Densification is integral to 5G for meeting the key objectives of high data rates, greatly increased capacity and to some extent ultra-low latency. Higher frequency radiation provides the capacity and bit rate, but in turn suffers more rapid degradation and requires near line of sight. Low latency requires edge computing and also close proximity to devices for some of the more extreme use cases. So while 4G RANs were about 10 times denser than their 3G predecessors, 5G will require a similar increase in density.
This will involve a mix of small and macro cells, along with in-building developments and also integration with other wireless technologies, especially WiFi. Indeed, as we have argued before, presentation of WiFi and cellular as competitors rather than complementary technologies should fade in the 5G era, even if they are binary alternatives in some use cases, such as V2X communication in the automotive sector.
Many studies, papers and seminars have been devoted to the technical aspects of 5G densification, but the greatest challenges are in fact commercial, political, regulatory, and logistical, especially relating to acquisition of sites and associated planning consents. It is those factors that account for nearly all time and cost associated with deployment of small cells, especially at new sites.
Depending on the territory, this may involve site and equipment approvals, negotiation of fees with municipalities, owners or landlords, as well as insurance that the deployment conforms with local environmental and aesthetic requirements.
There is also the question of backhaul, which could require permissions or partnerships, as well as delays and costs when fiber has to be laid for example. The problem is compounded by variations not just between countries, but down to regions and even individual cities. For these reasons, Sprint in the USA has indicated that while deploying a small cell might take just a day, it may take a year or more to obtain the necessary permit, imposing cost and delay.
It is true some countries have laxer regimes than others and one reason China has a lead over every other significant nation in 5G deployment is that any local opposition tends to be smothered, as in the case of all national infrastructure projects deemed to be of strategic importance there.
Elsewhere, regulators have acted to establish a more harmonized framework for densification and small cell deployment, aiming to streamline approvals and reduce delays. Sometimes this happens at just a local level, as in the city of San Jose in the heart of Silicon Valley, where individual deals extend to the whole area to avoid having to submit multiple planning applications.
At the other end of the scale is the European Electronic Communications Code (EECC), an EU Directive due to come into force at the end of 2020, setting consistent rules on issues such as rights to install new telecoms equipment. This should then be enacted by member states with the stated objective of accelerating densification. It is not clear though how the EECC will interact with local rules, or affect rights of individuals to object to planning applications. It is that process as much as any other that holds up cell deployments.
There is a natural overlap between these logistical and local aspects and technical considerations over design of the small cell stations themselves. It has created pressure to reduce size of cells and package them flexibly as well more aesthetically to suit varying local considerations. This brings in technical aspects not directly related to performance or robustness of the radio equipment itself.
The equipment must be housed in a small form factor that delivers the required RF performance and yet fit in the space available both horizontally and vertically, in different settings such as dedicated poles or existing utility ones like lamp-posts, rooftops and street furniture. It must be far more unobtrusive than macrocell counterparts and ideally either invisible or at least blending into the background.
There are also different financial considerations, with increasing pressure and incentive to share equipment under a neutral host model, whereby a third party pays for the small cells and rents access. Then multiple MNOs can come in, using different frequency bands that may in turn be accumulated via multiband carrier aggregation. Needless to say, this crowded arrangement both physically and spectrally requires strong signal processing, in turn raising the bar technically.
These hurdles were anticipated in the planning and standardization of 5G, especially in 3GPP’s 5G NR Non-Standalone (NSA) technology in Release 15, the first including 5G. NSA allow operators to deploy initial 5G services around existing 4G/LTE infrastructure, by upgrading with Massive MIMO technology, although this does not allow use of the new millimeter wave spectrum. Many operators have followed this line and confined their opening 5G services to the midrange sub-6 GHz frequencies, where MIMO is particularly effective. At the same time, many of these MNOs have expanded their spectrum in the midband by up to 120 MHz or so, while also including some unlicensed spectrum up to 5.9 GHz for LTE-LAA (Licensed Assisted Access, which allows 4G to run in the unlicensed bands provided there is an anchor network in licensed spectrum).
But over time operators will require new cells at greater densities to move up into the mmWave bands and exploit their higher capacities. A few have decided to jump straight there, such as AT&T and Verizon, which lacked access to midband airwaves and so made immediate use of their mm Wave spectrum to offer consumer broadband services in selected metropolitan areas.
Then there is the in-building aspect, where operators hope that 5G will at last overcome cellular’s great deficit to WiFi. While the contention between cellular and WiFi should diminish, operators are naturally keen to have the ability to offer ubiquitous services without necessarily having to partner with providers of WiFi or fixed broadband. They also aim to target larger campus type locations, including stadia, hospitals, office buildings and hotels, where there is a desire for ubiquitous services including indoors but not always reaching every single spot.
Therefore, unlike the situation for many WiFi deployments, these outdoor/indoor 5G networks do not need to be ubiquitous. Indeed, the cost of deploying small cells everywhere in such locations would be prohibitive and not justified by the revenues generated. This was one of the motivations for development of distributed RAN (D-RAN) as a cost-effective compromise, enabling a sparser network of passive components using low power radios to be set up. This comprises interface points with several outlets enabling flexible networks that can be recognized as required to meet varying capacity and coverage requirements, for example to cater for exhibitions and other events.
Then there is the backhaul question, which is forcing some MNOs into new partnerships, as well encouraging deployment of fronthaul for increased flexibility. Fronthaul, where the baseband units are separated from the remote radio heads, allows the former to be pooled centrally. This again reduces footprint and increases flexibility, as operators can dynamically configure multiple radio sites and therefore respond faster to changes in demand for capacity.
There is also some scope for introducing mmWave at existing LTE sites, on a co-siting basis, but again this does not address density demands. It can also fail to provide the same coverage at mmWave frequencies, for example in urban canyons where streets are hemmed in by tall buildings and line of sight is effectively required. So most operators will find themselves having to add sites dedicated just to mWave.
Small cells can be fronthauled over wireless mesh configurations and there is consideration of allocating some of the mm Wave spectrum to that, which could be cost effective where not all of that is required for the RAN. There is also the possibility of tapping small cells into cable TV networks, which some operators are doing, either through normal commercial arrangement or deeper partnerships.
In the UK, Liberty Global’s Virgin Media cable TV operation plans to merge with Telefónica O2’s cellular service with a view to developing converged broadband, TV and mobile services. Virgin Media’s cable TV network has sufficient capacity to backhaul or even fronthaul a lot of the O2 traffic.
There are also other factors in the mix, one being the lingering concerns over possible health impacts from exposure to the higher mmWave frequencies, particularly skin cancers and eye damage. This has been simmering ever since the days of 2G with no conclusive evidence of any serious impacts resulting primarily from localized heating by microwave radiation, but there have been some suggestions mmWave frequencies could cause oxidative damage to cells and tissues. While this does not look like delaying 5G deployments at present, it is smouldering in the background and could erupt if any more definitive links between the radiation and health emerge.