One of the key goals of the operators pioneering small cells in their public access networks is to break the lock-in of large OEMs. While some will cautiously listen to the major vendors’ arguments that it will be too complex and risky to roll out multivendor metrocell networks, many believe it is essential to stick to that ambition. The Small Cell Forum’s recent plugfest, its fourth, was a clear statement that, even in the complex LTE metrocell space, the vision is to have interoperable, affordable hardware from a variety of vendors, all working transparently with other suppliers’ cores and gateways.
That would see the small cell layer of the mobile infrastructure looking far more like public Wi-Fi, with low cost, mix-and-match, commoditized access points, relying on a supply chain that is far closer to consumer electronics than traditional base stations. In that model, hardware cost and margins are those of the Taiwanese ODMs, not the major infrastructure suppliers. Three factors, previously marginal in base stations, become important sources of power ‘ the reference design; the licensed multivendor software; and the gateway holding it all together.
The first of these offers a chance for non-traditional suppliers to enter the market, providing platforms which can allow manufacturers to make small cells cheaply and get them to market quickly. Makers of commodity hardware pre-integrate their cells with openly licensed protocol stacks and small cell software to create low cost, interoperable units with low development overheads.
Companies like Tektelic and Purewave make small base stations themselves, but are increasingly showing reference platforms which can be adopted by others, while this approach will start to filter into the Taiwanese white label model once metrocells pick up volume and become fully standardized. That opens up significant opportunities for companies which make the vital element of the reference design, the software, making firms like Radisys, Aricent and MimoOn very important in the evolving supply chain.
These dynamics are common in Wi-Fi and, of course, in home gateways and other equipment. This has been one of the most important impacts of the domestic femtocell, which showed how consumer kit economics could be applied to cellular infrastructure, with the small cell players increasingly getting their power and differentiation from software or silicon, both of these applied to a wide range of end products rather than a single vendor’s range.
Ubiquisys, now part of Cisco, saw this change coming when it decided that licensing its technology would become its main business model, rather than manufacturing. It also tapped into the norm of the consumer equipment space, the ‘out-of-the-box’ reference design which can then be used by and ODM to get a product to market quickly and cheaply. The Femto-Engine enabled Ubiquisys to claim a first in breaking the magic $100 price barrier in 2010. I
Alcatel-Lucent also went down the licensing route in 2012, and its small cell software, combined with reference designs which pre-integrated silicon, protocol stacks and intelligence, allowed large Taiwanese players like Alpha Networks and Askey to get into the market, creating economies of scale which will now, gradually, spread to the public access metrocell space.
These metrocells are more complex and expensive beasts than their residential cousins, of course, and have to be fully carrier-class, mirroring the functionality of a full macrocell in a small, low cost, low power form factor. That does not preclude the reference design approach, however much some OEMs would argue that these small base stations must be as highly engineered (and proprietary) as the macro versions. Greater complexity is required in the silicon ‘ as seen in the entry of Freescale and TI, which had shunned the residential space, with metrocell versions of their mainstream architectures. And there are major challenges to tick all the carriers’ boxes in terms of software, to prevent interference, distribute traffic intelligently and so on. But as in carrier Wi-Fi, the smarts lie in the software and the core networks, while the boxes themselves can be commoditized.
The new structure of the market has been highlighted by several interesting recent launches in the LTE metrocell space. Tektelic recently unveiled a pre-integrated outdoor small cell platform which can be the basis of third party products, not just its own, and can also be adapted for NLOS backhaul. The 2220 LTE product is particularly targeted at vendors which serve vertical markets such as emergency response and combines Texas Instruments’ system-on-chip with MimoOn’s small cell PHY, protocol stack and scheduler. Tektelic and Purewave both acknowledge they do not have the brand power of the big OEMs but can provide pre-integrated designs for others, drawing on the availability of openly licensed elements ‘ such as PHYs and protocol stacks ‘ which were previously developed expensively inhouse by each vendor.
