When we last looked at the landscape for urban small cell backhaul, we found some 20 or more companies touting a range of solutions. These congregated into four or five clear market segments based on technology. We spoke to a number of different vendors to gain their perspective and understand how technology and the market have progressed.
The structure of the wider cellular backhaul market
The majority of cellsites today would ideally be connected by wireline, preferably fibre, but that isn’t available everywhere. Today’s wireless backhaul equipment market size is around $4.5 billion, split between $2 billion being bundled in with the complete RAN network (sold by Ericsson, Huawei, Nokia etc.) and $2.5 billion being procured independently (Ceragon, Dragonwave, Aviat, NEC etc). Of this independent market, operators want to pick and choose a vendor for each type of problem although some of the larger backhaul vendors have an extensive portfolio.
Quite a few operators are looking into Centralised RAN/Distributed macrocell approach. These aren’t IP based and may use much higher rate CPRI links. Other C-RAN architecture splits are being considered which aren’t so backhaul bandwidth intensive but it’s generally thought these would still be fibre based, limiting the locations and pace of deployment. A few wireless vendors are touting support for the higher data rates required by C-RAN.
If when you consider that the equipment market size for small cells across all segments is perhaps $1 billion today, you can see why it’s not attracted so much excitement from the mainstream backhaul vendors, but there is plenty of opportunity for it to grow.
Perspective of a major independent backhaul vendor
Shai Yaniv, VP Marketing at Ceragon, told me he can see carriers in the USA are now taking a more concrete approach to small cells, but even there can’t say it is yet ramping up beyond pilot capabilities. Ceragon have invested in their own silicon chipset (IP20 Multi-Core) which supports multiple baseband chains within the same silicon chip. This same chip is used within both macrocell and small cell backhaul products. They can squeeze 1Gbps through a 28MHz channel, and where consolidated at aggregation nodes it can separate adjacent sectors by as little as 15 degrees.
Shai’s view was that fibre backhaul would be used when available at a cost where it makes sense. He wouldn’t agree that all urban areas will “fibre up” – even today a lot of macrocells today remain connected by wireless backhaul. Wireless is often much faster and more cost effective to deploy.
There are numerous smaller companies, including many startups, focussed on a specific technology. Examples (in no particular order) include:
Tarana, NLoS technology, continues to mature and expand their capabilities. Publicly showcased by Vodafone at the SAIL2015 event in Amsterdam, they can stream 200Mbps across quite hostile and changing environments. There’s no need to align antenna (although that further helps the link budget).
E-Blink, speaking at a recent Cambridge Wireless event, explained several practical use cases for their point to point link which operates in the unlicenced band. By using highly directional antenna, they can quickly extend the CPRI links across line of sight, such as for remote radio heads.
CCS, also speaking at the same event, have expanded the throughput of their multi-point to multi-point mesh solution. They claim this means it could be used for C-RAN deployments, although it might struggle to achieve full CPRI rates for multiple nodes. Their 270 degree antennas and mesh architecture facilitates rapid deployment and tolerance to outages, avoiding extra site visits.
We’ve also seen a number of other vendors visible at events including Dragonwave, Siklu, Fastback, CBNL, Vectron, Intracom to name just a few. All of these market their wares to several market sectors and don’t rely exclusively on the small cell segment. Popular applications include campus-wide data networks, bridging links between buildings, CCTV and public Wi-Fi.
A few of the small cell players offer their own integrated backhaul solutions with both Airspan and Parallel Wireless perhaps the most notable.
Battlelines being drawn up for the years ahead
When you consider the market forces, there are several “battle lines” to be drawn and considered.
- Between “bundled” and “independent specialists” backhaul vendors. Will the major RAN vendors ensure their portfolio includes the full set of technology choices, reducing the need for operators to shop directly with independents? We’ve seen a number of reseller agreements with some of the niche technology companies to fill out their portfolios, and in the long term these might evolve into acquisitions if business volume and differentiation justified it.
- Between the full portfolio and niche independent backhaul vendors. The larger independents – Ceragon, Dragonwave, NEC etc. – span a wide portfolio of technologies in which they invest. Several of the smaller startups focus on one or two very specific niches, and some have been quite well funded. Can the larger businesses continue to remain best of breed in the face of so much innovation using widely different new technologies?
- Wired vs Wireline market share. Based purely on technical performance, networks would prefer dark fibre to every site. Practically, that’s expensive, takes time and in many scenarios infeasible. The wave of small cell deployments is likely to rush ahead of a supply of plentiful dark fibre, shifting the proportion of backhaul for outdoor small cells in favour of wireless. On the other hand, I’d expect where C-RAN is deployed the mix would shift more towards fibre.
Those factors above will be skewed by many other factors including but not limited to:
- Availability of spectrum, driving urgency to densify via small cells
- Adoption of spectral efficiency improvements, ranging from Full Duplex, MIMO to CoMP
- Enthusiasm for fully independent Small Cells vs Remote Radio Heads or C-RAN
- Operator preference for investment in-building vs outdoor small cells
My take on the situation is that we are likely to see some of the smaller innovators drop out of the race or become subsumed/acquired. I’ve been pleasantly surprised by how well a few have done (including raising further funding). Many of those surviving in the race are strong contenders to win out in the longer term. Some will likely be acquired either by the larger independents or RAN vendors looking to balance their portfolio.
As urban small cells start to become more significant, we will be watching closely to spot any shifting patterns in these three sets of battle lines.