We reported Ericsson's announcement of their new RBS6402 multi-mode Enterprise Small Cell product at CTIA last month. It's one of several dual-mode 3G/LTE products – a capability that the industry has been setting as its ultimate goal for some time. Just like empty salt and pepper pots, these can be configured to serve up either selection. Even better they can be remotely filled with software and choose from one of several frequency bands to operate in.
We review the status of the various products announced, point out some trade-offs and consider whether they would suit every use case.
Dual-mode 3G/LTE products on offer today
It could be argued that DAS vendors reached this goal first. Many DAS systems are radio technology neutral and can handle 2G, 3G or 4G in any band – they simply distribute the RF around a building. However, many older products and existing installations may not support new frequency bands allocated for LTE. LTE features such as 2x2 MIMO may require different antenna and radio head functions. In practice, each installation would need to be assessed on a case by case basis.
The first commercially deployed dual-mode small cell was introduced by NTT DoCoMo in 2012 in a standalone residential product.
Huawei were also quick to develop their LampSite solution, which I see as more of a distributed base station architecture, akin to an RRH (Remote Radio Head). Each radio node can support 3G, LTE and Wi-Fi concurrently. This is a single network operator solution and connects through a centralised baseband controller in the basement. It's already commercially deployed in China and Europe.
I believe Spidercloud can claim the first true 3G/LTE Enterprise small cell to be commercially deployed. Their SCRN-310 supports both technologies and connects through their local onsite controller. It can be remotely configured between 3G and LTE for either radio, each operating at 125mW.
Cisco (through their acquisition of Ubiquisys) demonstrated dual-mode hardware last year and are probably the first with completely standalone dual-mode product. Their 7734 product spec indicates the two radios can be both 3G, both LTE or mixed. This is a standalone unit that doesn't offer Wi-Fi and doesn't plug into their Wi-Fi access points.
Alcatel-Lucent have been developing a high-spec 3G/LTE/Wi-Fi enterprise product based on the Qualcomm chipset. It will be the first multi-mode using that platform. They recently reconfirmed the development is on plan for field trials before end 2014 and commercial rollout during 1H15.
Ericsson launched their RBS6402 product at CTIA last month. This fits within their scalable RBS basestation product range and shares much of the software and technology. It's similar in spec to the ALU box, supporting 3G/LTE and Wi-Fi with 4 250mW radios (2x2 MIMO) and 1 200mW Wi-Fi. They are keen to emphasise it can be commissioned in 10 minutes through automated configuration – far shorter than the usual hours or days we associate with their larger macrocell cousins. It's marketed as a picocell rather than enterprise small cell and would use the Iu-b 3G interface to an RNC rather than Iu-h to a small cell gateway. I was told at the show that this is already at an advanced state of product readiness.
Ericsson have also been heavily marketing their Radio Dot solution for larger buildings. This is a similar architecture to the Huawei Lampsite. Commercial availability is planned for 2015 but I haven't found anything on their website indicating that it's capable of dual mode yet.
The above are all intended for Enterprise (i.e indoor non-residential) use. I've heard of a few vendors considering 3G/LTE dual mode residential products but apart from the DoCoMo product mentioned above, I haven't see others commercially deployed today.
There are many 3G or LTE standalone small cell products available today, but relatively few that do both yet. You could imagine that vendors who offer both separately could later merge these into a single platform.
My apologies if I've omitted or overlooked any other vendor announcements. Please feel free to add and correct this list by commenting below (can be anonymous if you prefer), and I'll update the article.
The underlying chipsets
A key factor in enabling such dual mode products has been the evolution of baseband chipsets with the capability to handle 3G or LTE through remote configuration. Some of the more powerful chips can handle both at the same time, while other architectures require a separate chipset per radio. Broadcom and Qualcomm seem to be well positioned for indoor enterprise dual-mode products, especially from new entrants. Freescale and TI are already embedded in many larger basestation products and have developed smaller scale versions suitable for this market. Intel and Cavium also offer reference designs.
Configurable frequency bands
Another useful capability is the ability to remotely reconfigure the frequency band used from a defined subset. For example, the Ericsson 6402 can operate across 10 bands (3 bands for 3G and 5 bands for LTE in addition to 2 for Wi-Fi).
While I expect this adds slightly to the cost and complexity, it provides a futureproof capability to refarm and reassign spectrum between technologies as the host network evolves. This helps make best use of the available spectrum as the industry migrates over from 3G and LTE throughout the next few years.
I confess to some initial surprise at the peak power consumption of some of these devices. Perhaps I'm used to residential 3G femtocells with a total power budget of 10W or less – some designs even fit within the 5W total capacity of a USB connection. Typical 3G/LTE/Wi-Fi multi-mode units can consume up to 50W of power and this just fits in the PoE+ (Power over Ethernet plus) specification – supplied using multiple pairs of a standard CAT6 Ethernet cable.
Some of the highest capacity DAS radio heads can take up to 90W of power, requiring proprietary schemes and/or extra cabling. It's much more attractive where the small cell can be connected with only a single CAT6 Ethernet cable which provides both power and backhaul.
The reason for the much higher power is the significantly increased RF output power of 250mW rather than 100mW. This can provide much greater range or penetration through walls but is still substantially less than an outdoor unit and is many cases won't be needed. There is also duplication where LTE uses two radios rather than one, using 2x2 MIMO for greater diversity which further increases range, coverage and performance.
Small Cell Gateways
The 3G small cell architecture uses the Iu-h interface to connect to a central small cell gateway, which is a mandatory network component. LTE small cells use the standard S1 and X2 interfaces so don't require one. For small scale trials and early deployment, they may not be needed. Some vendors such as Ericsson don't offer one. Experience from larger deployments, such as in South Korea, suggest that they are useful to scale and handle the signalling traffic, avoiding overload of the core network. This may vary depending on the capabilities of the core itself.
I've heard industry comment that many operators with sizeable residential and enterprise small cell deployments prefer to have a Small Cell gateway, allowing them to manage it independently. There are relatively few LTE small cell gateways available today – Cisco, NEC, Alcatel-Lucent, Contela spring to mind. Since operators wouldn't want to have a large diversity of different small cell gateways, I could expect the industry will consolidate on a few. Multi-mode 3G/LTE gateways would seem to be a good long term goal.