Other stack providers which are making an impact on this new supply chain include Radisys and Aricent, both of which, like MimoOn, have worked with specialist silicon providers, like Mindspeed/Picochip to launch reference designs. At this year’s Mobile World Congress, Mindspeed and Aricent showed off a 20MHz small cell reference solution which saw the former’s Transcede T3300 baseband processor and PHY software integrated with the latter’s stack. Aricent’s software supports functions such as VoLTE and self-organizing networks and is enabled for LTE-Advanced.
Another recent demonstration which indicated the power of reference platforms ‘ to prove new functions at low cost and risk and bring them to market quickly ‘ combined Radisys’s software with Airspan’s AirSynergy small cell, supporting LTE-Advanced features. The AirSynergy 4G harnesses Radisys’ Trillium TotaleNodeB metrocell software, which accelerates time to market considerably.
The mini-base station is an all-in-one compact outdoor picocell supporting key features from the LTE-A menu such as SFR (soft-frequency reuse via distributed eICIC); almost blanking subframes (ABS); X2-based automatic neighbour relations; and range extension.
The Trillium TotaleNodeB software handles radio resource management, SON and OAM functions and runs on top of the Trillium LTE protocol stacks to provide an integrated, pre-verified solution for vendors. It is verified across most small cell SoCs.
Aricent has also recently worked with Fujian Sunnada of China to develop a full TD-LTE small cell solution for the Chinese market. This integrates Aricent’s eNodeB stacks with Sunnada’s hardware, helping bring a non-established vendor into the market. Huang Haifeng, a Sunnada director, said: ‘Aricent’s readily available LTE software frameworks and expertise in integrating and building the complete solution has helped us create a highly differentiated product, while helping us reduce costs and time to market.’
Also important to this multivendor vision are high class gateways. NEC is a leader here and recently announced an LTE small cell gateway which can support 1,000 access points from a single compact unit ‘ with a successor capable of linking to 10,000 metrocells on its way. The gateway is based on standardized ATCA hardware and it can be scaled up with the addition of new blades. It can reduce the burden on the core network by aggregating many small cells, rather than each of these connecting directly to the core.
The new-look ecosystem could transform the economics of base stations, if it is allowed to mature, and it paves the way for fully multivendor networks, and therefore greater competition and price reduction. However, it is vital that interoperability ‘ the basis of a multivendor, low cost ecosystem ‘ does develop to the level in which carriers have confidence. As in Wi-Fi, this new structure makes interoperability testing vital, to ensure that the low cost base stations do talk to one another. Many challenges remain here, particularly in areas like interoperability between different SON (self-organizing network) schemes. And the Small Cell Forum is steering clear of devising a fully fledged certification system like that of the Wi-Fi Alliance. However, it is stepping up its activities in running plugfests, and in 2010, reference designs took part for the first time, indicating that they had become solid enough to enter the mainstream.
‘The fact that the plugfest was carried out so quickly after the completion of the 3GPP standard [Iuh for common small cell connectivity] is testament to the industry’s support for standardized femtocell access points and network equipment. This process will ultimately allow operators to multisource the technology as they do with mobile handsets today,’ said Natasha Tamaskar, who was then chair of the Femto Forum’s IoT Initiative.
This year, the Forum held its fourth plugfest, and its first focused on LTE, hosted by the Sintesio lab in Slovenia. The organization said this aimed ‘to cultivate an effective ecosystem of interoperable small cells and provide operators and consumers with a wider choice of small cell products while also facilitating economies of scale.’ The plugfest particularly concentrated on interoperability of multivendor small cells with different EPCs, security gateways, macrocells and HeNB gateways. Participants were Airspan, Argela, Athonet, Cisco, Fujitsu, JDSU, NEC, Node H, Qucell, Quortus, Radisys, Sanjole, Spidercloud Wireless, Stoke, Telekom Slovenije and Iskratel.
‘The Plugfest series is helping to drive forward LTE small cells as an open and interoperable technology ‘ a critical element in their successful uptake,’ said Neeraj Gupta, chair of the Forum’s LTE working group. If this outcome does indeed come about, it will bring a host of new names into the base station supply chain, while the large vendors will find their power shifting to integration services, and the software and core network elements to tie all the commoditized cells together intelligently.