LTE’s high performance combined with newly available sub-1GHz frequencies means it reaches places that 2G or 3G often can’t. But where VoLTE isn’t used, voice service suffers. This causes a dilemma over when and how LTE only vs MultiMode 3G/LTE small cells may be rolled out. We try to separate the facts from the myths to discern what steps will be taken on the path toward LTE only products.
Quick recap on voice and LTE
It often seems that LTE was designed primarily as a data service, with voice added as an afterthought. A simple solution would have been to pipe 2G/3G voice service (which is already conveniently sent in packets) over IP, perhaps with some simple prioritisation or scheduling.
Instead, the industry took a more radical decision to introduce a new IMS core and adopt SIP as the call control protocol, known as VoLTE (Voice over LTE). This allows a range of expanded services such as video calls, known as RCS (Rich Communication Services). Perhaps it was felt this allowed direct competition with emerging OTT services such as Skype, Facetime or Facebook Messenger. It may also provide more synergy with fixed line services.
Introducing VoLTE has been a very slow process. Installing the new switching nodes involves integration with back office systems – everything from post-paid billing, pre-paid online control to handling voicemail and emergency calls correctly. The new systems have to interwork seamlessly with the old. Many internal provisioning and configuration processes were built with the 2G/3G core network in mind and have had to be redesigned. A hidden cost is that the latest smartphones are required, not just any LTE one – Verizon Wireless don’t support VoLTE on any model prior to the iPhone 6.
Fallback to 2G/3G instead
So networks have instead relied on CSFB (Circuit Switched Fall Back), which reverts back to 3G or 2G to handle any incoming or outgoing call. When connected through LTE, notification of an incoming call requires a search for available 2G/3G service and handover to accept it. This adds delay, often unnoticeable to both callers (the period of ring tone just lasts longer), but in extreme cases results in the call being terminated before being answered.
In areas with good 2G/3G coverage, fallback isn’t a problem. But in areas where 2G or 3G previously didn’t reach or was poor, calls will drop or just not be connected. Users can’t quite understand why their phone shows 4G with perhaps 3 or 4 bars but can’t make or receive calls. If they turn off LTE, then they don’t get any service at all.
What to do if 2G/3G isn’t that good?
For in-building coverage, a 3G small cell should resolve the issue and generally does. An issue can arise where LTE service is being delivered from outside the building, but CSFB reverts to the outdoor macrocell 3G rather than the in-building small cell.
An example is the upper floors of high rise buildings, which can be within range of many macrocells from nearby rooftops. Interference from multiple cells can reach levels that make calls garbled and unusable.
A simple solution is to install one or more 3G small cells so that users on those floors are directly connected and handle all calls locally. This also benefits the outdoor macrocells by removing a source of common interference for them. I’ve come across several case studies where penthouse apartments have used these to great effect.
But what happens when 4G comes along?
Newer smartphones capable of LTE would normally be configured to seek out LTE where available. It offers faster data rates and greater spectrum efficiency.
When visiting buildings fitted with 3G small cells, they may well ignore them in preference to a weaker but usable LTE signal from outside. The CSFB issue highlighted above then becomes a problem. Voice service problems previously resolved using an inbuilding 3G small cell now return, because the small cell isn’t being used.
Typical 3G small cells are capable of data rates of 16Mbps or faster, shared between very few concurrently active users. They provide high quality voice and more than adequate data service compared with busier or congested outdoor macrocells.
A simple test if this is the case would be to configure smartphones not to use LTE, and force them to camp on to the inbuilding 3G small cells.
While this may solve the problem, it may be unreasonable to expect users to remember to switch on/off LTE every time they exit/arrive at the building. This has been a common issue with Wi-Fi, which some prefer not to use when out of the home/office environment.
Introducing LTE small cells
I've seen LTE only small cells exhibited at events for quite some time. A few networks have started to deploy them, such as China Mobile and Verizon. These have both invested in very comprehensive LTE coverage and for legacy reaons with their 3G networks want to make the switch as quickly as possible.
Elsewhere, they haven’t taken off because relatively few networks today have deployed VoLTE.
As with core networks, operators will need to run both 3G and LTE concurrently for some years. The more complex and modern technology means that LTE only small cells are more expensive than 3G only today. Inevitably a combined 3G/LTE product is more costly than either single mode.
However a single baseband chip can now drive both 3G and LTE concurrently, so the incremental cost may not be quite as large as you’d think. That could be further simplified if the same frequency band is used for both generations, sharing the front end RF stage. Once example outdoors we’ve seen is where 1800MHz being used for both 3G and LTE by EE UK.
So what strategy will operators choose?
There seems to have been reluctance on the part of many operators to introduce LTE only in-building small cells. Most of those already with 3G have retained them and continue to ship them to customers with coverage issues. Multimode 3G/LTE products are available from vendors, but need to go through an approval process before being more widely deployed.
My guess is that we’ll see a few LTE enthusiasts adopt LTE-only small cells while many stick with 3G ones for the time being. Once 3G/LTE multi-mode become commercially available (and approved), these will be the new gold standard. The incremental cost of the multimode small cell may be a lot less than upgrading outdated smartphones to VoLTE capable models, but this will change as the installed base of smartphones transitions to newer models.
In 3-5 years time, as VoLTE becomes more widespread – both in the networks and in the smartphones being used – operators may switch to LTE only products. I don’t think this will happen overnight, because there will still be demand to support older 3G phones and services including those visiting from abroad.
Outdoors, for urban small cells, I think LTE only is likely to be adopted from the outset. There is already generally very good 3G coverage in these areas, and the urban small cells are being installed primarily for capacity. For rural areas, 3G/LTE multimode would be the preference - it may take longer in those areas for the latest smartphones to be widespread.
These are the questions that small cell product managers (and their component suppliers) have to answer and justify future investment plans. Large bets have been made.
As always, if I’ve misrepresented or overlooked the issues, feel free to comment below (can be anonymous).