Security-by-Design at the Telco Edge with OSS: Challenges and Lessons Learned
Carmine Cesarano, Alessio Foggia, Gianluca Roscigno, Luca Andreani, Roberto Natella, Carmine Cesarano 1 , Alessio Foggia 1 , Gianluca Roscigno 2 , Luca Andreani 3 , Roberto Natella 1 · 2025
Read Security-by-Design at the Telco Edge with OSS: Challenges and Lessons Learned by Carmine Cesarano, Alessio Foggia, Gianluca Roscigno, Luca Andreani, Roberto Natella, Carmine Cesarano 1 , Alessio Foggia 1 , Gianluca Roscigno 2 , Luca Andreani 3 , Roberto Natella 1 (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2505-00111.txt.
Introduction
Five authors report lessons from securing a telecom edge platform with open source tools in this 2025 paper. The platform, GENIO, runs edge computing over passive optical network infrastructure, and the team applied security by design inside an industrial research project.
They mapped threats and matched them with defenses: system hardening, vulnerability management, digital signatures, and static and dynamic code analysis, all using open source software. The paper's value is practical. It records where these free tools proved mature enough for industry and where they fell short, so other teams building on open source at the network edge can plan around the gaps.
Catalog record
| CREATOR | Carmine Cesarano, Alessio Foggia, Gianluca Roscigno, Luca Andreani, Roberto Natella, Carmine Cesarano 1 , Alessio Foggia 1 , Gianluca Roscigno 2 , Luca Andreani 3 , Roberto Natella 1 |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 30 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2505-00111
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Security-by-Design at the Telco Edge with OSS: Challenges and Lessons Learned Carmine Cesarano, Alessio Foggia, Gianluca Roscigno, Luca Andreani, Roberto Natella, Carmine Cesarano 1 , Alessio Foggia 1 , Gianluca Roscigno 2 , Luca Andreani 3 , Roberto Natella 1 Abstract This paper presents our experience, in the context of an industrial R&D project, on securing GENIO, a platform for edge computing on Passive Optical Network (PON) infrastructures, and based on Open-Source Software (OSS). We identify threats and related mitigations through hardening, vulnerability management, digital signatures, and static and dynamic analysis. In particular, we report lessons learned in applying these mitigations using OSS, and share our findings about the maturity and limitations of these security solutions in an industrial context. {carmine.cesarano2, alessio.foggia, roberto.natella}@unina.it aggiungere autori dalle aziende I Introduction Edge computing has emerged as a transformative paradigm in the telecommunications and industrial sectors, enabling low-latency data processing closer to the end-users [ 1 ] . Traditionally, edge computing and broadband access networks have operated independently, with edge workloads running on dedicated servers, and Passive Optical Networks (PON) providing high-speed broadband connectivity. However, leveraging PON hardware infrastructure for edge computing presents a promising opportunity to create a high-performance and cost-effective platform for running edge services [ 2 , 3 ] . Unlike centralized cloud environments, deploying workloads on PON infrastructures introduces security risks. A primary concern is the physical exposure of hardware in uncontrolled environments, increasing the risk of tampering and unauthorized access [ 4 , 5 ] . Security risks are amplified by the complexity of software architectures, which rely on Commercial Off-the-Shelf (COTS) and Open-Source Software (OSS) components for virtualization, software-defined networking, and orchestration. Software reuse provides flexibility and cost-efficiency, but can also introduce software vulnerabilities and expose to compromised software dependencies [ 6 ] . Multi-tenancy introduces more security risks, as different edge applications share the same infrastructure. Maintaining isolation between tenants is critical to prevent escalation of security attacks [ 7 ] . The GENIO project [ 8 ] is a joint R&D initiative between academic and industry partners, aiming to achieve a secure-by-design edge computing platform integrated with PON networks. Unlike conventional edge models that rely on dedicated servers, GENIO leverages existing PON hardware to host multi-tenant edge services, directly within the telecom infrastructure. This approach can optimize resource utilization and create new revenue opportunities for telecom operators, without the need for additional investments in dedicated servers. One of the main objectives of the GENIO project is to align the platform with security regulations, such as the European Cyber Resilience Act [ 9 ] and CE marking certification [ 10 ] . This objective shaped the platform by guiding threat mitigations. Despite the growing emphasis on security-by-design, existing frameworks are not directly applicable for the design of the GENIO platform. Technical standards, e.g., from CISA [ 11 ] and NIST [ 12 ] , provide high-level security guidelines, but