case owners, we turn the spotlight on TUB (Technische Universität Berlin) and their contributions to Advanced Driving Assistance Systems (ADAS) through the SMARTY project. TUB’s focus lies in advancing cooperative perception for driving assist systems, ensuring ADAS can reliably interpret and predict road-user movements in diverse and complex traffic scenarios. By leveraging SMARTY’s secure and innovative solutions, TUB is advancing ADAS development and testing.

Read More

1612, 2024

SMARTY Use Case Owners Series: Meet InbestMe 

Continuing our series on SMARTY use case owners, we spotlight InbestMe and their pioneering work to revolutionize investment portfolio management.  Working with SMARTY, InbestMe aims to address key challenges in automated and personalized investment management, including scalability, security, and operational efficiency. By integrating cutting-edge solutions, InbestMe ensures data integrity and privacy of sensitive financial data and operations within cloud environments.  Key advancements introduced by InbestMe include:  With these innovations, InbestMe streamlines operations

Read More

1012, 2024

SMARTY Use Case Owners Series: Meet Italtel

In our series on SMARTY use case owners, we spotlight Italtel and their innovative efforts to redefine secure peer-to-peer (P2P) communication for telco operators. Through SMARTY, Italtel aims to address the critical need for security, scalability, and trust in edge computing platforms that handle sensitive telecommunications data without compromise in service requirements. By focusing on P2P communication within edge networks, the plan is to meet the growing demand for communication

Read More

2711, 2024

SMARTY Use Case Owners Series: Meet Continental

Launching our series on use case owners, we highlight Continental and their contributions to securing on-board electronics within the automotive sector. Continental’s work within SMARTY highlights the critical role of cybersecurity in advanced automotive systems like Vehicle-to-Everything (V2X) communication and Advanced Driver Assistance Systems (ADAS). With secure on-board electronics, Continental protects these systems from unauthorized access, ensuring both vehicle safety and data integrity. Through SMARTY’s collaborative innovation, Continental advances security in connected

Read More

SMARTY Kick-off
1806, 2024

SMARTY Kick-off

Hosted by CNIT in the beautiful, historical city of Pisa, the SMARTY Project has launched with a kick-off event. The consortium, comprising 33 international partners across 11 countries, is set to develop advancements in AI systems focused on transparency, security, and accuracy, with significant implications for sectors such as healthcare and autonomous vehicles. At SMARTY, we’re committed to advancing digital rights with a focus on traceability, security, and privacy, in

Read More