ELASTIC D4.1: Wasm, eBPF and TEE enablement on edge

This deliverable presents the enablement and evaluation of WebAssembly (Wasm), extended Berkeley Packet Filter (eBPF), Trusted Execution Environments (TEEs), and Federated Learning (FL) for IoT edge nodes. It includes an assessment of lightweight Wasm runtimes, such as WAMR, for executing portable workloads, alongside proposed extensions to the WASI interface enabling secure, low-level hardware access. Furthermore, it explores the use of eBPF/XDP for in-kernel network filtering and observability, including its integration with AI-driven Intrusion Detection Systems (AI-IDS) to support real-time threat detection at the edge. The deliverable also investigates the applicability of TEEs for secure workload execution and remote attestation in constrained environments. AI processing and distributed AI workloads are analysed, leveraging Wasm's lightweight footprint and fast startup times. The findings contribute architectural guidelines, performance evaluations, and preliminary insights toward the realisation of trusted and efficient data processing and orchestration across the edge–cloud continuum.