these lack practical, actionable implementation details. In real-world deployments, security must be tailored to the specific hardware and software technology and to the operational constraints of heterogeneous industrial environments, thus requiring customized solutions. Academic research has proposed security-by-design methodologies for specific types of systems, such as cloud applications [ 13 ] , smart grids [ 14 ] , big data frameworks [ 15 ] , and 5G networks [ 16 ] . However, no blueprint is readily available for security-by-design in PON-based edge computing. This paper presents our findings from the design of the GENIO platform, covering security risks across hardware, OS, middleware, and applications. We evaluated the maturity and limitations of security mitigations, based on OSS solutions, and documented several challenges due to architectural constraints and software heterogeneity. By aligning security research with deployment realities, this work provides key lessons learned, bridging the gap between theoretical security models and practical implementation in an industrial context. II The GENIO Project The GENIO project aims to develop a platform that integrates edge computing capabilities into telecom central offices, leveraging PON equipment. The architecture spans three layers: the cloud layer, the edge layer, and the far-edge layer, enabling flexible application deployments based on latency and computation requirements. Deployment . As shown in Figure 1 , the far-edge layer includes Optical Network Units (ONUs), which are deployed in residential and business premises, and which connect users to the fiber network. In GENIO, ONUs are equipped with additional low-end computing resources, enabling them to run applications with ultra-low latency requirements. The edge layer includes Optical Line Terminals (OLTs), which are devices located in telecom central offices, and which are repurposed in GENIO to serve as edge computing hubs. Originally designed for PON connectivity management, OLTs are enhanced with additional hardware and software to provide computational and storage resources. This layer is optimized for applications with strict latency and bandwidth requirements, balancing performance and resource availability. For applications with less stringent latency requirements, the cloud layer offers high computational and storage resources. The cloud layer also behaves as the orchestration center, managing resources across the edge and far-edge layers, and handling complex tasks that exceed the capabilities of local devices. Use cases . GENIO is designed to serve both business users and end-users . Business users, such as service operators and enterprises, are providers of edge applications to be shared as container images on a public registry of the GENIO project. Examples include ML workloads, real-time analytics, IoT data processing (e.g., from smart meters, cameras, and sensors), and telecom network functions. Business users can leverage the GENIO platform through an Infrastructure-as-a-Service (IaaS) model, by leasing computing, storage, and networking resources on the edge to run their applications. End-users, which include individual customers and businesses, interact with the platform via a Software-as-a-Service (SaaS) model, by consuming edge applications provided by business users. Software Architecture . The GENIO platform a distributed and multi-layered architecture to manage network operations, resource allocation, and application deployment. In particular, OLTs are based on x86 COTS hardware, and they integrate several OSS technologies for software-defined networking (SDN), including ONOS [ 17 ] , VOLTHA [ 18 ] and Open Networking Linux (ONL) [ 19 ] , to dynamically and efficiently manage PON network resources. OLTs are extended to support secure multi-tenancy for running edge applications, based on OSS virtualization technologies. The physical resources of the OLT are managed using a cluster of virtual machines, managed using the Linux/KVM hypervisor. Edge applications can run in either hard isolation (in dedicated virtual machines) or soft isolation (in containers and network namespaces within the virtual machines), to accomodate different security and performance requirements. Additionally, GENIO uses Kubernetes and Proxmox [ 20 ] for orchestrating the virtual machines and containerized applications, allowing scalable, resilient, and efficient workload management based on current network and computational conditions. Fig. 1: GENIO deployment across cloud, edge and far-edge layers. III Threat Modeling for GENIO Securing the GENIO platform required developing a comprehensive threat model to identify risks across the cloud, edge, and far-edge layers. Using the STRIDE methodology [ 21 ] , we systematically identified threats, from physical tampering of ONUs and OLTs, to software vulnerabilities and misconfigurations of orchestration services. This process led to the categorization of risks into Infrastructure-level , Middleware-level , and Application-level threats.