A common mantra in the telecoms sector is that 5G and IoT are two sides of the same evolutionary coin – that 5G will enable profitable IoT apps, and that IoT is the chief business case that will justify 5G investment. And yet LTE will still exist, while certain IoT apps will thrive without 5G via LPWA. Which raises the question: Is the link between 5G and IoT/M2M really as inextricable as the hype machine claims?
This question kicked off a lively panel discussion at the CommunicAsia2017 Summit. Here is what the panelists had to say.
Better economics and low latency
Andrew Scott, director of technology, Telstra: It is lower battery consumption and scale that 5G can do and LTE cannot. Scale and signaling, and the ability for user plane separation means that the networks can be so much more economical.
Low latency will mean autonomous vehicles and remote operations in the healthcare industry. VR needs 1ms latency to prevent motion sickness. VR feeds into healthcare, as a surgeon can perform an operation in VR at 1ms latency but not at 20ms. But the QoS of the surgeon must be guaranteed and not decimated when someone starts streaming Netflix.
5G will be so much more flexible. In the past, we had mobile internet and broadcast. With 5G you can have reliability, massive gigabit connectivity or massive deployment. You can pick and choose the parameters for what you want. The question, then, is whether regulators will see each mix as a different category or different flavors of the same?
Remember: 5G isn’t replacing LTE
Sandeep Girotra, senior VP of APAC and Japan at Nokia: As an industry, we are accustomed to selling B2C. Eventually this will be B2B2X – telcos will sell services to a company that will sell to an end user. In my opinion 5G is more about the business model.
4G to 5G is not a flick of a switch that will happen on Day 1. 5G will become mainstream around 2020 but between now and then the [LTE] evolution is not stopping. Today we can already do a couple of hundred megabits per second with 10-15ms latency. With 4.5G or 4.9G that latency goes down to 10ms and with 5G it will be 1ms.
The question is who will make the money? That is not clear and is different in each vertical. This panel has already mentioned about 50 different use cases, and no telco can do all 50. So who will make the money? Will it be some sort of OTT?
New partnerships and regulations
Robert Le Busque, Regional MD of Australia, New Zealand and India at Verizon Enterprise Solutions: 5G’s characteristic of low latency will lend itself to real-time apps and ones that are not just point-to-point. Autonomous vehicles and other real-time apps will be the applications that really drive the infrastructure, whether it is fiber or 5G. Whether it is autonomous vehicles or some sort of social infrastructure, the partnerships for 5G will need to be very different than the ones forged seven to eight years ago.
One problem is that 5G use cases are outside of the current regulatory framework. Some regulators say we do not need a new regulatory framework for 5G, others say we do. The answer is probably somewhere in the middle. The question is how is this regulation is going to affect innovation in 5G?
Smart cities enabled by eMTC
Dr Abdul Memon, director of marketing at Huawei South Pacific: It was because eMTC was delayed so much that LoRA and Sigfox took market share in the early stages. If we look at the applications it is mainly about battery and power consumption. In some cities, we see smart sensors only around poles because they need power from those poles. Battery intensive applications were not there in NB-IoT, and it is eMTC that will enable us to get to smart cities and smart meters and realize the huge cost savings possible.
No one company can make a IoT solution. It has to be alliances. We need an ecosystem program to help developers to NB-IoT an eMTC.
LTE-based IoT isn’t profitable IoT
Andrew Mackay, head of mobile solutions APAC at Cisco: We must remember that LTE stands for ‘long term evolution’ LTE will continue to evolve in the 5G era. Features such as network slicing will come to LTE. Where will LTE stop and 5G start? The 3GPP says that the initial version of 5G will be non-standalone and use an LTE core. That will be early 5G. In the 5G era a lot of the features will come through in Evolved LTE. New radio and slicing in particular will enable ultra-reliable communications.
5G beamforming and massive MIMO will enable connectivity for drones in the sky. Traditional networks have been designed for people on the ground. Up there a drone can see up to 18 cell sites which can lead to a huge amount of interference.
There is also the issue of 5-year-plus battery life for IoT sensors. This means that the network still needs to be working five years later. That means three, four or even five 3GPP specification releases later.
The 5G use cases will mean revenue of 10 cents or less per month per device. On LTE that cannot be profitable, and you need to think how you can monetize and provide more services to end users.
Security is clearly a problem for IoT devices, so it will be up to the network to detect and isolate compromised devices as close as possible to the edge and at least reroute their attacks via DNS.