Researchers have introduced OpenLain5G, an open-source framework designed to automate the deployment, orchestration, and systematic validation of cellular core networks across experimental testbeds. Developed by Willian R. Canazas-Rosas and colleagues, the platform addresses a fundamental hurdle in mobile network research: while telecommunications standards define Fourth Generation Evolved Packet System and Fifth Generation Standalone core protocols, they do not provide an integrated, ready-to-use software product for practical laboratory operation. Consequently, researchers, engineers, and telecommunications instructors have traditionally been forced to assemble disjointed open-source projects, configure intricate host routes, reconcile disparate subscriber credentials, and write ad-hoc deployment scripts to establish functional test environments. OpenLain5G resolves these procedural challenges by consolidating scenario provisioning, preflight host diagnostics, isolated containerized execution, real-time observation, and stage-specific protocol validation under a single control plane that transitions seamlessly between simulated software routines and physical radio-frequency trials.
The motivation behind the software emerges directly from the broader evolution of mobile telephony infrastructure toward Open Radio Access Networks, known as O-RAN. By adhering strictly to Third Generation Partnership Project standards and implementing vendor-neutral interfaces, open cellular architectures aim to eliminate the costly vendor lock-in and rigidity characterizing traditional monolithic telecommunications deployments. Although ongoing academic and industrial debate continues regarding whether real-world commercial O-RAN adoption will match aggressive market projections, the open architecture represents an indispensable paradigm for indoor enterprise services, academic research facilities, and experimental testbeds. However, building an end-to-end cellular test bench remains an intricate endeavor requiring an Evolved Packet Core or Fifth Generation Core, a compatible radio access network, subscriber credentials, data-network routing, and an optional IP Multimedia Subsystem for advanced communications. Without a cohesive management harness, academic teams frequently struggle to differentiate between container launch failures and protocol-level attachment errors.
To eliminate this operational friction, OpenLain5G provides an architectural framework structured around a lightweight software stack that coordinates independent cellular functions without altering their underlying codebases. Users interact with the framework through either a dedicated command-line interface named lain5g or an intuitive graphical web dashboard built using React and TypeScript. Both operator interfaces communicate directly with a backend application programming interface built on Python and FastAPI, which governs the experimental lifecycle and provides real-time diagnostic observation. When an operator triggers an experiment, the orchestrator evaluates a declarative profile template, securely merges it with an uncommitted local private configuration file, verifies prerequisite parameters, and generates an isolated multi-container runtime environment utilizing Docker Engine and Docker Compose version two. Public scenario templates define non-sensitive service configurations and virtual network topology, while sensitive information—including subscriber authentication keys, radio-frequency transmission profiles, and equipment serial numbers—remains strictly segregated within ignored local configuration files.
The released codebase supports four standardized experimental deployment profiles tailored to distinct phases of telecommunications research and education. For baseline computational reproduction without physical radio hardware, the software provides two completely virtualized environments: the 4g-lte-sim scenario and the 5g-sa scenario. The 4g-lte-sim profile integrates an Open5GS Evolved Packet Core with srsRAN Fourth Generation evolved NodeB and User Equipment software processes interconnected through a simulated ZeroMQ software-radio link. The 5g-sa profile integrates an Open5GS Fifth Generation Core with UERANSIM next-generation NodeB and simulated User Equipment instances. For laboratory environments equipped with transmission hardware, OpenLain5G provides two guarded radio-frequency profiles: 4g-lte-x310 and 5g-sa-x310. These physical profiles integrate compact Kamailio IP Multimedia Subsystem deployments, srsRAN radio access stacks, provisioned commercial subscriber identity module cards, and Ettus Research X-Series Universal Software Radio Peripheral devices. Furthermore, community software patches extend the native hardware compatibility of the srsRAN radio stack beyond USRP hardware to alternative platforms including ADALM-PLUTO, LimeSDR, LibreSDR, and BladeRF.
A central innovation of the OpenLain5G architecture is its rigorous stage-specific validation engine, which establishes transparent diagnostic evidence boundaries throughout every phase of network activation. In typical containerized deployments, monitoring tools merely report whether individual process containers are running, which often masks deeper protocol-level registration failures. OpenLain5G systematically decouples raw container service availability from protocol establishment, subscriber registration, data session creation, Internet Protocol address allocation, and user-plane network reachability. During each execution run, the validator interrogates the environment through sequential checkpoints, recording the exact stage at which any operational failure or protocol deviation occurs. The generated diagnostic telemetry includes the scenario identity, software release tag, configuration hash, execution runtime, summary check matrix, and bounded diagnostic service logs. By storing this structured execution evidence entirely apart from immutable scenario templates, the platform allows experimenters to perform direct comparisons across independent trials without altering the underlying procedural definitions or source code.
To ensure software reliability and reproducible computational execution, the development team implemented an exhaustive continuous quality assurance pipeline integrated directly into the source repository. The core test suite, executed locally via a consolidated automated software check target, runs three hundred twenty-three backend unit tests and fifty-one frontend verification tests, achieving seventy-eight percent line coverage across backend control modules. This automated validation routine enforces strict TypeScript static type-checking, compiles a production frontend bundle, validates declarative deployment profiles against strict schemas, checks release citation metadata, and scans the repository surface to prevent accidental inclusion of sensitive files or private credentials. In addition to local verification commands, an identical non-operational quality gate executes automatically through GitHub Actions on clean Ubuntu twenty-four point zero four virtual runners configured with Python version three point twelve and Node.js version twenty-two, confirming that the control logic remains robust, syntactically correct, and free from internal regressions across diverse execution environments.
