Sustainability is becoming a central objective for cloud and edge infrastructures, particularly in telecommunications, where large-scale deployments increase energy costs and environmental impact. This paper evaluates WebAssembly (Wasm) as a next-generation execution layer, examining its ability to combine high performance with improved resource efficiency. Leveraging our proposed open-source WasmPulse benchmarking framework, we compare Wasm with containers and bare-metal across artifact size, startup time, networking, and resource usage. The results show that while bare-metal remains the most efficient, Wasm achieves performance close to containers while offering distinct sustainability benefits: up to 98% faster startup, significantly smaller artifacts, and more stable memory scaling under load. These advantages reduce computational overhead and resource consumption, enabling efficient and rapid service deployment.Obtained results position WebAssembly as a lightweight, performant, and sustainable execution model, with strong potential for telco-grade cloud-native infrastructures.

As mobile networks advance, so do the use cases and quality demands they must meet, such as achieving “seven nines” reliability. Meeting these demands requires implementing intelligent and autonomous networks. Observability plays a crucial role in enabling data-driven Orchestration within this context. This paper presents Telco-Observer, an innovative solution tightly integrated with Management & Orchestration systems, designed to collect and provide efficient access to telemetry data across distributed cloud native edge environments. To the best of our knowledge, no existing solutions fully incorporate Observability at the Telco edge while supporting efficient Orchestration. Our performance evaluation demonstrates that Telco-Observer incurs minimal overhead while collecting a broad spectrum of telemetry signals and providing fast data access.

The convergence of 5G and Edge computing has revolutionized the technology landscape, ushering in a new era of innovative use cases and accelerating the realization of an intelligent, fully connected digital world at an unprecedented pace. However, the demanding and ever-changing specifications of these emerging applications present significant challenges in terms of orchestration, requiring innovative solutions and efficient management strategies to ensure optimal performance and resource utilization. The mobility of end-users is a critical factor expected to significantly impact Edge operations. As a result, Edge-enabled 5G systems face the daunting task of tracking moving users and meeting their Quality of Experience (QoE) demands simultaneously, which presents a formidable technical challenge that must be addressed for the realization of a truly seamless and ubiquitous user experience. In this paper, we put forward a novel solution based on Reinforcement Learning to optimize the relocation of applications between Edge Hosts while meeting their demands in terms of latency and resource requirements. Our proposed solution aims to achieve a balance between two key objectives: maintaining the Quality of Service (QoS) of Edge Applications and optimizing the utilization of infrastructure resources. By achieving this balance, we can ensure that Edge-enabled 5G systems operate efficiently, with minimal resource wastage and maximal QoS satisfaction for end-users. The intensive simulations and evaluation results demonstrate its effectiveness in terms of QoS assurance and resource load bnlancing.

The convergence of 5G and Edge computing is fostering the development of innovative use cases making the dream of a fully connected, intelligent digital world almost true. However, the stringent requirements coupled to the high dynamicity of these new applications make their orchestration extremely challenging. Specifically, the mobility of end users will undeniably impact Edge operations. Indeed, the Edge-enabled 5G systems need to provide the capability to follow moving users while respecting their latency requirements. To achieve such a goal, we propose, in this paper, an innovative solution optimizing the relocation of applications between Edge clusters while respecting their requirements in terms of latency and resources requests. In doing so, our solution is capable of ensuring the QoS of Edge applications while optimizing the infrastructure resource consumption. Performance evaluation in real platform put forward the efficiency of our solution in terms of (i) QoS assurance, (ii) resource load balancing and, (iii) convergence time.

In this paper, we summarize the accomplished PhD work on data plane programmability for software datapaths in a virtualized network infrastructure. In particular, we explain our motivation, the up-to-date state of the art, and propose two new solutions for programmable software datapaths, namely P4rt-OVS and NIKSS. Finally, we present research outcomes and future work directions.

