PaperlayerView exact revision on arXiv

Source-linked AI summary

Open, Programmable, and Virtualized 5G Networks: State-of-the-Art and the Road Ahead

Leonardo Bonati, Michele Polese, Salvatore D'Oro, Stefano Basagni, Tommaso Melodia

arXiv:2005.10027v5cs.NI

TL;DR

Heterogeneous 5G use cases require flexibility beyond monolithic black-box cellular networks, while the growing open-source ecosystem remains difficult to assess cohesively. This paper surveys and taxonomizes open, programmable, and virtualized 5G software, hardware, frameworks, and testbeds, detailing their interactions and usage. It concludes that these technologies are not yet ready for prime-time commercial 5G deployment and identifies standards support and other showstoppers as remaining challenges.

  • Problem

    Heterogeneous 5G requirements expose the limitations of monolithic black-box networks, while the expanding open-source ecosystem has unclear functionality, scope, and interoperability.

  • Method

    The paper provides an end-to-end survey and taxonomy of open-source 5G software, hardware, frameworks, and testbeds, analyzing capabilities, interactions, and usage combinations.

  • Results

    The survey presents a cohesive account of open, programmable, and virtualized 5G solutions and explains how heterogeneous components fit together across the ecosystem.

  • Takeaways & Limitations

    The reviewed ecosystem supports a transition from black-box cellular management toward white-box principles and broader software-based innovation.

  • Takeaways & Limitations

    The surveyed software-defined technologies are not yet ready for prime-time commercial 5G networks, with standards support remaining a major challenge.

Abstract

from arXiv · show

Fifth generation (5G) cellular networks will serve a wide variety of heterogeneous use cases, including mobile broadband users, ultra-low latency services and massively dense connectivity scenarios. The resulting diverse communication requirements will demand networking with unprecedented flexibility, not currently provided by the monolithic black-box approach of 4G cellular networks. The research community and an increasing number of standardization bodies and industry coalitions have recognized softwarization, virtualization, and disaggregation of networking functionalities as the key enablers of the needed shift to flexibility. Particularly, software-defined cellular networks are heralded as the prime technology to satisfy the new application-driven traffic requirements and to support the highly time-varying topology and interference dynamics, because of their openness through well-defined interfaces, and programmability, for swift and responsive network optimization. Leading the technological innovation in this direction, several 5G software-based projects and alliances have embraced the open source approach, making new libraries and frameworks available to the wireless community. This race to open source softwarization, however, has led to a deluge of solutions whose interoperability and interactions are often unclear. This article provides the first cohesive and exhaustive compendium of recent open source software and frameworks for 5G cellular networks, with a full stack and end-to-end perspective. We detail their capabilities and functionalities focusing on how their constituting elements fit the 5G ecosystem, and unravel the interactions among the surveyed solutions. Finally, we review hardware and testbeds on which these frameworks can run, and discuss the limitations of the state-of-the-art, as well as feasible directions toward fully open source, programmable 5G networks.

I. INTRODUCTION

5G’s heterogeneous use cases expose the inflexibility of monolithic 4G-style cellular infrastructure, motivating open, programmable, softwarized, and virtualized designs. This survey organizes the resulting open-source ecosystem across the network stack and explains how its components interact.

  • Motivation: 5G applications impose heterogeneous requirements that traditional monolithic, black-box cellular infrastructure cannot flexibly meet.The paper highlights mobile broadband, ultra-low latency, and dense-connectivity scenarios as examples of this diversity.
  • Motivation: SDN, virtualization, MEC, network slicing, and cloud RAN enable more agile network deployment, control, management, and resource sharing.These paradigms support dynamic operations such as frequency planning, user scheduling, and mobility management.
  • Open-source ecosystem: Open-source and programmable solutions make cellular networking more accessible while enabling software-based instantiation and reconfiguration on SDRs and white-box servers.Examples include GNU Radio, OpenAirInterface, srsLTE, and O-RAN.
  • Open-source ecosystem: The proliferation of heterogeneous open-source components makes their functionality, scope, and interoperability difficult to assess.The survey addresses this by organizing solutions into ecosystem building blocks and describing their interactions.
  • Survey scope: The survey provides an end-to-end taxonomy, architectural commentary, interaction analysis, and usage guidance for combining open-source 5G components.It also distinguishes its coverage from previous surveys through extensive treatment of softwarized architectures and framework interactions.

