OpenAirInterface Software Alliance

• OpenAirInterface Software Alliance is a non-profit consortium uniting academic and industrial experts to develop an open-source wireless stack compliant with 3GPP standards.
• OSA’s framework enables reproducible experimentation and validation in 4G, 5G, and 6G networks, with practical applications like multi-user MIMO and UL-TDoA positioning.
• Its transparent governance and flexible licensing facilitate deep collaboration and rapid innovation in areas such as network slicing, AI/ML integration, and 6G research.

The OpenAirInterface Software Alliance (OSA) is a non-profit industry-academic consortium that develops, governs, and maintains the OpenAirInterface (OAI) open-source software stack for 4G, 5G, and emerging 6G wireless networks. Headquartered in Sophia-Antipolis, France, with EURECOM as a founding member, OSA facilitates the convergence of industrial and academic contributions to create a comprehensive, standards-compliant wireless platform, spanning the Radio Access Network (RAN), Core Network (CN), and network management domains. OSA’s open-source framework, flexible licensing, and transparent governance have established OAI as a principal vehicle for R&D, validation, and pre-standardization work across cellular wireless, O-RAN, and emerging 6G research (Kaltenberger et al., 2024).

1. Historical Overview and Organizational Structure

OSA originated in the early 2000s with the initial development of OAI RAN prototypes at EURECOM on software-defined radios (SDRs), bridging 3GPP standards and experimental research. The formal establishment of OSA in 2014 catalyzed greater scaling, professional engineering, and the creation of meritocratic governance structures. The Board of Directors consists of EURECOM and partner representatives, complemented by a full-time core technical team responsible for development, lab validation, and codebase integration.

Advisory and working groups—with open monthly meetings—cover RAN, Core, OAM, CI/CD, and the 6G roadmap. All technical roadmaps, meeting minutes, and feature branch statuses are publicly accessible, promoting transparent community engagement (Kaltenberger et al., 2024).

2. Membership, Contribution, and Licensing Model

OSA membership has two categories: Associate Members (mainly academic and non-profit institutions, contribution- and usage-based, no fee) and Associate Partners (for-profit, with membership fees and strategic commitments). Technical contributions require signature of the OAI Contributor License Agreement (CLA), adherence to the OAI Public License v1.1, and compliance with FRAND patent pledges.

The OAI Public License, derived from Apache 2.0, introduces an additional patent-licensing clause harmonized with 3GPP FRAND IPR policies. Key provisions include royalty-free use for research/experimentation, patent pledges from contributors for commercial exploitation, and legal certainty for contributions relating to 3GPP standard essential procedures. This structure enables SEP holders to contribute with confidence and encourages deep industry engagement (Kaltenberger et al., 2024).

3. Core Projects: Architecture and Key Capabilities

3.1 OAI-RAN

OAI-RAN implements the 3GPP gNB with standard functional splits: DU (PHY, MAC, RLC), CU-UP (SDAP, PDCP), CU-CP (RRC), interconnected by 3GPP-defined F1 and E1 interfaces and O-RAN eCPRI (split 7.2) (Kaltenberger et al., 2024). Supported features include:

• OFDM-based PHY, TDD/FDD, SCS = 15/30/120 kHz, and BW up to 100 MHz (FR1), 200 MHz (FR2)
• MIMO (up to 4×4 DL, 2×2 UL), Turbo/LDPC/Polar coding
• Monolithic or fully split architectures via F1/E1/7.2e
• Hardware accelerator integration: NVIDIA Aerial (in-line FAPI), Intel/AMD look-aside bb-dev
• Interoperability: Open5GS, free5GC, commercial CNs, various O-RUs

OSA also enables real-time uplink Multi-User MIMO (MU-MIMO) via SRS-based channel estimation and regularized zero-forcing combining, as shown in cell-free O-RAN testbed experiments (Uçak et al., 14 Jan 2026). The scheduler supports non-orthogonal PRB reuse per MIMO layer, transparent to unmodified UEs.

