02 December 2019 | Annie Turner
The telecoms industry has a history of collaboratively addressing tech and operational issues through trade associations and standards bodies. Digital transformation has proved a bonanza. Annie Turner looks at four of the main ‘transformers’.
The great thing about digital transformation is that it has immensely complex, with many facets. From the tech end of the deal, it is generally seen as shorthand for the necessary overhaul of operations, IT and network architecture to support new business models as old sources of revenue erode, including the deployment of 5G.
Digital transformation also means being a customer-centric (instead of network- and process-centric) business, and armed to make smarter, data-based decisions by applying analytics and artificial intelligence. It involves adopting DevOps approaches to make operations more agile, network virtualisation, becoming cloud-native and platform-based, automation, orchestration, partnerships, ecosystems and much more.
No wonder the standards organisations and trade associations have been kept busy, typically with clutches of techies working on projects that are impenetrable to most of telcos’ C-suite, not least because of the blizzard of abbreviations and acronyms they generate.
Here’s a short guide to the main would-be transformers:
The O-RAN (for open radio access network) Alliance was created by the merger of the xRAN Forum and C-RAN Alliance. It caused a stir when announced at MWC in 2018 by its founding members – AT&T, China Mobile, Deutsche Telekom, NTT DOCOMO and Orange. They had two intertwined goals: to evolve the RAN through virtualised radio network elements, white-box hardware and standardised interfaces to gain greater interoperability and reduce costs.
More bluntly, they saw the digital transformation of the RAN as being in large part to counter the large, established equipment makers’ (Ericsson, Huawei and Nokia) practice of using proprietary connections between the radios and other elements thereby obliging operators to buy complete, expensive systems. Many smaller and newer market entrants saw the move to disaggregated, virtualised and software-controlled RAN equipment as a good way to compete against their powerful counterparts.
As Andre Fuetsch, President AT&T Labs and CTO at AT&T, said at the time, “To take full advantage of the flexibility of 5G, we have to go beyond the new radios and change the overall architecture of the end-to-end system”.
The original five were joined by Bharti Airtel, China Telecom, KT, Singtel, SK Telecom, Telefónica and Telstra. These 12 operators co-signed the Constitution Articles of the O-RAN Alliance at MWC Shanghai in June 2018.
In February, the O-RAN Alliance announced the release of the first O-RAN standard Open Fronthaul Specifications which includes protocols for the control, user, synchronisation and management plane. The following week, it ran six demos of the specifications in action.
By that time its ranks had swelled to include operators, KDDI, SoftBank, TIM and Verizon (but note not Vodafone – see below). Vendor members include Cisco, Ericsson, Fujitsu HFR Networks, Intel, Keysight Technologies, Mavenir, NEC, Nokia, Pivotal Commware, Radisys, Samsung Electronics, SOLiD and VIAVI. For some, this might look like turkeys joining the Christmas Club, but you have more influence inside the tent than outside it.
Despite complaints that some vendors have a different definition than everyone else of what ‘open’ means and simply can’t kick their proprietary code habit, it seems progress is being made. For example, as we went to press, NXP Semiconductors announced new 5G processors that reportedly work with the Alliance’s specs.
Already there is unease in some quarters about the dangers of operators trading in reliance on big equipment makers for dependence on a handful of system integrators, as operators increasingly turn to them for help in managing larger numbers of smaller hardware and software companies.
In step with the shift towards open source by telcos, which was an extraordinarily long time coming, in April 2019 the Alliance joined forces with the open source Linux Foundation on the creation of the O-RAN Software Community (O-RAN SC) to develop software for the RAN. The idea is to draw on some of the Foundation’s other projects, but that is not without its critics either (see Open Network Automation Platform – ONAP – below).
A final note…
Do not confuse the O-RAN Alliance with OpenRAN, the project spearheaded by Vodafone and Telefónica since 2017 within the Telecom Infra Project (TIP), which Facebook initiated. Its goals of hardware- and software-neutral, interoperable RANs, at less cost and a greater pool of suppliers are strangely familiar.
In October 2019, Vodafone Group’s CEO, Nick Reid, sent a ripple across the industry when he announced, “We are pleased with trials of OpenRAN [in Turkey and South Africa] and are ready to fast track it into Europe as we seek to actively expand our vendor ecosystem”. Vodafone’s European rollout will begin in the UK. It will also deploy OpenRAN in the DRC and Mozambique – after all one of TIP’s original aims was to bring affordable to poorly served communities.
