CS Paper Radar

Benchmarking Open-Source O-RAN and 5G NTN Stacks on Embedded ARM Platforms for Space Payloads

Esteban Municio, Oscar Baselga, Oscar Lasierra, Gines Garcia-Aviles, Xavier Costa-Pérez

cs.NI cs.PF ·submitted 2026-11-18 21:30 UTC · v1
benchmarkcode/data

Abstract

Regenerative satellite payloads, which host the 5G Next Generation Node B (gNB) onboard instead of transparently relaying the radio signal, are a key building block of the Non-Terrestrial Network (NTN) integration that 3GPP started in Release 17. Combining them with the Open Radio Access Network (O-RAN) architecture adds software-defined control through RAN Intelligent Controllers (RICs), enabling in-orbit reconfiguration, multi-tenancy, and data-driven radio resource management. Open-source implementations make this vision accessible to research institutions and New Space missions, yet there is little quantitative evidence on how they behave on the embedded processors that a satellite can actually carry. This paper presents an out-of-the-box experimental benchmark of two open-source gNB stacks (i.e., OpenAirInterface/Duranta and srsRAN/OCUDU), and two Near-Real-Time RICs (i.e., the ORAN-SC Near-RT RIC and FlexRIC). All components are deployed on an ARMv8 platform representative of the processing units used in small-satellite payloads, driving a USRP B210 software-defined radio. We compare them in terms of functional compliance (3GPP NTN features, O-RAN interfaces, multi-tenancy), resource efficiency, real-time behavior, and engineering suitability. Results show that these platforms can host a joint gNB + near-RT RIC deployment using any of the considered components, although further optimizations are still required for carrier-grade performance. However, the performance and standards compliance vary significantly among them. Building on these results, we derive a decision framework that maps mission profiles and Size, Weight, and Power (SWaP) budgets to gNB/near-RT RIC combinations. Finally, we discuss the open gaps remaining before these stacks can be flown in NTN regenerative payloads.

Comment
11 Pages, 6 Figures, 6 Tables. Aeroconf

Closest papers

Nearest neighbours by meaning across the whole index.

Open-Source 5G RAN Platforms: A Dual Perspective on Performance and Capabilities

cs.NI ·2026-08-26 ·Maria Katarine Santana Barbosa, Iasmin Gomes, Vinícius Melo, Kelvin Lopes Dias
code/data similarity 0.70
Abstract · This paper has been published in IEEE WCPNS 2025

Open-source implementations of the fifth generation (5G) Radio Access Network (RAN), such as OpenAirInterface (OAI) and srsRAN, have become increasingly relevant for industry and academic research, rapid prototyping, and low-cost private 5G deployments. The state of the art presents different evaluation scenarios for both platforms; however, these are often limited to individual testbed evaluations, end-to-end studies, or partial comparisons between them. In particular, there is a lack of evaluations considering different Software-Defined Radios (SDRs), RAN configurations, user applications, and numbers of real UEs. To address this gap, this paper presents a comparative evaluation of OAI and srsRAN. The study includes a qualitative assessment of supported features and deployment flexibility, followed by a quantitative performance analysis covering radio resource control setup procedures, compliance with theoretical data rates, and real application workloads with diverse quality of service requirements. Specifically, we consider Video on Demand (VoD) for data rate evaluation, Live Streaming (LS) and Cloud Gaming (CG) as latency-sensitive applications. These workloads provide a comprehensive view of each platform capability.

0.63

Toward Lightweight Aerial 5G gNBs: Reproducible OAI Testbed on Commodity ARM Platforms

cs.CR eess.SP ·2026-10-04 ·Marc Duboc, Ammar El Falou
similarity 0.68
Abstract

For a battery-powered UAV base station, every gram and watt devoted to RAN-compute competes with flight-duration margin. We therefore keep the 5G Core (5GC) and central unit (CU) on the ground and design the airborne unit around only the backhaul endpoint, distributed unit (DU), and radio. The key question is then a practical one: can widely available commodity ARM computers sustain a real radio OpenAirInterface (OAI) 5G DU with useful performance over heterogeneous F1? We answer it with Raspberry Pi~5 and Jetson Orin Nano as DUs, a USRP B210, a commercial handset, and Ethernet, Wi-Fi/GRE, and 5G/WireGuard backhaul. Jetson preserves 88.4% of x86 split-DU Ethernet DL throughput and 87.5-89.0% across all three bearers; 5G/WireGuard preserves 76.4-77.1% of each host's wired DL rate. The validated Jetson/B210/RM500Q-GL electronics weigh 657.4 g and draw about 28 W under sustained traffic (757.4 g with integration allowance). A controlled link adaptation intervention raises split DL from 23.4 to 99.4 Mb/s as the dominant Modulation and Coding Scheme (MCS) moves from 3 to 26. Finally, synchronized radio, F1-U, CPU, and UHD evidence narrows the remaining monolithic-split gap to split-path scheduling/timing behavior. Beyond performance, this capability serves the emergency-response use case that motivates the aerial cell: once deployed above an affected area, it can broadcast a PWS warning message. We release the OAI patch that carries the single-segment Write-Replace Warning over F1 from CU to DU, where SIB8 is scheduled to the handset. The public artifact makes the payload, performance, and emergency-broadcast baseline reproducible with laboratory-accessible hardware.

