The telecoms network is becoming increasingly open. That means the big players, who once would have relied on patents and standards bodies to ensure they had power in new platforms, are now seeking that influence by initiating and controlling open source initiatives.
The year has begun with a long list of open source projects engaging in public relations activity to drum up new members and try to establish themselves as the dominant initiative in a particular area of strategic interest for 5G. AT&T has been particularly adept at pushing open source processes into the heavily defended fortresses of the mobile network vendors. It has contributed the foundations for three significant open platforms and is now busily building support, even among its rivals, in a bid to make these into cornerstones of 5G architectures – possibly at the expense of traditional standards organizations such as ETSI.
The trio are:
- XRAN, which seeks to define a “software-based, extensible RAN and to standardize critical elements of the x-RAN architecture”.
- dNOS (Disaggregated Network Operating System), an OS for telco white boxes.
- ONAP (Open Network Automation Platform), for management and orchestration (MANO) of virtualized telecoms networks.
XRAN recently released its first specifications, working to decouple the RAN control plane from the user plane. It has been working first on three areas which will help to transform the RAN:
- standardized northbound and southbound interfaces to make it easier to deploy services and achieve multivendor networks, in a decoupled, software-defined environment.
- implementation of RAN software and network functions on off-the-shelf hardware,including the modular xRAN LTE eNode B base station software.
- and the flexibility for operators to run software functions at the macro cell site or at the edge of the network, depending on availability of fronthaul links.
The dNOS effort is at an earlier stage – AT&T first announced it, with a white paper, in November, but claims it is already attracting interest and that “we shouldn’t have a problem building a community around it,” as Chris Rice, SVP of AT&T Labs, Domain 2.0 Architecture and Design, told FierceWirelessTech.
Rice said dNOS will provide a unified way to support OS functions on white box switches and routers. “What we would like to be able to do is have that done in a very consistent way so that it easily interfaces and connects to our network,” he said.
While dNOS is a network OS for white boxes, ONAP aims to be the network OS for the cloud, which makes the two initiatives complementary, says AT&T, and it seems probable that dNOS will find its home at the Linux Foundation, as ONAP did when its was formed from a merger of AT&T’s ECOMP project and China Mobile’s Open-O.
ONAP is less than a year old and now has about 50 members, and last week announced that AT&T arch-rival Verizon had joined as a platinum member. It was the eighth operator to join at this level and the sixteenth overall, and means ONAP now supports MNOs serving almost 60% of the world’s mobile subscribers. That, argued Arpit Joshipura, general manager of networking and orchestration for the Linux Foundation, makes it a de facto standard.
“This basically means we now have ONAP as the de facto standard that the operators have embraced globally,” he said. “Verizon is joining to be able to influence where ONAP goes. They recognize this ecosystem is becoming the de facto standard for the industry and they want to have a say, through contributions, in what happens. For some of the smaller operators, it will be more about guiding their decisions.”
Verizon’s VP of technology and supplier strategy, Srinivasa Kalapala, said he hoped ONAP would be ready and robust enough for the telco to deploy it in about a year’s time. That would certainly help to soothe a wave of fears, expressed by some operators last fall, that ONAP has been rushed out and will not be truly carrier-grade for a few more releases.
Verizon contributed code to ONAP’s first release, Amsterdam, before it joined, as part of its trial of the would-be standard platform to decide whether it met the telco’s needs.
Kalapala told LightReading: “We looked at the Amsterdam code and I think by the Beijing release [due in late May], we see full modularity and interoperability. That is when we would start trying to operationalize it.” By the time of the Casablanca release, due in late 2018, he expects Verizon will have some ONAP modules up and running, with commercial deployments in early 2019.
The two specific use cases covered by Amsterdam are a virtualized core to manage VoLTE services, and virtualized residential CPE. Although there were some doubts about how robust this early code would be for production implementations, initial reception was positive overall.
Amsterdam combines the contributed code of ECOMP and Open-O and removes duplication, but also adds significant new features. These include a correlation engine called Holmes which joins the former ECOMP Data Collection, Analytics and Events (DCAE) module; and a new module called Control Loop Automation Management Platform (CLAMP).
Much of the code is already in use at AT&T, so has some production track record, and some has been in use at China Mobile. Bell Canada plans and Orange plan to use the system soon and other operators are evaluating it (including new recruit Vodafone) or running proofs of concept.
But while open source processes encourage fast pace of innovation and lower barriers for a wide variety of players to contribute to de facto standards, the downside is that many different projects can end up battling for attention. The merger of ECOMP and Open-O came as a relief to many operators, which were unsure which MANO platform to back, and the willingness of Verizon to join an AT&T-initiated consortium will add further reassurance that the need for common interfaces and interoperability will outweigh politics.
But there are still other MANO initiatives which may limit ONAP’s power, or at least spread uncertainty or delay deployment decisions among MNOs. The most prominent is ETSI’s Open Source MANO (OSM), to which Telefonica has been a prominent contributor. Last summer, there were reports that the Spanish telco might join ONAP too, when its planning and technology director, Javier Gavilan, said in an interview: “OSM cannot be compared with ONAP because the scope of ONAP is bigger and OSM is only a small part of the Telefonica transformation project. We are transforming our full stack and this is something we are doing and it could be a part of ONAP.”
