The mobile carriers have talked for several years now about moving to new types of networks built around small cells and cloud-based virtualization, and last year the more advanced of them started to practise what they preached with actual roll-outs. Public access small cells and the related trend of cloud-RAN will not become mainstream for another 12 months at least, and there are many issues ‘ such as interoperability and backhaul – still to be resolved, not in expensive semi-proprietary ways for a select band of pioneers, but in a standard way for all operators. However, nobody now doubts that small cells will be important in the 4G RAN, and that, along with the shift to an all-IP platform, could create the first major shake-up of the mobile infrastructure vendors for many years.
Fully aware of that, and conscious of being considered the plodding dinosaur in the small cell world, Ericsson unleashed a torrent of announcements in its pre-Mobile World Congress update, designed to stake its claim to be a leader for the new IP world, as it has been in 2G and 3G. However, while it unveiled additions for its AIR (antenna integrated radio) line, it has still stopped short of a real metrocell, as its attention increasingly shifts away from the base stations which made its fortune, to the platforms to control them.
The Swedish giant knows it will have to adapt to new platforms, margins and competitors, a point made very clearly by CEO Hans Vestberg at an analyst meeting in London this month. In the past decade, since Ericsson turned itself around after its turn-of-century collapse, the main changes in the competitive landscape have been consolidations, which have served to place more power and market share in the market leader’s hands. Its traditional rivals have either disappeared (Nortel) or merged but still failed to gain significantly on the leader (Alcatel-Lucent, Nokia Siemens/Motorola Networks). Huawei has replaced those companies as the main threat to Ericsson, but despite the Chinese firm’s broad reach, it has fought in wireless networks on very traditional lines. Even in the early 4G contracts, the competitive process, and the type of networks procured, have been old-style. Ericsson’s chief headache has been the tendency, especially in Europe, to focus on less profitable modernization schemes rather than big-bang capacity roll-outs.
But in the next wave of LTE, there will be far more deals that rely on new topologies like small cells, carrier Wi-Fi and convergence with other IP systems. Ericsson has expanded enthusiastically on the IP side in recent years, wooing wireline, media and quad play providers with content platforms, smart routers and IMS. It has been slower to embrace the new RAN, sniffing at the small cell and insisting it would be a medium term option for carriers, which would first concentrate on squeezing more capacity out of their macrocells. It has been proved right in many areas ‘ for every SK Telecom, AT&T or KDDI, with massive metrocell roll-out plans, there are a dozen carriers with a far more cautious approach.
But the momentum is building, and Ericsson is being dragged, however reluctantly, into the small cell platform which not only changes its hardware margins, but threatens to deprive it of control over its customers (if carriers manage to implement multivendor, interoperable networks), and to open it up to new rivals which work in a different way from its usual foes. Some of these will be smaller pioneers, like Ubiqusys, which can combine small cell innovations with powerful partnerships to tap into operators’ RANs. Others will be giants from the IP world, like Cisco, pushing itself right to the edge of the network with plans for its own metrocells.
All this could create major shake-up. According to a survey by Telecoms.com, two-thirds of telecoms professionals believe that the move to LTE will change the rankings of the equipment suppliers and open up routes for firms like Cisco and Avaya. And over half also thought firms which are outside the mobile infrastructure big six ‘ start-ups as well as Samsung and NEC – will be able to harness LTE and small cells to improve their standing.
Ericsson will certainly not let that happen easily. After several years when it sat aloof from the buzz about small cells, it has come out fighting. It acquired BelAir Networks a year ago to bring Carrier Wi-Fi expertise into its portfolio and now proclaims that the day of the small cell has arrived. Hans Wibergh, head of the networks division, said this month that Ericsson would ship its first commercial small cells this summer, and that the age of the metrozone had arrived. One powerful factor in the decision to concede to the small cell brigade will have been the US carriers, which have shown rising interest in deconstructed network approaches for their second wave of 4G roll-out. Ericsson is an LTE supplier for all the big four US cellcos ‘ with Alcatel-Lucent at Verizon, AT&T and Sprint (the latter also has Samsung), and with Nokia Siemens at T-Mobile. Verizon and AT&T started their 4G deployments conventionally enough with the larger cells that their 700MHz spectrum supports best, but will move to denser networks as they bring AWS and other higher band frequencies into play. AT&T, in particular, has an aggressive strategy to roll out 40,000 small public access cells this year, to add to its already significant investment in Wi-Fi and indoor femtocells.
All the national cellcos are planning to use a mixture of all-in-one metrocells, and distributed base stations (with remote radio heads and virtualized baseband processing, like ALU’s light Radio or NSN’s Liquid Radio). Such architectures are also becoming de rigeur in east Asia, where Ericsson will hope to gain ground against local suppliers like NEC in Japan, Huawei and ZTE in China and Samsung in Korea.
Initially, Wibergh sees public metrocells as an extension of the traditional network, with the network passing users from the macro layer to a small cell to gain greater capacity and quality of service. The next stage will be for devices to link to multiple cells, including Wi-Fi, simultaneously and without interference, one of the key goals of LTE-Advanced standards and the evolution of HetNet.
We will wait to see whether Ericsson unveils a true metrocell ‘ a compact unit based on a system-on-chip, with radio, antenna and baseband combined ‘ perhaps even with a roadmap to integrate BelAir Wi-Fi in full. For now, its main contribution to the deconstructed RAN centers on its AIR (antenna integrated radio) line, which provides a low-footprint unit for the cell site (though this can serve large as well as small cells). The radio/antenna is linked to a remote server where baseband processing for many cells is combined and virtualized.
The latest offering in the family is the AIR 32, which allows operators to add LTE or HSPA(+) without deploying additional antennas or remote radio heads (RRHs) on cell sites. The new model claims up to 70% higher throughput than its predecessors by implementing 4×4 MIMO antenna arrays, and up to 25% better indoor coverage, says Ericsson. It supports multiple active frequencies, reducing radio equipment by 50%.
AIR is a key element of Ericsson’s network modernization offering, which has been at the core of its mobile revenues in the past year. T-Mobile USA is using the platform in its own complex network transformation, which involves moving its HSPA services from the AWS band to some refarmed PCS frequencies (while leaving GSM in operation, at least for a while), and then deploying LTE in the vacated AWS spectrum. That will bring TMo’s 3G and 4G networks into the same bands commonly used by its rivals, opening up better roaming and device procurement opportunities, including the prospect of an iPhone.
TMo is currently using AIR 21 and will evaluate its new stablemate. CTO Neville Ray commented that AIR ‘offers a fast track way of getting mobile broadband services such as LTE into the marketplace. We can save months in terms of deployment with this product on a per-site basis. In addition, there are many performance gains and benefits in radio performance that will enhance the user experience of mobile broadband services.’
AIR is not in itself revolutionary and the time may well have passed for Ericsson to lead the way in pioneering new base station formats. And as the base station itself commoditizes, that could be a poor use of the Swedish firm’s R&D. Instead, it may rely on partnerships and acquisitions, like the BelAir one, for the site hardware, and turn its attention to the real source of power ‘ the smart network hardware and software which will control all these thousands of cells and direct traffic between them most efficiently. Here, just as Cisco says it will be reaching out from its core network kingdom into actual small cells, Ericsson will be going the opposite way, seeking a greater share of the IP intelligence activities dominated by its sometimes-partner, sometimes-rival. In this IP battle of the titans, there will be chances for innovative smaller players to gain a market position in areas where they can surround the giants’ main offerings ‘ and increasingly, and for the first time, the base station itself may become a ‘surrounding product’ for Ericsson, opening up opportunities for Ubiquisys, ip.access and other specialists.