Regulators around the world are under pressure to make spectrum available for 5G. The core 5G band is considered to be 3.4 to 4.2GHz, and most countries have either made the most of this available band or are planning to do so.
This may be good news for mobile subscribers, but is less appealing to the incumbents – those already using these bands. Incumbents include satellite users, fixed links and in many countries, especially those in Asia and Africa, those implementing fixed wireless access (FWA) solutions.
For FWA operators, there is a serious risk of losing their entire network and for their customers, there may be no other alternative to provide home and office broadband.
Displaced incumbents need to go elsewhere, but where?
The move to higher frequency bands
5G presents a host of challenges for wireless operators. While it is intended to support faster mobile broadband speeds and lower latencies new and improved services will require wireless operators to have access to larger amounts of spectrum in order to deliver new services.
Spectrum is a scarce resource and consequently operators are likely going to have use a mix of low-band, mid-band and high-band spectrum to deliver the experiences being promised by 5G.
Incumbents will need to look at moving their networks to higher frequency bands, as the bands below 3GHz are already intensively used. There are a range of candidates, and some of the main bands for FWA solutions are in 5, 6, 10, 12, 24 to 30, 37 and 60GHz. However as always, there is a trade-off that will need to be considered. For example, higher frequencies have greater bandwidth, but less coverage.
For an operator moving from 3GHz, costs can be minimised if the same base station locations can be re-used. This implies frequencies need to be as close to 3GHz as possible in order to match the propagation.
However, operators may also be finding that their networks are becoming congested as user bandwidth requirements grow. Due to this, there may be benefit in finding bands that can deliver substantial increases in capacity.
Spectrum above 6GHz
Spectrum above 6GHz will be needed for 5G services, such as ultra-high-speed mobile broadband, without which it won’t be possible to deliver the faster data speeds being promised by service providers.
However, while it might appear that 5/6GHz is the most obvious choice, being closest to 3GHz, 5/6GHz is unlicensed and is widely used for Wi-Fi. This results in low power limits and potential interference for FWA operators, making the delivery of a stable service difficult and resulting in relatively short range.
The next highest frequency is 10GHz, which is a much more viable alternative. This band is already widely used for FWA and fixed links, so there is a range of low-cost equipment available alongside greater experience in deployment and operation. The bandwidth is substantial, in some cases 500MHz or more, and the band is often less congested.
The band 24-30GHz is also tempting for operators. It is designated for 5G and is becoming available in some countries, opening the door to a wave of opportunities. It has even more bandwidth than lower bands, however there are still down-sides for operators migrating from 3GHz that should be considered.
Primarily it has a shorter range and it runs the risk that operators might be displaced yet again as regulators look to auction the band for 5G. Where there is more certainty as to spectrum ownership, then it should definitely be considered alongside 10GHz.
Above these bands the range becomes so short that a 3GHz operator would likely need a complete network re-design. Due to this, 10GHz or 24-30GHz are the most likely candidates for migration.
Once a decision is made on the new frequency band, network planning can start. The first test is to determine what would happen if just the existing base stations are used. If this still allows for the vast majority of subscribers to be covered, then the migration is relatively simple and cost-effective. For the few that cannot be reached, a decision can be made as to whether additional base stations or repeaters are worthwhile or whether the service should reluctantly be discontinued.
There are various approaches that can be used to help deliver this outcome. For example, bandwidth can always be traded off against range. If there is more bandwidth at 10GHz than 3GHz, which is likely, then lower modulation levels can be used at 10GHz allowing a better link budget and greater range. This might balance the loss due to higher frequencies.
Alternatively, higher-gain dishes at the subscriber terminals can deliver sufficient enhancement in the link budget. Moving to more sectors is also an option, for example from 3GHz to 6GHZ, as this can deliver greater range through more directional antennas at the base stations.
However even using these approaches, there are often significant areas where coverage is poor or lacking. These will typically require in-fill base stations – locations selected to cover these specific areas. These are often lower height and lower power than existing base stations, as they may only need to have a relatively short range. Minimising the number of these will reduce the largest cost element of the migration.
Getting it right first time
There is a complex re-design process with a large number of variables, including extended coverage, more base stations, trading bandwidth against range, disconnecting a few subscribers or potentially enabling new ones to be connected.
When it comes to spectrum, it has a key role to play in the type of 5G service that is made available to customers. However, high-band spectrum will only be able to provide high speeds and lots of capacity if the users are located close to a cell site, and while using lower bands may provide better levels of coverage when it comes to levels of performance it may be little better that 4G.
Working through all of this will require propagation planning tools, network models and cost models, all leading to a range of scenarios from which the highest net present value (NPV) scenario can be selected. The migration process will also need care to ensure downtime for subscribers is minimised as much as possible, as well as to limit the inconvenience of needing to visit their home to install new equipment while not requiring substantially more staff.
Getting the migration wrong could be very costly, and few operators will have the skills to do this themselves.
However, quality equipment vendors will have this capability and operators should maximise the in-depth experience vendors have in conducting similar exercises across the world.
Migrating may be a costly and time-consuming change in the short-term, but it could lead to a whole host of new opportunities with a network with increased bandwidth, up-to-date equipment and the ability to add additional spectrum in higher bands as it becomes available.
Author details: William Webb is CTO at Cambridge Broadband Networks Group (CBNG)