What do you get when you mix an academic, an optimistic technologist, a sceptical regulator, and the Great Telco Debate? A pragmatic view of where 5G is and where it is going.
I thought it timely to revisit the 5G debate. The expert witness comments at the Great Telco Debate 2017 from Professor Rahim Tafazolli from the University of Surrey 5G Lab, Hossein Moiin from Nokia and Clive Carter from Ofcom gave us plenty to consider on the day. Mobile World Congress gave me an additional chance to test these thoughts against the glitz and glamour of Barcelona and subsequent posturing from suppliers and mobile operators alike are bringing the 5G investment question into stark highlight.
Key threads to emerge from the debate:
- 5G is an engine, a framework for future proofing the whole network
- It will start with mobile broadband because the industry is comfortable with the business model of connectivity. The newer business models requiring interworking with the automotive, health, energy sectors will need some time for both sides to agree how to work together. It must be a win/win situation where savings of 60-80% sited around healthcare for example, would be shared between parties, benefiting all.
- 4G is a very good system. We can add more speed e.g. massive MIM) to 4G to improve its performance.
- 5G will support higher speed broadband connecting people and ‘Things’ of different capabilities, but most importantly it will support automation.
- 5G does not treat IoT in isolation, automotive in isolation or mobile in isolation, but all together in a new all-inclusive system. These can’t all have separate networks so the single network has to be programmable. The reason for virtualisation is not about customer service but about reducing the cost of supporting this ever-expanding demand on the network.
- Mission criticality and resilience are what is needed to support verticals like health and automotive but also to support the swift roll out of services contrary to traditional telecoms timescales. More important than low latency is guaranteed latency. Knowing what you have to work with when building and managing applications is vital. Take out the variability. We need more and more capacity because more and more applications include video content. The structure of the wave form and the frame size in 4G just can’t cope with this explosion of connections from low to high bandwidth and from low to high latency requirements.
- Analytics gives the underlying fabric of 5G the intelligence to build future systems to support all activities. This turns the dumb pipes into intelligent pipes that the customers need.
- We need 1 million connections per square kilometre! Where is the money to support this? ARPU is flattening or shrinking in some countries. The new money for the telcos is in automation, powered by data analytics and ultra-reliable, low latency communications. It is also important to note that data analytics and AI is not so much about what has happened in the past, but what will happen in the future and how to adapt the network around this demand. Excel spreadsheet can do the retrospective but AI/Analytics can do the future. Leveraging MEC and Analytics to have the content cached at the nearest cell site would save everyone a lot of trouble and improve performance as well as satisfaction.
- We will see islands of 5G emerging: some from the telcos, some from other parties like cities, governments, major manufacturers. It will result in a very localised, scalable network infrastructure linking closely into the activities of people and things across the network.
- The 5G Innovation Centre (5GIC): The role of Service Providers, Network Equipment Providers (NEPs), cloud providers, academia and regulators is to create an ecosystem capable of supporting the new industry.
- The more and more personalised service that services and analytics brings to the table just isn’t the model the industry has been built on in the past. We built large scale networks that did lots of things for lots of people. This stuff really matters to society, to the economy, and of course, to the individual.
- Universal service for everyone is somewhat at odds with the notion of more personalised service and making more money out of certain individuals. Targeting individuals can make the telecoms players more profitable. Great, but can you make the networks available at a uniform price to all customers?
- In economic terms, we are redistributing the pie, not creating new ones or bigger ones. From a regulator (referee’s perspective), the redistribution of the pie results in people complaining to the regulator about losing a slice or two. The industry is giving productivity back to the customer through lower prices and the Skype dividend rather than giving it to the telcos and providers of equipment and devices. There is inevitably going to be a bun fight between suppliers for a share of a fixed pie.
- 5G rollout will vary dramatically based on regional differences, control economies, political motivation and regulation
- Mobile operators can only afford 5G with additional sources of financing, income or perhaps a removal of the spectrum auction.
As the debate motion posited: The market can’t afford 5G. Well, it has to. It requires a rethink in terms of a broader investment for society as a whole. Benefits will accrue to citizens and not necessarily to the telcos. So how do governments and regulators encourage the investment? Tying it all back to economics, perhaps it’s time for a reconsideration of the value being created. The business case is built on saving money and driving operational efficiency into an industry that has gorged itself on fat margins and is now waking up to digital reality.
The Great Telco Debate 2018 on 29th November in London is open for registrations. Please contact me for speaking opportunities.