This paper introduces NIKSS (Native In-Kernel SDN Switch), a novel programmable software switch for Software-Defined Networking. NIKSS leverages P4 as a high-level programming abstraction, Portable Switch Architecture (PSA) as a fully-featured P4 forwarding model and eBPF as a packet processing engine. The main component of NIKSS is the PSA-eBPF compiler that translates P4 programs into code that runs inside the eBPF execution environment. However, enabling a high degree of programmability for data planes without sacrificing performance or operability is a challenge. Therefore, NIKSS provides two alternative designs of generated code, with different trade-offs between supported PSA features and achieved performance. We demonstrate that NIKSS can be used successfully as a P4 software switch, with an acceptable performance gap compared to native eBPF programming. We also provide an extensive performance evaluation, proving that NIKSS might be a viable alternative to existing software switches.

A hypervisor switch, such as Open vSwitch (OVS), plays a key role in the virtualized data centers implementing overlay networking to provide network isolation. The hypervisor switch is running on each compute node as a software application and switches packets between virtual machines. Software switches frequently require upgrading and customization of network protocol's stack to introduce novel or domain-specific networking techniques. However, it is still difficult to extend OVS to support new network features as it requires mastery of network protocol design, programming expertise, and familiarity with the complex codebase of OVS. Moreover, there is currently no solution that enables the deployment of network features recompilation of OVS. In this article, we present P4rt-OVS, an original extension of OVS that enables runtime programming of protocol-independent and stateful packet processing pipelines. It extends the forwarding model of OVS with the userspace Berkeley Packet Filter (uBPF), bringing a new extensibility mechanism. Moreover, P4rt-OVS comes with a P4-to-uBPF compiler, which allows developers to write data plane programs in the high-level P4 language, compile them to the BPF bytecode and inject them to the OVS packet processing pipeline at runtime. Our design results in a hybrid approach that provides P4 programmability without sacrificing the well-known features of OVS. We provide thorough performance results, including end-to-end performance tests, microbenchmarks and a performance test of an exemplary network application (Broadband Network Gateway). The performance evaluation shows that P4rt-OVS does not introduce significant processing overhead, yet enables runtime protocol extensions and stateful packet processing. Moreover, we discuss features of P4rt-OVS and provide programming guidelines to help developers achieve the best performance of P4 programs for P4rt-OVS.

In the paper, we present the framework to offload VNF's data plane to the cloud infrastructure by leveraging the capabilities of P4. This approach allows tenants to run data plane functions directly on the programmable switches. Moreover, we propose the P4-based VNF offloading management interface as the extension to the DPPx framework - the target-independent P4 plugin for OpenStack.

In this paper, we propose a P4-based Data Plane Programmability and Exposure framework (DPPx), which allows to enhance Network Function Virtualization (NFV) services. Network functions can be offloaded from VMs to the virtualized network infrastructure. We present a general approach, system architecture and use cases enabled by DPPx. The DPPx framework has been preliminarily evaluated in a testbed environment as a prototype implementation based on OpenStack Neutron, P4 language and BMv2 software switch.

In this paper we describe the IIP System designed, implemented and currently tested in the frame of the Polish national project entitled "Future Internet Engineering"1(in Polish "Inzynieria Internetu Przysz"- IIP). The IIP system uses virtualized network infrastructure that allows to set a number of, named, Parallel Internets essentially differing in solutions for the data and control planes. More precisely, we design three Parallel Internets: (1) IPv6 QoS that basically explores Diff Serv and NGN architectures and uses TCP/IP, (2) Content Aware Network (CAN) specially designed for efficient content delivery, and (3) Data Streams Switching (DSS) for handling constant rate traffic with strong QoS requirements. For the IIP System we design a novel network infrastructure, which is based on the virtualization of the network elements (links and nodes). Such approach is on the line with the expectations of the network infrastructure for the Future Internet. In the paper we mainly focus on the system architecture, Parallel Internets and some aspects of network virtualization.