What’s the story - Network or GPS synchronisation?

Relentless data growth is driving spectrum exhaustion; forcing MNOs to re-think their 4G-network strategy to introduce TDD and the advanced features of LTE-A to keep pace. Then there’s 5G to prepare for. Time & phase synchronisation needs to be an integral part of your architecture.

4G has been live in the UK for five years now. In that time there’s been mass adoption by all but the lightest users of mobile services. Moving traffic to LTE spectrum gave MNOs some breathing space for a while, relieving congestion on their 3G networks and allowing some spectrum re-farming. But LTE also unlocked latent demand, and gave consumers a real broadband experience for the first time, resulting in a big change in user behaviour – the mobile device has become integral to the modern lifestyle. The result is that MNOs are once again up against it on spectrum capacity.

There are two things you can do to relieve spectrum congestion. One is to throw more spectrum at it, the other is to make more efficient use of the spectrum you have. Rolling out new spectrum has been hampered in the UK by the ongoing delays to the auction of the 2.3GHz and 3.4GHz bands. MNOs are showing an increased interest in adopting the advanced features of LTE-A, which improve cell edge coordination and can increase spectral efficiency by up to 80%.

These features all rely on accurate time and phase synchronisation. As do spectrum bands licenced for Time Division Duplex (TDD) use only, which includes the new bands and some of the frequencies of existing spectrum holdings. Adjacent mobile cells must be aligned in absolute time and phase to be able to coordinate the transmission and reception of data from multiple cells to the end device, or to coordinate transmission and reception time slots in TDD. Without adequate synchronisation, these features simply won’t work, and massive interference can occur, which, in turn, will have a negative impact on customer experience.

So the industry is in agreement that time and phase synchronisation is needed, the big question remains on the optimal build. Initially, MNOs will need it only in dense urban areas, where demand is highest and performance is most constrained by limited spectrum. One might be tempted to take a tactical approach for limited deployment. But isn’t that just putting off the inevitable? Over time, the requirement will extend over the whole network. Shouldn’t we fix the roof while the sun is shining, not wait for the downpour that’s coming? MNOs need to make a strategic decision on how to enable time and phase synchronisation network wide and with the least risk of disruption in-life.

The strategic direction to achieve this time and phase alignment is open to debate. Firstly, there are network based time and phase synchronisation solutions available. Using a network-based universal time source will bring a number of key benefits to MNOs:

• An extremely reliable and secure timing solution which has been engineered to always meet the performance requirement demanded by the standards
• Economies of scale achieved through a “build once, consume by all” model
• Easier transitions, changes and upgrades, including everything from LTE-A requirements to more stringent requirements for 5G ultra-reliable networks
• Extension to the potential scale and geographical reach, whilst also simplifying the provision process – negating the need for individual site surveys and mitigation of no line of sight
• Fewer points of maintenance and failure are added to the network, significantly reducing outage time and the cost to maintain, align and upgrade via site visits as required (in comparison to GPS antennas that need to be re-aligned every 18 months)
• Finally and most importantly a network-based solution on a common time source is the only solution that will guarantee that, in a GPS failure, cells will stay in sync with every other cell on that timing domain – avoiding the loss of significant coverage and capacity.

As a second option, MNOs may choose to deploy GPS/GNSS solutions to each site. To date, GPS has been a strong technology due to the simplicity and incremental cost to deploy on a single basis, but does it scale? As an industry in the medium term, we need to think through some further considerations with GPS:

• For network critical applications is it a sufficiently robust approach?
• Is it economic to deploy network wide?
• How does it scale for small cells?
• How do we manage, service and maintain, tens of thousands of GPS sources from an operational perspective in 5G?
• How do we manage potential disruption of the signal as evidenced by past GPS failures caused by software issues, or on a local scale by accidental or deliberate jamming?
• And lastly, how do we manage complex operational SLAs in the LTE-A, 5G world?

All of these considerations need to be carefully and thoughtfully planned before embarking on a mass GPS deployment within a mobile network domain.

MNOs are also contemplating how to build the architecture for low latency and high coordination going forward. It’s likely that they will start to look to points of interconnect for local groups of cell sites in order to realise a C-RAN design. Many of those locations will be telephone exchanges, and they will require time and phase synchronisation. It makes sense to start with an architecture that enables those locations by default, and can easily deliver the service to existing cell sites in the process.

As the network demand grows so will demand for phase based timing. Having a plan that meets not only your short term 4G timing requirements but also provides a strong evolution path to 5G will be key. With the challenges of ultra-reliable networks now being planned by industry, the challenge of choosing the correct timing solution is now open for debate.

If you want to find out more about BT’s time and phase synchronisation solutions, please download our latest whitepaper.