B. Enabling technologies for softwarized 5G cellular networks

Softwarization separates network control from forwarding, while NFV virtualizes network functions on general-purpose infrastructure. Together, these technologies support flexible management and orchestration of heterogeneous 5G networks.

  • Overview: 5G management must integrate heterogeneous technologies and orchestrate network services across a complex infrastructure.The paper frames this as a management and control problem addressed through cloud-computing-inspired architectures.
  • Software-defined Networking: SDN decouples network control from the forwarding plane, separating routing and control processes from data handling.In 5G, this principle extends to detaching RAN and edge hardware from networking and service capabilities.
  • Software-defined Networking: 5G softwarization enables RAN disaggregation, with radio units acting as transceivers while control and processing run through software, open interfaces, and APIs.The paper identifies open-source SDN solutions such as ONOS and related frameworks as examples of this approach.
  • Network Function Virtualization: NFV virtualizes network services as software-based VNFs decoupled from hardware and executed on VMs running on general-purpose infrastructure.Atomic VNFs can be combined to create more complex and customized network services.
  • Network Function Virtualization: NFV architecture comprises an orchestrator, VNFs, and NFVI, while OPNFV promotes common infrastructure adoption through testing, compliance, and verification tools.The orchestrator manages VMs and the services they run on physical infrastructure.

C. RAN and Core Network Slicing

Network slicing virtualizes shared infrastructure into isolated, tenant-controlled networks spanning RAN and core functions. The section also connects edge computing and machine learning with the resource, latency, and automation demands of softwarized 5G.

  • Network slicing: Network slicing abstracts network functionality from hardware and software, allocating shared physical resources to independent tenant slices.Tenants control their slices but remain isolated from other slices through slice isolation or orthogonality.
  • Network slicing: Slices can span RAN and core functions and target specific services, users, applications, or traffic classes.Examples include private cellular networking and resource-intensive virtual or augmented reality applications.
  • Network slicing: Software-controlled slicing supports real-time, on-demand instantiation, reconfiguration, and revocation to adapt resources to changing demand and service agreements.Slices may receive different resources and security properties for service differentiation.
  • Network slicing: Dynamic slicing requires orchestration, SLA compliance, and resilience mechanisms supported by OSS/BSS and closed-loop control.The paper identifies these operational systems as important to the success of network slicing.
  • Edge computing: MEC moves content and functionality toward the network edge to reduce latency and offload the core, while supporting localized services.Applications include private cellular networking, IoT monitoring, augmented reality, telesurgery, and industrial use cases.
  • Intelligence in the Network: Machine learning supports automated, data-driven optimization for traffic forecasting, core-resource scaling, HARQ prediction, and beam adaptation.Its integration into real deployments requires collecting and aggregating real-time telemetry and data.

III. THE RADIO ACCESS NETWORK

Open-source RAN frameworks provide software implementations and interfaces for deploying, simulating, and optimizing cellular networks. OpenAirInterface supports broad LTE functionality and experimental NR development, while CellOS demonstrates software-driven scheduling optimization.

  • OpenAirInterface: OpenAirInterface provides software-based LTE implementations for base stations, UEs, and the EPC, plus a layer-2/layer-3 simulation environment.NR-compliant base-station, UE, and core-network applications were under development.
  • OpenAirInterface: OAI-RAN supports FDD and TDD operation, multiple bandwidths and transmission modes, standard LTE control channels, and several MAC scheduling algorithms.Its eNB stack includes proportional-fairness, greedy, and fair round-robin scheduling.
  • OpenAirInterface: OAI-RAN can connect to commercial or open-source EPCs and SDR platforms, although X-series USRP execution was reported as less than fully stable.Both commercial smartphones and SDRs can serve as UEs, but development emphasis on the eNB may cause connectivity issues.
  • OpenAirInterface: CellOS uses RIC functionalities with enriched OAI eNBs to optimize user scheduling for three eNBs serving nine commercial smartphones.The experiment compares CellOS optimization with OAI-RAN proportional-fairness and greedy scheduling.
  • OpenAirInterface: CellOS significantly increases network throughput and reduces convergence time to stable high throughput relative to the other schedulers.The experiment illustrates software access to scheduling parameters and protocols for swift automatic reconfiguration.

