It often seems there is a binary choice for RF planners between making macrocells even more spectrally efficient versus deploying large numbers of small cells. Techniques such as sector splitting can add extra capacity to macrocells, but are insufficient to meet growing demand. Small cells add enormous amounts of capacity, but can't always be positioned exactly where required. In some cases prospective small cell sites can be either too costly or simply not feasible.
Expanding the capacity of small cells through multi-sectors could be a useful option in some situations. We look at one way of achieving that using Luneberg lens based antenna solutions.
Multi-sector small cell base stations can be deployed in dense urban areas or sites associated with very large concentrations of people (e.g. stadiums, city centre squares, outdoor music festivals). LTE is already demonstrating very good performance with up to 15 sectors deployed at a single macro base station site. Combining multi-sectors with small cell technology promises to become very cost effective method of delivering multi-gigabit throughput for these challenging locations.
Outdoor Urban Small Cells
The total cost of an outdoor urban Small Cell is much higher than indoor small cells. Factors include site acquisition, greater difficulty to gain site access, providing power and backhaul, plus coping with the environmental temperature and weather variations.
So there is a greater incentive to make each Small Cell as productive as possible, with operators looking ideally for tri-mode products that deliver 3G/LTE and Wi-Fi at multiple frequencies. Today these are all single sector (i.e. transmitting the same signal in all directions). Huawei have included beam-forming capability in their Atomcell, so that the coverage footprint need not be the same in all directions (i.e. not circular). This helps system efficiency by reducing interference with neighbouring cells, especially macrocells. Alcatel-Lucent's Metrocells can also be positioned to direct the signal away from the macrocell (zone exclusion).
A further step would be to add a multi-sector antenna array to a Small Cell location. Each sector would be connected to a separate small cell radio board, sharing a common high capacity backhaul link. An RF lens based antenna would need to be a size suitable with street furniture; size will vary with frequency bands supported (i.e. higher frequency means smaller size). One example could handle up to 9 sectors low band and 18 sectors high band, each providing both 3G and LTE service at different frequencies. If we assume even distribution of users, this would represent a capacity increase of more than 12 times and easily providing in excess of 1Gbps throughput.
The physical format of such an installation would be much bigger than a typical small cell but considerably smaller than an equivalent macro base station deployment. The range and capacity would be phenomenal – several Gbps overall. Imagine covering a street square or busy transport hub such as a railway platform or concourse with just one or two of these.
Enterprise Small Cells
The relatively low cost of an indoor Enterprise small cell suggests to me that in most cases it is likely to be cheaper to install one or two more small cells where capacity is required. It's the more challenging situations where very high capacity density is required, or where additional cells can't be physically located that this technology could be more appropriate. Indoor stadiums, conference centres and places with very high footfall might be most relevant. Using a multi-beam antenna solution also allows for excluding coverage by switching off one or more of the beams. This can be done to help with load balancing and optimization. Candidate sites for small cell deployment that require zone exclusion for either load balancing or integration optimization will work best with a multi-beam antenna solution. Lunberg lens based multi-beam antennas provide very narrow beams to minimize the adverse effect of excluding too wide of an area due to wide beam antenna use..
The size of a Luneburg lens antenna is closely related to the frequencies used and the number of sectors required. Most small cells will be operating between 1.7GHz and 2.7GHz which reduces the antenna size needed. Weight can be a factor, with each sector contributing a little less than 3kg to the total.
An example is this Matsing 4 beam antenna with 360 degree coverage pylon shaped enclosure which can be driven from any small cell radio with external antenna capability. This weighs less than 20kg hence can still be pole mounted. A much more complex 45 sector unit weighs less than 120kg.
|Frequency||698-960 MHz||1710-2690 MHz||4000-6000MHz|
|Polarisation||Dual Slant +/- 45o||Dual Slant +/- 45o||Dual Slant +/- 45o|
|Dimensions||Base: 120cm/48" by 120cm/48"|
Fig. 1 - 9 low band beams from Matsing small cell antenna
Fig. 2 - 18 high band beams from Matsing small cell antenna
Fig. 3 Street mounted Small Cell and Matsing antenna (white globe)
Integrating into the network
Sectorisation is not a new concept for the mobile industry and using these on small cells would follow the same principles as for macrocells. With such small sector sizes, you would want to avoid areas with fast moving traffic that generates large numbers of handovers. You would also be more reliant on SON technology to configure the neighbour lists and other parameters to optimise handovers, coverage and capacity. With the addition of a switching matrix it is possible to vary the sector count with demand which will become more viable with the evolution of SON systems. Among other positive attribute when utilizing a switching matrix, by varying sector count to match network demand at each site pooled resources such as backhaul can be better utilised.
A single RF lens antenna is radio technology agnostic, and works well with 3G, LTE, LTE-Advanced and/or TDD mode. Refarming of spectrum between 3G and LTE doesn't require new hardware or parts, and may be possible remotely if the small cells support it.
LTE-Advanced includes many enhancements including augmentation that will require antenna solutions with a very wide band performance which is a corner stone of Luneburg RF lens technology. Any Luneburg lens is comprised of dielectric material which performs well up to a cutoff frequency, Matsing solutions utilize lens with cutoff frequency at 10GHz meaning all existing frequency bands are supported through single lens. Frequency specific feeds are placed around the lens to create customer matching solutions.
While the small cell industry is generally trying to deskill and simplify small cell installation to enable mass rollout and deployment, this technique is one that will require RF experts. It's intended for those special cases of super high capacity where experts will choose the best approach from each situation – choosing between the technologies of larger numbers of small cells, multi-sector small cells, macrocells and/or DAS. RF planning tools will be essential in these scenarios. Standard installation and commissioning practices can be used onsite – there is nothing out of the ordinary required to make it work.
Once commissioned, the use of SON (Self Organising Networks) software is important to fine tune and improve overall performance, making the most of both coverage and capacity.
Fig 4: Commissioning an outdoor pole mounted unit covering a busy market area
With small cells evolving into widespread deployment over the next couple of years, it may be some time before we see many combined with RF lens technology. However, this does illustrate one direction the industry could take to extend small cell capacity further and meet longer term future data capacity demands. This is particularly useful in the more challenging locations, where traditional approaches may be too costly or constrained.
Our thanks to Tony DeMarco from Matsing for explaining this technology to us.
Matsing are a sponsor of ThinkSmallCell