3.2 OAI-5GC

The 5G Core (OAI-5GC) is based on Service-Based Architecture (SBA) using HTTP/2 + JSON over SBI between NFs (AMF, SMF, UPF, AUSF, UDM, UDR, NRF, NSSF, PCF, NWDAF, LMF) (Kaltenberger et al., 2024). Advanced features include:

• Network slicing (per UE/PDU), QoS Enforcement (eBPF in UPF)
• NWDAF-driven analytics for AI/ML
• LMF-based UL-TDoA positioning, compliant with NRPPa and LPP
• Deployment via Docker Compose, Helm, or Kubernetes

3.3 OAI-OAM

Operations, Administration, and Maintenance functions leverage FlexRIC near-RT RIC, E2 Agent, xApp SDK (C/C++/Python), and service models (KPM, RC, others), with sub-millisecond E2 latency on ARM/x86. Integration with orchestration frameworks (ONAP, OSM, Nephio, Sylva) achieves closed-loop automation for RAN and Core (Kaltenberger et al., 2024).

4. Development Methodology, Interoperability, and Testing

OAI’s development workflow is governed by feature/integration/develop branch hierarchy with date-coded release tags. Automated CI pipelines in GitLab orchestrate compilation, unit tests, static analysis, and generation of container images for deployment on Docker Hub.

Testing is performed via:

• Lab testbenches (EURECOM 5G Lab): USRP SDRs, O-RAN split RUs, COTS UEs, RF simulators for 4G/5G NSA/SA system tests
• Large-scale external testbeds (Colosseum, X5G, POWDER, ARA Lab, AERPAW, COSMOS): over-the-air and channel emulator platforms, multi-vendor RAN/CN, spectrum, and mobility experiments

For validation of advanced features, such as 5G positioning with UL-TDoA, setups use both OAI-RF-simulator-based platforms and hardware-based O-RAN localization testbeds, demonstrating mean positioning errors from ~0.35 m to ~5.4 m in real-world deployments (Malik et al., 2024).

5. Standards Compliance and Integration with Research Initiatives

OAI is closely aligned with 3GPP Releases 15–18, covering LTE, 5G NSA/SA, Non-Terrestrial Networks (NTN), network slicing, massive MIMO, sidelink, and integrated support for O-RAN interfaces (E2, O1, A1). The OAI RAN stack implements all 5G NR sidelink PHY blocks and supports configuration for various numerologies, MIMO modes, and slot structures, in full compliance with 3GPP standards (Elkadi et al., 2023).

OAI’s codebase enables reproducibility, transparency, and compatibility with O-RAN Software Community (OSC), ONF, LF-Networking, and ETSI NFV platforms.

6. Performance Metrics, Research Directions, and 6G Roadmap

OSA supports deep academic-industry engagement for system-level metrics and experimental research:

• Spectral efficiency: $\eta = R/B$ (bits/s/Hz)
• Throughput: $C = B\log_2(1+\mathrm{SINR})$ (SISO), $$C = \sum_{i=1}^{N_{\rm streams}} B\log_2(1+\lambda_i/\sigma^2)$$ (MIMO)
• Latency: $$T_{\rm proc} = T_{\rm rx} + T_{\rm fft} + T_{\rm demod} + T_{\rm decode}$$; RTT includes all air, core, and processing steps

Current use cases include private 5G Open RAN testbeds, network slicing with scheduler xApps, NTN (GEO/LEO) end-to-end demos, and UL-TDoA positioning validation (Kaltenberger et al., 2024, Malik et al., 2024).

The 6G R&D roadmap encompasses:

• Dynamic spectrum access in FR3 (7.125–24.25 GHz), sub-THz prototypes (>60 GHz/130 GHz)
• Reconfigurable Intelligent Surfaces (6G-BRICKS RIS agent in gNB)
• AI/ML integration for air-interface, RAN, and core-level optimization, e.g., CSI compression, ML-based beam management, NWDAF inference for capacity prediction
• Massive MIMO (beyond 32×32), tactile internet, TSN with sub-millisecond latency (<0.1 ms)

7. Collaborative Model and Impact on Open Cellular Systems Research

OSA’s governance and code contribution pipeline—open CLA, public GitLab, documented CI, hackathons, community tutorials—lower both legal and technical barriers to participation. This open-source approach has enabled reproducible experimentation in advanced areas, from 5G NR sidelink V2X (Elkadi et al., 2023) to multi-user MIMO testbeds (Uçak et al., 14 Jan 2026) and real-time 5G positioning (Malik et al., 2024). The alliance has established OAI as the reference for RAN/CN experimental research and 6G proof-of-concept development in spectrum, MIMO, RIS, and AI/ML, thereby accelerating the pace and rigor of wireless innovation worldwide (Kaltenberger et al., 2024).