By Gerry Foster, SON Engineering Director, and Steve Bowker, CTO, AIRCOM International, a TEOCO company
The global emergence of LTE has seen new levels of innovation being introduced in terms of mobile network architecture. The ability to add new software layers to existing infrastructure provides a cost effective way of delivering a better quality experience to users while reusing all existing network assets, wherever possible.
LTE innovation enables operators to deploy new RAN technology in a controlled, cost effective way that sweats all legacy equipment. Critically this means deploying new technology in areas of heavy use, where the benefits of faster speeds and greater capacity can be instantly monetised for optimal ROI. This is certainly true with current new LTE deployments in all major countries around the world.
While the principle of the Hetnet has been a huge success, it is becoming increasingly clear that if 5G is to become a viable reality, network operators are going to have to dispense with at least one RAN technology. This is mainly due to the cost of supporting the multiple radio bands that these different technology layers require to function.
The battle of the bands
At present, operators must typically support seven bands globally for GSM, UMTS, HSPA, LTE and LTE-A (450, 700, 900, 1800, 1900, 2100, 2200). In some regions LTE and LTE-A add three or four different bands (2.5 GHz, 2.6GHz, 3GHz and 3.5GHz). Device manufacturers must also contend supporting additional bands for Bluetooth and Wi-Fi (2.4GHz and 2.6GHz). 5G technology will add many more bands. Then there are the additional technologies like CDMA and EV-DO that sit outside of 3GPP that still require supporting in several significant global regions.
The reality is that at least one RAN technology is likely to be switched off before 5G is introduced. We have already seen operators re-farm 3G spectrum to better support LTE deployment. This migration is likely to continue as LTE can offer better spectral efficiency as well as faster and richer mobile services. It therefore seems likely that the arrival of 5G will signal the death of 3G in many global territories.
The technical and financial reality
From a technical perspective it is possible to deliver up to eight bands simultaneously. However, simple network economics means that operators and OEMs are unable to support more than six different bands to deliver cost effective services. The early standards that 5G will be based on centre around the enablement of carrier aggregation – the ability to combine multiple disparate bands for new levels of spectral efficiency and to unlock extra capacity. This will be compromised if operators intend to juggle myriad spectrum bands.
Even disregarding carrier aggregation, the sheer cost of having to manage five network technologies and systems concurrently will have most operator CFOs and CTOs worried. The respective handovers between these multiple bands will all need to be optimised for best possible quality of service and experience. This involves a great deal of technical complexity, especially when also addressing the signalling overheads that these multiple radio access technologies will present.
Work to do now
As with LTE, higher data rates will not provide enough of a pull to encourage 5G take up alone. Subscribers will migrate to 5G if it offers a better mobile data experience in more places thanks to better coverage. This includes the provision of seamless offload to other technologies (including WiFi and shared public small cells) and better access to data thanks to cloud servers working in conjunction with mobile and WiFi channels.
Less than half of the world’s antennas are even RET tiltable, less than ten per cent are RAS steerable and there are very few MIMO systems in operation
The mobile industry must also actively invest in developing antenna technology. The ability to automatically tilt an antenna as part of a broader SON architecture is crucial to making 5G (and indeed LTE-A) a viable technology. Less than half of the world’s antennas are even RET tiltable, less than ten per cent are RAS steerable and there are very few MIMO systems in operation. This will need to change if operators are to realise the instant, real-time configuration enhancements that user-centric 5G networks will need to become a reality.
Failing to plan is planning to fail
The added complexity linked to managing multiple technologies and bands will herald new network planning techniques for 5G technology. Advancements in data analytics, geolocation technology, SON and CEM systems will lead to a more detailed view of user behaviour – not just at a network level, but also at a subscriber level. The placement of base stations and small cells will not be determined by network-centric analysis, but by the customer lifetime value of the users impacted by dropped calls, data buffering or poor download speeds in any given area.
This enhanced view will have a profound impact on an operator’s ability to maximise profitability by taking a business view of network performance and customer experience. Such a detailed perspective will tempt operator marketing executives to consider either monetising this generalised view to third parties, or pick up the mantle themselves and deliver innovative new offers and services to drive increased revenues and longer-term retention. The mobile industry is well on the way to achieving this reality, with much of this capability and functionality possible today.