B. srsLTE

srsLTE is an open-source LTE software suite spanning eNB, UE, and EPC functions, with Linux compatibility, SDR support, simulation utilities, and selected NR-layer development. Its programmable components support experimentation in slicing, security, and edge networking.

  • srsLTE: srsLTE provides LTE Release 10 software implementations of the eNB, UE, and EPC, with selected features from higher releases including NR Release 15.The suite is written in C and C++ and distributed under the GNU AGPLv3 license.
  • srsLTE: srsLTE supports multiple FDD bandwidths from 1.4 to 20MHz, several transmission modes, and physical-layer modulations up to 256 QAM.Compatible RF front-ends include USRP B- and X-series boards, limeSDR, and bladeRF.
  • srsLTE: The srsLTE eNB implements the E-UTRAN protocol stack, interfaces with the core through S1AP and GTP, and supports IPv4 connectivity for COTS and SDR UEs.The UE additionally includes NAS and gateway functionality supporting IPv4 and IPv6 connectivity.
  • srsLTE: Its tools can simulate uncorrelated fading, propagation delays, and radio-link failures, while initial NR support was being developed across MAC, RLC, RRC, and PDCP layers.The suite also supports soft and hard USIM cards with XOR and Milenage authentication.
  • Use cases: srsLTE has been used for LTE security analysis, device fingerprinting, control-plane fuzzing, layer-2 analysis, and MEC slicing on heterogeneous edge devices.CellOS and Sl-EDGE are used to compute and apply differentiated resource allocation for premium and regular slices.

A. Evolved Packet Core

Open-source EPC software implements core-network functions across several projects, with differing interfaces, operating-system support, features, and deployment architectures. The surveyed implementations include OAI-CN, srsEPC, Open5GS, and OMEC, alongside rural-connectivity and emerging 5G-core use cases.

  • EPC architecture: LTE EPC implementations provide mobility-management, subscriber-database, serving-gateway, and packet-gateway functions for control, authentication, tunneling, routing, and address allocation.The surveyed implementations expose different subsets of relevant 3GPP interfaces summarized in Table III.
  • EPC implementations: OAI-CN is an Apache-licensed C/C++ EPC for Intel x86 and Ubuntu that requires kernel modifications for real-time capabilities and supports dynamic policy-based QoS.Its QoS functions include establishing, modifying, and removing multiple dedicated bearers.
  • EPC implementations: srsEPC is a C++ GNU AGPLv3 implementation supporting configurable subscriber parameters, XOR and Milenage authentication, per-user QCI, and dynamic or static IP configuration.It is compatible with Ubuntu and Fedora Linux.
  • EPC implementations: Open5GS is an AGPLv3 EPC supporting several Linux distributions, FreeBSD, and macOS, with VoLTE and SG-SMS services for voice calls and text messages.These services extend LTE connectivity beyond traditional data transport.
  • EPC implementations: OMEC is an Apache-licensed, NFV-based LTE Release 13 EPC designed for scalability and usable standalone or within larger frameworks such as COMAC.It offers connectivity, billing, and charging features and was developed with operators and industry partners.
  • Use cases: Open-source EPCs support applied deployments including rural connectivity through CoLTE and ongoing development of 5G NR core functions from OAI-CN.CoLTE modified OAI-CN for user billing and accounting and was being ported to Open5GS.

B. 5G Core

The surveyed 5G core and RAN frameworks span open-source implementations, standardized interfaces, intelligent control, virtualization, and multiple deployment options. O-RAN emphasizes open interfaces, interoperable components, and programmable RAN control, but remained below production level at the time described.

  • 5G Core: free5GC implements 5G core access, mobility, session management, and network-function discovery under the Apache License v2.0.It is written in Go and supports Ubuntu Linux.
  • 5G Core: Its 3GPP interfaces connect the AMF, SMF, UPF, UDM, AUSF, RAN, and data network for mobility, sessions, user-plane functions, authorization, subscriptions, and authentication.The listed interfaces include N1/N2, N3/N4/N6, N8, N10/N11, and N12/N13.
  • O-RAN: O-RAN defines non-real-time and near-real-time RICs, RAN components, and open interfaces for policy control, machine-learning management, and near-real-time actions.The A1 interface connects the two RICs, while O1 and E2 connect controllers with RAN elements.
  • Deployment and software: O-RAN supports regional-cloud, edge, cell-site, O-Cloud, and proprietary deployment options, with Scenario B preferred: RIC in the regional cloud, CU/DU at the edge, and RU at cell sites.The O-RAN Software Community contributes containers, component source code, virtualization infrastructure, and interface simulators.