0.14

Virtualized 5G Tesbed using OpenAirInterface: Tutorial and Benchmarking Tests

cs.NI ·2026-09-11 ·M. Dória, V. Sousa, A. Campos, N. Oliveira, P. Eduardo +7 more
publishedbenchmarkhighly cited similarity 0.67
Abstract

The development of 5G and its evolutionary path to 6G brings virtualization as close as possible to the antennas. Native 3GPP systems are now software running at servers boosted by accelerator cards to cope with computationally intense signal processing. Meanwhile, the Radio Frequency (RF) front-end is still proprietary hardware, sheltering specific PHY-layer procedures like passband amplification/modulation. This approach takes advantage of the software's flexibility while keeping the complex microsecond signal processing performance from modern telecommunication systems. This paper provides tutorial material on the Core Network and Radio Access Network of OpenAirInterface (OAI) 5G stack on top of Universal Software Radio Peripheral (USRP) platforms. A set of blueprints showcases OAI's ability to provide a virtualized 5G network with different transmission capabilities and the possibility to use it with commercial mobile phones. Configuration discussions and throughput benchmark analyses follow installation and deployment instructions. Our results show that 5G prototyping using OAI and USRP frontends can lead to good reproducibility and consistent throughput.

0.49

The OCUDU dApp Platform: An Open Runtime and E3 Interface for Real-Time AI-RAN

cs.NI cs.LGeess.SP ·2026-09-07 ·Timothy O'Shea, Matthew Pennybacker, Andriy Kharchenko
code/data similarity 0.60
Abstract · 17 pages, 13 figures, 9 tables. Preview accompanying the OCUDU dApp platform preview release (OCUDU AI-RAN WG2): https://gitlab.com/ocudu/wo…

Machine learning has shown its largest gains in the band below 10 ms inside a 3GPP new radio (NR) 5G distributed unit (DU): link adaptation, per-slot scheduling, channel estimation, and the receiver itself. No open platform has let independently built software run there. Prior dApp frameworks reached the band only as external observers of an export stream. This paper is a guided introduction to the OCUDU dApp platform, an open runtime and E3 interface under which signed AI-RAN applications execute inside a production DU under three timing contracts: resident on the GPU receive chain (Class A), inside the scheduler's 100 us admitted deadline (Class B), or as never-blocking observers whose results the scheduler consumes (Class C). The conventional path is never displaced, and every authority is typed, validated, and operator-bounded. The paper explains how the runtime, the embedded E3 agent, and the three public repositories fit together; shows a dApp's source, its signed package, and its lifecycle state machine; defines the contracts a module is written against; and shows how one management surface serves a Python script, an operator's console, and an LLM agent. On a GB10 gNB with attached handsets, dApps of all three classes, including an out-of-tree neural equalizer, ran together on a live cell without a single fallback, and equalizer variants were compared over the air by lifecycle operations alone. Every measured checkpoint is reported with its conditions and its gaps. Platform, SDK, and a zero-hardware quickstart are public under BSD-3-Clause-Clear as a preview release of the OCUDU AI-RAN Working Group 2, inviting feedback, new use cases, and independent vetting ahead of upstreaming into the OCUDU mainline.

0.45

PRO-RAN: Processor-Level Characterization of Open RAN Centralized and Distributed Units

cs.NI cs.PF ·2026-08-27 ·Moojan Kamalzadeh, Larry Horner, Linqi Xiao, Abhishek Bhattacharyya, Ehsan Bahaloo Horeh +3 more
hot direction similarity 0.60
Abstract · 7 pages, 6 figures

Open Radio Access Network (O-RAN) disaggregates RAN protocol functions and enables Centralized Unit (CU) and Distributed Unit (DU) software to execute on general-purpose computing platforms. Different CU and DU protocol responsibilities produce different processor workloads and execution paths. Conventional performance metrics, including CPU utilization and throughput, quantify aggregate resource usage without identifying function-level execution costs or processor microarchitectural bottlenecks. Processor-level characterization, on the other hand, provides insights into resource provisioning, function placement, software optimization, and hardware acceleration. The paper describes a controlled characterization framework that evaluates independently deployed CU and DU functions under matched hardware and traffic conditions. The experimental platform integrates the Linux Foundation OCUDU implementation with an emulated User Equipment, a ZeroMQ-based radio interface, and an Open5GS core. Automated validation confirms registration and bidirectional packet delivery before process-scoped Intel VTune Hotspots and Top-Down Microarchitecture Analysis. Under traffic load, accumulated process CPU time increases from 17.3 s to 37.4 s for the CU and from 462.0 s to 628.4 s for the DU during equal 300-s profiling intervals. The measurements identify distinct CU and DU execution characteristics and motivate function-specific processor analysis and optimization.

