RAN and Core ran-frameworks
This page lists the main RAN, edge, and core ran-frameworks
The following frameworks enrich the capabilities of the radio access or core network.
The O-RAN Alliance is an industry consortium that promotes the definition of an open standard for the vRAN, with two goals. The first is the integration of machine learning and artificial intelligence techniques in the RAN, thanks to intelligent controllers deployed at the edge. The second is the definition of an agile and open architecture, enabled by well-defined interfaces between the different elements of the RAN. Since all O-RAN components must expose the same APIs, it is easy to substitute components with others offering alternative implementations of the same functionalities. This allows O-RAN-based5G deployments to integrate elements from multiple vendors, thus opening the RAN market to third-party entities providing new functionalities and diversified services. Moreover, it makes it possible to adopt COTS hardware, in an effort to promote flexibility and reduce costs. Eventually, following the trend started with cloud-native infrastructures, the O-RAN Alliance also aims at promoting open source software as part of the consortium effort.
COMAC is a platform that targets the integration of multiple access and CN technologies, including 4G and 5G cellular networks, broadband, fiber and cable networks, and Wi-Fi deployments. The framework provides a common data plane in the core, which aggregates user data to and from different access technologies, and the possibility of managing users’ subscriptions and identities with a single management platform. COMAC is based on the SEBA platform (a lightweight multi-access technology platform, which provides high-speed links from the edge of the network to the backbone of the infrastructure), and on multiple ONF Component Projects, such as OMEC, for the mobile core and edge, and CORD for the broadband subscriber management. Moreover, it will exploit O-RAN (with the SD-RAN implementation) for the control plane of the mobile cellular access.
SD-RAN is an open source RAN framework that follows the O-RAN development by building and trialing
O-RAN compliant open source components. In particular, SD-RAN is developing a cloud-native near-real-time RIC
(nRT-RIC) and a set of exemplar xApps to control the RAN. This effort has a tight interplay with the other
ONF frameworks, including COMAC, ONOS, and Aether.
Aether streamlines the deployment of private enterprise cellular networks. It combines three main elements, namely,
a control and orchestration interface to the RAN, an edge cloud platform (the Aether edge), with support
for cloud computing APIs, and a central cloud (the Aether core), for orchestration and management.
The Aether project will build and integrate several ONF efforts, including SD-RAN, ONOS, CORD and OMEC.
At the time of this writing, the source code and the deployment pipeline are not publicly available.
When the code will be released, besides providing an opportunity for private 5G networks,
Aether could be effectively used to deploy and manage integrated RAN-edge testbeds for 5G research
Magma is a framework developed by the Facebook Connectivity initiative for simplifying the
deployment of cellular networks in rural markets. Notably, its goal is to avoid dependence on a
specific access technology (i.e., cellular or Wi-Fi) or on a generation of 3GPP core networks.
Moreover, it avoids vendor lock-in for telecom operators, while offering advanced automation and
federation capabilities. The latter is particularly relevant in rural and under-developed scenarios,
as it allows the pooling of resources from multiple network operators. Magma is composed by three components,
i.e., an access gateway, to interface RAN and core networks; a cloud-based orchestrator, for control; and
a federation gateway, which can be used to bind together multiple core networks.
Akraino Radio Edge Cloud (REC)
Akraino Radio Edge Cloud (REC) is a blueprint to support and meet the requirements of the O-RAN RIC. It is part of the Telco Appliance blueprint family. Its features include automated configuration and integration testing to facilitate the management and orchestration of the virtualized RAN. The blueprint is made up of modular building blocks and provides an abstraction of the underlying hardware infrastructure, allowing O-RAN RIC to run on top of it, and to seamlessly interface with the provided APIs.
NVIDIA Aerial is a set of Software Development Kits (SDKs) that allows to build Graphics Processing
Unit (GPU)-accelerated software-defined, cloud-native applications for the 5G vRAN. At the time of
this writing, Aerial provides two main SDKs: cuBB and cuVNF.
The following frameworks provide support for network slicing.
5G-EmPOWER is an operating system for heterogeneous RAN architectures. It consists of an open
source and reprogrammable software platform abstracting the physical RAN infrastructure and providing
high-level APIs to control RAN functionalities.
5G-EmPOWER currently supports several mobile Radio Access Technologies (RATs) such as LTE via srsLTE,
Wi-Fi, and LoRa. The 5G NR is not supported yet. Integration of diverse RATs is obtained through
agents embedding specialized wrappers, one for each RAT. While the general architecture of the
agent is RAT-independent, the wrapper is RAT-specific. For instance, new RATs (e.g., 5G NR) can
be integrated by implementing new wrappers.
FlexRAN leverages abstraction and softwarization technologies to develop a RAT-independent RAN
management platform. FlexRAN embraces SDN principles to decouple control and data planes.
The control plane is orchestrated by a real-time centralized controller, which controls a
set of agents, one for each network element. FlexRAN implements a set of REST APIs in JSON
format describing the northbound interface of FlexRAN. These APIs are used by the agents to
interface with base stations, thus enabling control of the protocol stack and functionalities
of the base stations (i.e., MAC, RRC, PDCP).
The following frameworks enable deployment of edge computing solutions.
The CORD framework is based on multiple software solutions that, together with reference hardware design, realize a reference MEC architecture based on SDN, NFV and cloud-native solutions. CORD aims at (i) reducing deployment costs by using commodity hardware, and (ii) enabling innovative services, thanks to well-defined APIs for accessing edge computing facilities and multi-domain security. Moreover, CORD can be easily extended to address the heterogeneous requirements of different markets. In particular, two CORD architectures specific for mobile and residential services have been spawned off into two Exemplar Platforms (SDN-Enabled Broadband Access (SEBA) and COMAC). CORD is one of the ONF projects with the largest number of contributions by the open source community. It includes detailed installation, operation and development guides, and a set of repositories with its source code.
LL-MEC is an open source MEC framework for cellular systems compliant with 3GPP and ETSI specifications.
This framework merges SDN, edge computing and abstraction principles to provide an end-to-end platform
where services requested by mobile users are executed on edge nodes of the network. LL-MEC consists of
two main components: The Edge Packet Service controlling core network elements (e.g., routers and gateways)
via OpenFlow APIs; and the Radio Network Information Service interfacing the data plane
and physical RAN elements (e.g., eNBs) via the FlexRAN protocol. Aside from MEC capabilities,
LL-MEC supports network slicing for differentiated services applications with diverse latency and
LightEdge is a MEC platform for 4G and 5G applications compliant with ETSI MEC specifications.
LightEdge allows network operators to provide MEC services to mobile users through cloud-based applications.
The framework provides a Service Registry summarizing services and applications registered to
the MEC platform. LightEdge also includes modules and libraries for real-time information exchange
across applications and services, and to perform traffic steering to and from the cellular network.
LightEdge supports multiple eNBs and is compatible with several open source projects such as
srsLTE, Open5GS, and srsEPC.