B. Open Networking Foundation Frameworks

ONF organizes open-source networking efforts into modular Component Projects and deployable Exemplar Platforms. Its mobile-network portfolio combines SDN, programmable hardware, edge-cloud, core-network, and access technologies.

  • ONF structure: The ONF is a consortium of telecom operators contributing open-source code and frameworks for network deployment.Its projects use modular designs that facilitate component integration and incorporation of new open-source projects.
  • ONF structure: ONF distinguishes Component Projects serving specific purposes from Exemplar Platforms combining projects into deployable proof-of-concept designs.Table V summarizes dependencies across these projects and platforms.
  • Component Projects: The ONF oversees 10 open-source component projects spanning SDN, transport networks, programmable networking hardware, and mobile networks.Examples include OMEC, SD-RAN, CORD, and ONOS.
  • Component Projects: ONOS provides a common substratum for SD-RAN and Exemplar Platforms including Aether and COMAC.ONOS has primarily been used for wired SDN deployments.
  • Exemplar Platforms: COMAC extends CORD to integrate multiple access and core technologies, using a common core data plane and unified subscription and identity management.Its components include SEBA, OMEC, CORD, and SD-RAN-based O-RAN control.
  • Exemplar Platforms: Aether targets private enterprise cellular networks by combining RAN control and orchestration, an edge cloud, and a central cloud.It integrates ONF efforts including SD-RAN, ONOS, CORD, and OMEC.

C. Other Frameworks and Projects

Other projects address heterogeneous RAN control, cellular deployment, MEC, programmable RAN hardware, and GPU-accelerated vRAN processing. Several frameworks provide slicing or orchestration capabilities, while some lack 5G NR support.

  • RAN platforms: 5G-EmPOWER abstracts heterogeneous RAN infrastructure through an open, reprogrammable platform with high-level control APIs.It separates control and data planes using a centralized controller and network-element agents.
  • RAN platforms: 5G-EmPOWER supports LTE, Wi-Fi, and LoRa but not 5G NR, while its wrapper architecture allows additional RATs to be integrated.It already supports customized isolated RAN slices specified by SLAs and user assignments.
  • RAN platforms: FlexRAN provides RAT-independent RAN management through a centralized real-time controller, agents, and JSON REST APIs for slicing policies and requirements.It interfaces directly with OAI and therefore does not yet support 5G NR communications.
  • Deployment frameworks: Magma simplifies cellular-network deployment in rural markets while avoiding dependence on a specific access technology, 3GPP core generation, or vendor.Its access gateway currently supports an LTE EPC and has been tested with commercial LTE base stations.
  • MEC frameworks: LL-MEC merges SDN, edge computing, and abstraction into a 3GPP- and ETSI-compliant end-to-end platform for executing mobile services on edge nodes.LightEdge provides ETSI MEC services, a service registry, inter-application information exchange, and traffic steering.
  • Programmable hardware: NVIDIA Aerial supplies GPU-accelerated vRAN SDKs: cuBB runs PHY processing on GPUs, while cuVNF bypasses the CPU for direct NIC-to-GPU packet placement.The cuVNF SDK also supports additional VNFs such as deep packet inspection and firewalls.

VI. OPEN VIRTUALIZATION AND MANAGEMENT FRAMEWORKS FOR NETWORKING

Open virtualization and management frameworks decouple network services from hardware and coordinate their deployment across virtualized infrastructure. The survey covers NFV architectures, virtualization mechanisms, VIMs, and MANO frameworks, while focusing primarily on common NFV approaches rather than serverless computing.