0.28

Bringing dApps to OCUDU: An E3 Controller for Real-Time Open RAN Intelligence

cs.NI ·2026-09-02 ·Angelo Feraudo, Ravis Shirkhani, Stefano Maxenti, Andrea Lacava, Michele Polese +1 more
code/data similarity 0.59
Abstract

Real-time control loops in Open RAN are increasingly colocated with the gNB, where dApps, i.e., programmable applications with sub-millisecond access to PHY- and MAC-layer signals, enable latency-critical use cases such as spectrum sharing, channel-aware scheduling, and integrated sensing. What turns dApps from a per-vendor mechanism into a portable component of the emerging AI-RAN ecosystem is the E3 interface: this interface defines how a dApp subscribes to RAN telemetry, receives indications, and issues control actions back to the RAN. To date, E3 has been implemented on OpenAirInterface and on NVIDIA Aerial; OCUDU, i.e., the Linux Foundation's open-source CU/DU project, has lacked a comparable E3 path. We close that gap with an open-source E3Controller for OCUDU, designed as a sidecar daemon so the E3 protocol stack, service-model logic, and dApp interactions live entirely outside RAN core processing, which is touched only through well-defined instrumentation hooks. We validate the controller with two reference service models, Spectrum and L1, that deliver O-RAN fronthaul I/Q samples to a spectrum-sensing dApp over a Foxconn RPQN 4800 radio unit. Our evaluation shows that the E3 path imposes no measurable throughput penalty on the OCUDU RAN, demonstrating that a real-time dApp/RAN interface is feasible on production-grade open RAN software.

0.65

AI-Native Open RAN: A Roadmap from xApps and rApps to Autonomous Network Agents

cs.NI ·2026-09-14 ·Ryan Barker, Alireza Ebrahimi Dorcheh, Tolunay Seyfi, Mohammad Raihan Uddin, Alireza Mohammadhosseini +4 more
similarity 0.58
Abstract

Open Radio Access Networks (O-RAN) have emerged as a transformative paradigm for future wireless systems by introducing openness, virtualization, disaggregation, and programmable intelligence through the RAN Intelligent Controller (RIC). The availability of standardized interfaces and near-real-time control loops has created unprecedented opportunities for integrating artificial intelligence (AI) into radio access network management and optimization. Over the past several years, a broad range of AI techniques have been proposed to address key O-RAN challenges such as radio resource management, network slicing, traffic prediction, mobility management, interference mitigation, and spectrum sharing. Despite significant progress, existing solutions often remain task-specific, require extensive retraining, and exhibit limited generalization across deployment environments and network conditions. This paper presents a comprehensive review of AI-enabled O-RAN systems and provides a unifying perspective on the evolution of intelligence in wireless networks. We first examine the O-RAN architecture and the role of intelligence within near-real-time and non-real-time RIC frameworks. We then develop a taxonomy of AI approaches for O-RAN, covering machine learning, deep reinforcement learning (DRL), digital-twin-assisted optimization, and emerging foundation-model-based architectures.

0.04

A Ray-Traced Digital Twin for 5G NR Positioning with OpenAirInterface

cs.IT ·2026-10-09 ·Nima Chaharbaghi, Florian Kaltenberger, Bartosz Podrygajlo, Rakesh Mundlamuri, Roberto Verdone
similarity 0.58
Abstract · submitted at 22nd Wireless On-demand Network systems and Services Conference

Evaluating 5G positioning on a physical testbed ties every measurement to one propagation environment and one antenna layout, so any question about an alternative deployment requires a new field campaign. We present a ray-traced digital twin of the GEO-5G testbed at EURECOM that drives an unmodified OpenAirInterface (OAI) 5G stack with per-antenna channels computed by Sionna-RT over a 3D model of the site. The distributed gNB antennas are modelled as independent radio devices rather than as elements of a compact array, which preserves the per-antenna absolute delays on which time-difference-of-arrival (TDoA) positioning depends. With them preserved, the twin drives OAI's SRS-based UL-TDoA pipeline to metre-level accuracy consistent with that reported for the physical deployment. We then use the twin to answer a question the physical testbed cannot. Widening the carrier truncates the tail of the horizontal error distribution without improving a typical position, and leaves the vertical error untouched, because the latter is limited by the array geometry rather than by the measurement. That limit is predictable from the antenna coordinates alone, and a height-diverse layout reduces the RMS height error by a factor of four.

0.20