Symmetricom provide the end-to-end timing solutions and grandmaster clocks used by the majority of telecommunications networks worldwide. They work closely with the majority of small cell designers, vendors and operators. I asked Manish Gupta, VP Marketing and Business Development, to share his insights into how the small cell marketplace is evolving, and explain what issues are holding back large scale deployment of small cells.
He reveals what current small cell hardware costs, believes the number of small cell products coming to market will increase dramatically over the next 12 months, but forecasts a rationalisation of the number of ODM small cell vendors.
How do Symmetricom perceive the small cell market evolving?
There's no uniform answer to this question – it is very much dependent on which region and which operator you look at. Examples include Free France, who have been very aggressive, not making any distinction between enterprise and residential use. Whereas at the other side of the world, we find both Korea and Japan have moved ahead strongly with both metrocell and enterprise but without a significant focus on residential.
In the USA, all four main operators have publicly announced substantial plans for small cells. Overall, we've seen steady ramp for residential, good early growth into the enterprise and a lot of focus and interest in public access/urban small cells.
If you are looking for a uniform rollout of small cells in the near term, then I doubt you'll see that. Instead look closely where LTE has already been extensively deployed and we should see small cells becoming centre-play. These implementations are starting now and will be much more visible over the next 12 months.
How has this affected vendors and the supply chain?
Most small cells deployed today are 3G only, and while that functionality isn't being abandoned, we can see the industry moving towards both hybrid 3G/LTE and LTE only products. Silicon vendors are investing heavily to develop next generation SoCs (Systems-on-a-Chip) with multi-mode capability and low overall power consumption.
Significant players, such as Alcatel-Lucent, have made huge organisational shifts across their business prioritising small cells at the centre of their wireless strategy going forward.
Small Cell product cost remains a significant factor, with the economics underpinning large-scale rollout dependent on low cost per unit. We are seeing the BOM (Bill of Materials, excluding software licences) for residential 3G sit across the $50 to $100 range, with LTE significantly higher. The BOM for low power metrocells is in the $200 to $300 range for either 3G or LTE only. Longer term, prices continue to trend down as volumes grow. Multi-mode 3G/LTE and/or higher RF power radios are more expensive.
How has this affected the ODMs (Original Device Manufacturers)?
We are seeing the same evolution cycle in the Small Cell world that we've previously witnessed in other domains such as wireline broadband DSL or WiMAX.
When we reach a point of design maturity where it becomes straightforward to design a box – at least a lower powered one – and there is a promise of significant equipment volumes, it becomes exciting for the ODM community to join in. We've noticed an increased number of participants over the last 18 months, especially in the major hubs such as Korea and Taiwan. We discuss this frequently with a large number of players but I believe we won't see many more new companies entering this market, although we will see the number of small cell products increase dramatically. At the same time, some of the ODMs and OEMs who haven't captured the larger deals will lose interest.
After the early explosion in the number of new participants, I wouldn't be surprised if there was some contraction in the number of vendors playing in this field. Small cell products are beginning to reach levels of maturity and uniformity, so we won't see radically new form factors and instead a better cost structure and implementation.
I expect that cycle to reach maturity level within the next 12-18 months.
Given that LTE-Advanced features demand tighter phase timing, what are the implications for network designers?
Of the operators we've spoken with over the last 6 months, 18-20 are contemplating LTE-Advanced services in the near term. Operators can select which of the LTE-A features they want to adopt as well as where and when. Two popular options are eMBMS (Multimedia Broadcast Multicast Service which allows the same datastream to be multicast to many smartphones in the same sector) and CoMP (Co-ordinated Multipoint- where the same data is transmitted from multiple base stations and especially helps improve performance at the cell edge), which both require eICIC (evolved Inter-Cell Interference Co-ordination – efficiently shares the same spectrum between large and small cells).
The design of small cells and end-to-end network timing architecture needs to support that, and these features need to be supported in designs being developed today for commercial deployment in 12 months. There is wide appreciation of the difficulty of achieving the phase sync in this ever more complex environment. Major operators recognise that they need to work with a major sync vendor at an end-to-end network level – standalone solutions such as embedded 1588 aren't enough on their own.
To meet these tight phase requirements, critical considerations include end-to-end time error budget from the grandmaster to the small cell and asymmetry correction. These are difficult topics that few people in the industry have tackled. This translates into the need for an operator to carefully engineer their network and consequently place their sync and timing devices closer to the small cells than in the past with 3G where phase was not required.
A standalone edge master is the answer that supports the critical PRTC (Primary Reference Timing Clock) requirements defined by the ITU-T. Sharing timing capabilities with other functions like switching or demarcation present some challenges which Symmetricom advises against given the tight phase challenges.
eICIC requires 5us or better sync accuracy. LTE-TDD defines 1.5us requirement and services such as CoMP can even go down to 0.5us. These are on-going discussions both within operators and standards bodies, and these debates are likely to continue for some time. Meanwhile, some operators have a greater sense of urgency and want to be first to market, so are taking their best shot with a view to refine/revise their specifications as they learn through experience.
This is by no means a uniform requirement worldwide, but the vast majority of discussions we are having involve small cells being a tightly integrated part of a full HetNet environment rather than having separate zones with or without small cells. Zone separation could become an incredibly complex operational exercise for operators to delineate and manage.
Is that also the case for indoor small cells?
I could agree that these LTE-A features are unlikely to be used all the way through to residential, but we would expect it for most public indoor deployments.
Symmetricom has undergone significant organisational change during 2013. How has this affected their commitment and investment towards the small cell market?
We have a new CEO in place and other significant management changes, which have led to refocusing on specific growth areas. We had a number of legacy product lines and diverse exposure to multiple market places, and need to prioritise our R&D dollars into those areas that will give us the best returns.
LTE small cells and related packet sync definitely falls into one of these primary growth areas, alongside atomic clocks and government sectors. We continue to see small cells to be very important and continue to offer an end-to-end solution, with our timing embedded into silicon designs and edge-master portfolio for deployment at cell sites/indoor hubs complementing our core network grand master clocks.
If anything, our commitment has only increased because of these changes that allow us to prioritise our investments and focus on this growing opportunity.
Symmetricom established a strong eco-system called SyncWorld last year. How has this evolved and is it still relevant?
Syncworld has been successful to a level above and beyond our initial expectations. We now have over 50 partners including major basestation and switch routing manufacturers – covering over 80% of the marketplace. The ecosystem includes small cell vendors, backhaul players, semiconductor and software providers, and test/measurement equipment.
We have now reached a point where we are no longer aggressively trying to increase the size but are focused on strengthening the quality of engagement between partners. We have had various interoperability testing sessions, and make reports available to operators specifying the pre-determined performance they can expect. This provides much greater value and shorter time to market than before, when only theoretical/predicted figures could be given.
And finally, are there any specific milestones or capabilities we should be looking out for which will facilitate the wider adoption of small cells?
Broadly speaking, the biggest issue for small cells today is around the ease of deployment – how do you ensure they have full HetNet capability in a simple to use "plug-and-play" installation process. From a timing standpoint, the standards continue to evolve and agreements reached, which is making it easier for operators to become more comfortable with what is essential for a synchronized HetNet environment.
Symmetricom is a sponsor of ThinkSmallCell. You can read their white paper on Timing and Sync for Small Cell networks here.