The GSM Association has finally defined what it means by 5G wireless in a major report released overnight in Europe.
The group offers eight technical objectives which it says will define 5G. They are: 1 to 10Gbps connections to end points in the field (i.e. not theoretical maximum), 1 millisecond end-to-end round trip delay (latency), 1000x bandwidth per unit area, 10 to100x number of connected devices, (Perception of) 99.999% availability, (Perception of) 100% coverage, 90% reduction in network energy usage and up to ten year battery life for low power, machine-type devices.”
Of these, the two most profound objectives are for bandwidth and latency. The bandwidth objective, a gigabit to ten gigabits as a real world end user product, speaks for itself. For all those commentators—the latest being the Australian’s Mark Day and Business Spectator’s Mark Gregory—who think that an FTTH NBN is necessary they really need to get a handle on what 5G will do to the economics of network topologies. Fibre to the distribution point—that is, a point where the core can be delivered by means of last 500m wireless or fixed access medium—will clearly be the most desirable architecture for the network in the decade to come. The NBN mixed technology model has it right.
As for latency, the GSM Association admits that its one millisecond objective is incredibly ambitious. The network architecture changes such an objective requires will have profound effects on the economics of both fixed and wireless telecommunications if they come to pass.
“Achieving the sub-1ms latency rate identified as a technical requirement for 5G necessitates a new way of thinking about how networks are structured, and will likely prove to be a significant undertaking in terms of technological development and investment in infrastructure,” it says.
This is because a key determinant of latency rates is the physical distance between the end user and the location where the content they are accessing is being served.
“Subsequently services requiring a delay time of less than 1 millisecond must have all of their content served from a physical position very close to the user’s device. Industry estimates suggest that this distance may be less than 1 kilometre, which means that any service requiring such a low latency will have to be served using content located very close to the customer, possibly at the base of every cell, including the many small cells that are predicted to be fundamental to meeting densification requirements. This will likely require a substantial uplift in CAPEX spent on infrastructure for content distribution and servers,” the GSM Association says.
In a multi-operator environment, this would equally necessitate interconnection at every base station. Roaming would also need to be revamped, with connections no longer hopping to a customer’s home country but also being able to access local servers in the visited country.
The inevitable consequence of all this is an NBN-style model for mobile networks, a scenario the GSM Association readily admits is extreme, given its report\’s audience largely resides in normative telecom industry environments and not outliers such as Australia’s.
“In the most extreme case, it would make sense for a single network infrastructure to be implemented, which would be utilised by all operators. This would mean all customers could be served by a single content source, with all interaction and interconnect with localised context also being served from that point at the base station. This would also imply that only one radio network would be built, and then shared by all operators,” the GSM Association says.
The GSM Association rightly says that such a model would considerably reduce CAPEX in the network build but would require unprecedented levels of co-operation between operators. It would also have the effect of making competitive spectrum auctions irrelevant while presenting new challenges for competition regulators.
But the pay-offs? Augmented and virtual reality. Immersive Internet. Autonomous driving with mobility. Pervasive wireless cloud. This is big.