The functional reproducibility of the platform was empirically established through five independent cloud-hosted workflow jobs running release version one point one point three on clean virtual infrastructure. Each validation ran on fresh GitHub-hosted Ubuntu twenty-four point zero four long-term support runners provisioned with four x86_64 virtual central processing units, sixteen gigabytes of random-access memory, and fourteen gigabytes of solid-state storage. Across all five clean virtual machine executions, both software-only profiles achieved a one hundred percent pass rate across their principal operational success criteria. The Fourth Generation simulation successfully completed all fourteen required functional checks during every run, yielding a cumulative total of seventy passed checkpoints, with profile execution durations ranging from forty-five to forty-nine seconds and averaging forty-seven point four seconds. The Fifth Generation Standalone scenario successfully completed all fifteen required validation checks across every iteration, totaling seventy-five passed checkpoints, with step execution durations ranging between twelve and seventeen seconds and averaging thirteen point four seconds. User-plane verification confirmed bidirectional Internet Control Message Protocol packet routing through the user equipment tunnel directly to the local packet core gateways at internal addresses 10.57.0.1 for LTE and 10.45.0.1 for 5G Standalone.
Beyond virtualized simulations, OpenLain5G incorporates a strictly controlled operational workflow to govern physical radio-frequency trials with commercial hardware. To prevent accidental interference with licensed telecommunications bands, the software disables and blocks all radio-frequency emissions by default, requiring explicit administrative authorization via the command-line interface or web dashboard before activating transmission hardware. Operating guidelines mandate controlled laboratory precautions, including cabled coaxial setups equipped with inline signal attenuators or radio-frequency shielded Faraday cages. During physical laboratory validation using Ettus X-Series software-defined radios and programmable subscriber identity module cards, commercial handsets successfully registered with both Fourth Generation and Fifth Generation networks. The LTE test handset established active voice and data connectivity at a measured signal level of negative eighty-eight decibels relative to one milliwatt, alongside verified IP Multimedia Subsystem SIP registration. The 5G Standalone commercial handset established active connectivity at negative eighty-three decibels relative to one milliwatt with valid IMS registration. Physical link performance evaluations conducted using Ookla Speedtest across a reference backhaul offering twelve megabits per second downlink and ten megabits per second uplink demonstrated ten megabits per second downlink and nine megabits per second uplink on LTE, and nine megabits per second downlink and three megabits per second uplink on 5G Standalone.
The framework couples this experimental flexibility with rigorous operational security controls and strict computational provenance tracking. Because executing Docker containers and managing host network interfaces requires privileged administrative access to the Docker daemon socket and file system, the graphical user interface defaults strictly to an observation-only mode. Mutation actions, including scenario provisioning and profile activation, require explicit opt-in confirmation by authorized operators. To maintain strict hygiene across public version control platforms, sensitive subscriber records, cryptographic keys, hardware identifiers, and frequency assignments are stored exclusively within local uncommitted directories protected by restrictive operating system permissions. Software reproducibility is further reinforced by an upstream version dependency matrix, documented input policies, immutable Git release tags, a partial CycloneDX Software Bill of Materials, and permanent archival on Zenodo under the permissive Massachusetts Institute of Technology open-source license. However, the authors emphasize that these laboratory security protections are designed solely for controlled research benches and do not convert the local application into a multi-tenant or public Internet-facing cloud service.
Despite its comprehensive automation capabilities, the authors delineate clear methodological boundaries regarding the experimental interpretation of OpenLain5G results. The reported execution durations and gateway reachability checks establish functional reproducibility rather than formal statistical benchmarks for latency, network jitter, or maximum channel throughput. Furthermore, while physical handset trials confirmed successful network attachment, IMS SIP registration, native Short Message Service delivery over NAS and IMS, and over-the-top messaging connectivity, they did not evaluate native circuit-switched or IMS-based Voice over LTE or Voice over New Radio call establishment, media transport performance, or Mean Opinion Score voice quality metrics. The researchers also distinguish between reproducible source code and reproducible binary execution environments, noting that because upstream container images are not immutably mirrored within the core repository, external dependency rebuilds could introduce untracked changes unless users explicitly pin image digests. Future development iterations will focus on publishing fully immutable container image sets, broadening automated clean-host verification routines with pinned external dependencies, extending validation to comprehensive VoLTE and VoNR voice services, and releasing structured empirical datasets complete with acquisition telemetry.
Subject of Research: Automated cellular core orchestration and stage-specific validation for experimental testbeds
Article Title: OpenLain5G: Ready-to-Use 4G/5G core software with orchestration and validation for testbeds
Article References: Canazas-Rosas, W. R., Prieto-Tito, M. I., Hilario-Tacuri, A., & Tamo, A. (2026). OpenLain5G: Ready-to-Use 4G/5G core software with orchestration and validation for testbeds. SoftwareX, 36, Article 103225. https://doi.org/10.1016/j.softx.2026.103225
Image Credits: AI Generated
DOI: 10.1016/j.softx.2026.103225
Keywords: OpenLain5G, Open5GS, srsRAN, UERANSIM, Kamailio, Docker Compose, FastAPI, O-RAN, 5G Standalone, LTE Evolved Packet Core, Universal Software Radio Peripheral, Software-Defined Radio, IP Multimedia Subsystem, ZeroMQ, Protocol Validation, Cellular Testbeds, Reproducible Research
News Source: Denise Maddox. (November 17, 2026). OpenLain5G Automates Cellular Core Orchestration and Stage-Specific Validation for Testbeds. Scienmag.