  • MANO: An NFV orchestrator combines virtualization-infrastructure management, MANO functions, and the VNFs it manages, using northbound and southbound APIs.VIMs such as OpenStack, Kubernetes, and Docker connect orchestration with physical resources.
  • MANO: The survey compares ONAP, OSM, and Open Baton as MANO frameworks for provisioning network services.ETSI defines common features that NFV MANO frameworks should provide, primarily for network-function orchestration.
  • NFV foundations: NFV decouples network services from hardware by packaging applications as hardware-independent virtual machines, enabling scalability across physical machines.The architecture improves scalability and portability by separating VNFs from general-purpose infrastructure.
  • Virtualization techniques: The main NFV virtualization approaches are traditional VMs, bare-metal hypervisors, containers, and unikernels.The paper focuses primarily on these common approaches and excludes serverless computing from its main treatment.
  • Virtualization techniques: Containers package application code and dependencies while sharing the host operating system and kernel, requiring less virtualization than VMs.Docker modularizes applications and dependencies into shareable layers inside containers.
  • Infrastructure management: A VIM controls NFV infrastructure resources and coordinates virtual-machine instantiation on physical hosts.Hypervisors provide isolation, resource allocation, and scheduling for guest machines.

B. The Open Network Automation Platform

ONAP is an open NFV framework for designing, orchestrating, monitoring, and optimizing physical and virtual network services. Its 5G evolution adds awareness of heterogeneous functions, slicing, closed-loop control, and O-RAN-oriented interfaces.

  • ONAP is a Linux Foundation NFV framework supported by major mobile operators and integrated by leading network vendors.It is deployed in several commercial cellular networks.
  • ONAP orchestrates physical and virtual infrastructure, manages service lifecycles, and supports vendor-agnostic network-service operations.Its design framework models resources, services, products, and management functions.
  • The architecture comprises management, design, and run-time frameworks supported by Kubernetes and Consul-based component management.The run-time framework uses a microservices bus, automated control loops, data collection, analytics, APIs, and dashboards.
  • Integration with 5G networks: The Dublin release introduced joint discovery of virtual and physical network functions and preliminary network-slicing support.The Frankfurt release added end-to-end slicing, service orchestration, user-defined control loops, and harmonization toward O-RAN compatibility through O1 and A1 interfaces.
  • Integration with 5G networks: ONAP’s 5G blueprint supports dynamic parameter optimization through KPI collection, analysis, and automatic optimization or resource scaling.The data collection platform transfers relevant KPIs from the network edge to central processing facilities in a matter of minutes.

C. Open Source NFV Management and Orchestration

This section surveys OSM and related open orchestration components, together with compatible SDR hardware. OSM provides end-to-end MANO capabilities for network functions, services, and slices, while the hardware supports experimental deployments across varied scales.

  • OSM: OSM is an open-source ETSI-aligned MANO framework developed by network providers and tailored for end-to-end mobile-network service orchestration.Its architecture interacts with managed network functions and VIMs through the typical NFV orchestrator structure.
  • OSM: OSM’s Information Model represents network functions, services, and slices as deployable packages describing required resources.The model follows an information model provided by the ETSI MANO framework.
  • OSM: OSM automates service lifecycles from instantiation through scaling and deletion, with service configuration performed through Juju Charms.Northbound and southbound APIs enable integration with modeling components, orchestrators, and OSS/BSS systems.
  • Integration with 5G networks: OSM’s December 2018 5G-ready release added management of virtual and physical network functions, network slicing, policy-based closed control loops, and extended analytics and interface frameworks.European 5G projects used or contributed to OSM for transport, CDN, computing-resource management, and other experiments.
  • Related orchestration frameworks: Open Baton is an open-source, modular NFV orchestrator compliant with ETSI NFV MANO that composes VNFs into network services over virtualized infrastructure.It supports VIM drivers, including OpenStack, and adds network slicing and MEC through its Network Slicing Engine.
  • SDR support: The surveyed software mostly runs on commodity hardware, while SDR physical-layer signal processing generally runs on FPGA hardware.Compatible SDRs range from compact small-cell devices to higher-capacity platforms supporting larger MIMO configurations and networked operation.
  • SDR support: LimeSDR QPCIe supports 4 × 4 MIMO, but OAI and srsLTE support for these LimeSDR models was not explicitly reported.

VIII. TESTBEDS

The surveyed testbeds provide increasingly diverse environments for deploying and evaluating open-source 5G software, from indoor SDR facilities to city-scale and aerial platforms. They support reproducible experimentation across RAN, NFV, orchestration, edge, spectrum, and advanced wireless scenarios.