TM Forum is arguably the granddaddy of them all, founded in 1988, but has had a few name changes since, most recently from the TeleManagement Forum to TM Forum. For a long time, TM Forum was primarily about streamlining and improving the interoperability of BSS/OSS, and its eTOM and SID blueprints were and are used by operators all over the world.
One of its co-founders, Keith Willetts, was quick to understand the importance of digital transformation to telcos, and after a few false starts, TM Forum hit upon a plan in 2016. Led by Vodafone with BT and Orange, it launched an open APIs programme to help telcos’ adopt platform-based operational and business models. These are the most efficient and effective models ever devised for bringing disparate parties together to interact, as the staggering success of the FAANGs has illustrated.
TM Forum’s suite now comprises more than 50 REST-based open APIs which are contributed by member organisations then rounded out and standarised collaboratively. Whether they will ever be widely deployed by operators to provide digital services to businesses and consumers through an ecosystem of partners, which is the goal, remains to be seen. They are however being used by some (notably BT, Vodafone and Orange) as part of a platform approach to streamline their internal operations.
At the launch of its Open API programme in 2016, the Forum said it would have 200 operators signed up by 2018. AT&T, China Telecom Chunghwa Telecom, Deutsche Telekom and Salesforce signed up this year, joining the original signatories taking the total to 17 operators, although most of the biggies are in there, plus 36 others, mostly vendors. Only Ericsson has signed up from the ranks of the very largest.
MEF was the Metro Ethernet Forum at launch back in 2001 when it was dedicated to developing and positioning Ethernet as an access network technology. As infrastructure started to become ever more software defined and controlled, and will be even more so as 5G unfolds, MEF moved into standards for virtualised networks and changed its name to MEF Forum (more commonly referred to as MEF).
MEF claims more than 200 member companies, and has published more than 60 technical specifications and implementation agreements. Its current scope, MEF 3.0, work includes optical, transport, carrier Ethernet, IP, software-defined wide area networks (SD-WAN) and cloud services, as well as lifecycle service orchestration (LSO) – the term originated with Oracle.
MEF and TM Forum are a little like stalagmite and stalactites, getting closer to meeting in the middle. This two-way flow is because the routers, element managers, datacentres and so on that sit below BSS/OSS typically were traditionally the domain of 3GPP, European Telecommunications Standards Institute (ETSI) and MEF, but now they are exposing software-enabled services up to the BSS/OSS.
Against this backdrop, MEF, TM Forum and the Linux Foundation announced collaborative partnerships in 2017 to better coordinate their efforts.
This bore fruit the following year in ONAP’s second, Beijing Release, published in June 2018. The Release enables orders to be placed and fulfilled as well the monitoring the quality of service, with TM Forum’s Open APIs carrying MEF-defined payloads to provide the northbound interfaces. Their standards and assets have been incorporated in subsequent ONAP releases.
Operators flocked to get involved with the Open Network Automation Platform (ONAP) in 2017 and 2018, which runs under the aegis of the Linux Foundation. Many had become somewhat frustrated with the ETSI approach, known as OSM for open source management and orchestration. ONAP brought together the ECOMP framework developed by AT&T and the Linux Foundation’s Open Orchestrator Project.
ONAP is designed to simplify the operations, management and control of network elements and services across multiple domains, including wireless and optical transport networks. Its common, open framework is supposed to make it easier to integrate solutions from many vendors.
Most of the usual suspects are members, with the notable exception of Telefónica, which is only indirectly involved, for example, NEC and its subsidiary Netcracker have been long-term strategic partners for years with projects in many countries and they participate in ONAP.
Naturally there’s trouble in paradise. There are grumbles that progress is too slow and Ibrahim Gedeon, CTO at TELUS, claimed earlier this year that open source is just “the new model of taxes on telcos”.
At Broadband World Forum in Amsterdam, Mansoor Hanif took a pot shot directly at ONAP. He is the CTO of the UK regulator Ofcom and former BT executive, and warned against “over dependency on any single automation platform”. Clearly not everyone is delighted at the rising power of open source and the Linux Foundation in telecoms.
04 December 2019 | Natalie Bannerman
04 December 2019 | Gareth Willmer
04 December 2019 | Natalie Bannerman
04 December 2019 | Geoff Bennett