Verizon has also said that it sees ONAP has being broader, providing a full service management platform, while OSM is mainly an orchestrator for virtual network functions (VNFs).
Telefonica has not yet joined ONAP, but if it does end up belonging to both groups, it could be a catalyst for convergence, which would help address one of the many sources of fragmentation in the immature world of carrier SDN/NFV. Gavilan said last August that he had been in contact with AT&T about aligning ONAP and OSM efforts to develop common information models and processes. He said: “The industry has two different initiatives running in parallel with a lot of common points and the idea is to align them as much as possible.”
The same challenges will face dNOS if it starts to gain traction too during 2018. There are several past or ongoing efforts to create a white box network OS, including those within the Facebook-inspired Telecom Infra Project (TIP), now a major home of activities which could compete with those of the Linux Foundation or ETSI. Others include Microsoft SONiC and another Linux Foundation project, Free Range Routing.
Just because software is open source does not make it easy to implement – indeed, many early movers in NFV/SDN, such as BT in the UK, have pointed out that they have had to invest heavily in skills and tools to make open source MANO or other systems capable of supporting their full requirements and integrating with their network architectures.
This opens up opportunities for software houses and systems integrators to increase their telco business, and even enter the telecoms space for the first time, from enterprise or cloud markets where virtualization and open source have been longer established.
Another of the prominent open source projects in the software-defined telco world, CORD (Central Office Re-architected as a Datacenter), is also touting for more members and claiming it is ready to hit the big time in 2018. Part of the Open Networking Foundation’s ON.Lab initiative alongside ONOS, its aim is to enable the telco central office to function as a data center in a distributed and virtualized environment. The CORD community includes service providers such as AT&T, China Unicom, Google, NTT, SK Telecom and Verizon, as well as vendors Ciena, Cisco, Fujitsu, Intel, NEC, Nokia, Radisys and Samsung. CORD is hosted by The Linux Foundation.
Its challenge is not industry acceptance – it has 60 members and claims a large number of operators are testing implementations – but the shortage of integrators with relevant expertise. Timon Sloane, VP of standards and membership at the ONF, said in an interview: “Operators want to be able to buy this in a supported way. “The market – the systems integrators who are going to step up and say I can deliver this – has not emerged, and that is the challenge for the community as we move forward.”
It is also important that more code comes from the wider ecosystem rather than the ONF’s own engineering staff of 30. “Right now, half of the code, or maybe a little more, comes from ONF engineers,” Sloane said. “Our aspiration is for that to shift over time, what we really need is a full embrace of the ecosystem.”
Some vendors, such as Radisys and Mavenir – both of which have been heavily involved in CORD – are grasping the opportunity to make a business around CORD software and integration, but more are needed.
Sloane still expects CORD to emerge as a de facto standard for telco edge computing, a market he sizes at $300bn. There are challengers for that crown too – ETSI’s Multi-access Edge Computing (MEC) is particularly telco-focused, like CORD, while the Cisco-initiated Open Fog Consortium comes from an IT heritage.
In late 2017, CORD announced a consolidated platform, CORD 4.1, that combined its residential, mobile and enterprise versions into a single system that allows all kinds of services to be supported on a single edge cloud. CORD 4.1 combines NFV infrastructure with the virtual network functions (VNFs), plus orchestration to turn services up and down.
ETSI kicks off zero-touch initiative:
ETSI has officially launched its latest industry specification group (ISG), focused on zero-touch management of telecoms networks, and hoping to drive significant progress to the goal of full automation in 5G.
The initiative, called ZSM (Zero-touch network and Service Management) ISG, originally announced last autumn, has signed up over 40 operators as members and held its first meeting last week. Its first priorities will be to explore use cases and requirements, and build a reference architecture to deliver an end-to-end view of zero-touch systems to enable automation and network slicing.
ETSI said software-defined networking, NFV and MEC would lead to an “exponential increase” in complexity, making automation a necessity across all processes, including delivery, deployment, configuration, assurance and optimization.
Klaus Martiny, senior program manager at Deutsche Telekom, was elected chair of the group. He said: “While 5G and its building blocks are being developed, it’s time to offer an end-to-end view focusing on automated end-to-end network and service management. We want to offer the market open and simple solutions. A continuous feedback from all stakeholders will lead to the first implementations of the specifications which will be tested through proofs of concepts, the outcome being fed back to improve existing specifications. A strong collaboration and cooperation with others standards bodies and Open Source projects is important for the ISG.”
Some of the founding members include Amdocs, Hewlett Packard Enterprise, IBM, Red Hat, Telecom Italia and Telstra while CenturyLink, Ciena, Comarch and EXFO are among those joining as participants. CableLabs has joined as a member.
Nurit Sprecher of Nokia (a former chair of ETSI’s MEC group) and Christian Toche of Huawei were elected as vice chairs. Ashiq Khan of NTT Docomo was elected chair of the Network Operators Council (NOC) advisory group, while Serge Manning of Sprint was named vice chair.