  • The testbeds support instantiation of softwarized 5G networks using open-source utilities, frameworks, and hardware components.Their capabilities are summarized in Table VIII.
  • PAWR platforms: PAWR platforms target heterogeneous city-scale wireless experimentation with reproducibility, interoperability, programmability, and open research access.The program includes platforms for campus, neighborhood, and aerial experimentation.
  • Colosseum: Colosseum combines large-scale compute, storage, FPGA, SDR, and channel-emulation resources for installing open-source RAN, core, and O-RAN components.Its MCHEM can emulate virtual worlds with up to 65,536 concurrent wireless channels and realistic path loss and fading.
  • Indoor testbeds: Arena provides an indoor office-like environment with 24 USRPs, 12 high-performance servers, and an 8×8 antenna grid for low-latency radio experimentation.It supports evaluation of wireless protocols, communication traces, cognitive radio, and standard-compliant cellular networks through OAI and srsLTE.
  • NFV and orchestration testbeds: 5TONIC offers NFV infrastructure, servers, workstations, and SDR platforms for complex orchestration frameworks such as OSM.It has supported NFV MANO and mmWave experimentation.
  • Federated testbeds: FED4FIRE+ federates wireless, 5G, IoT, cloud, and big-data testbeds compatible with OAI and srsLTE.Its NITOS, IRIS, CORNET, and FIT facilities support MANO, MEC, Cloud-RAN, NFV, SDN, spectrum, and cellular experiments.
  • Indoor testbeds: Drexel Grid provides 24 SDRs and a channel emulator for controlled, repeatable indoor evaluation of 5G and cognitive-radio technologies.Its USRP X310 devices can run open-source RAN software such as OAI and srsLTE.

IX. SOFTWARIZED 5G: LIMITATIONS AND ROAD AHEAD

Open, programmable, and virtualized 5G technologies remain constrained by deployment readiness, standards evolution, RAN participation, scalability, and security. The paper frames these limitations as priorities for advancing the ecosystem beyond current black-box management.

  • Road ahead: The surveyed ecosystem is organized as heterogeneous but modular software and hardware components intended to support a transition from black-box management to white-box principles.The authors associate this transition with faster innovation, shorter time-to-market, and higher customer satisfaction.
  • The survey concludes that software-defined 5G solutions are not yet ready for prime-time commercial deployment and require several showstoppers to be overcome.
  • Standards evolution: Open-source cellular software must keep pace with evolving standards such as 3GPP NR and mmWave technologies.The surveyed RAN libraries had not completed support for the relevant emerging technologies at the time discussed.
  • Latency and scalability: SDR-based systems face latency and scalability pressure because processing and computing requirements generally grow with available bandwidth.Higher future data rates therefore demand highly efficient and robust radio-stack processing chains.
  • RAN contributions: Open-source RAN projects receive less support from major operators and vendors than open-source core-network and MANO projects.Academic and small-company contributions dominate, while intellectual-property concerns discourage broader vendor participation.
  • Deployment readiness: Most surveyed frameworks and libraries remain incomplete, integration-dependent, insufficiently robust, or poorly documented for actual network deployment.The paper points to container-driven development as one way to simplify and expedite deployment.
  • Security: Open-source 5G software requires stronger security-by-design practices because exposed APIs can introduce vulnerabilities.Community scrutiny, audits, reviews, and responsible disclosure can help identify bugs and security holes, but appropriate security remains lacking.

APPENDIX A ACRONYMS

This appendix defines acronyms used throughout the paper, spanning 5G generations and core, radio, networking, orchestration, and testbed technologies. It also includes abbreviations for software frameworks, network functions, protocols, and wireless research platforms.

  • 3GPP, 4G, 5G, and 5GC denote the 3rd Generation Partnership Project, fourth generation, fifth generation, and 5G Core, respectively.
  • The glossary also defines software, networking, and experimentation terms such as API, CORD, ONAP, ONOS, OSM, SDN, SDR, OAI, PAWR, ORBIT, COSMOS, and AERPAW.The listed terms include interfaces, open-source platforms, software-defined networking and radio, orchestration systems, and wireless testbeds.
  • Acronyms cover cellular network components and functions, including AMF, AUSF, CN, CU, DU, RAN, RIC, RU, SMF, and VNF.These abbreviations refer to access, authentication, core-network, radio, control, session-management, and virtualization concepts used in the paper.
Loading 2005.10027v5…