On 26 September 2019 over 150 senior engineers and researchers assembled for the CW Technology & Engineering Conference to dive into the application of Massive MIMO, Satellites and other new technologies that are transforming the networks, now and in the future.
What has Massive MIMO ever done for us?
BT Technology outlined the performance, challenges and evolution of active antenna systems in RAN while also educating the audience on how exactly Massive MIMO works. Anvar described the process as not magic at all, but simply following the fundamental principle of constructive / destructive addition of wave fronts:
- Antenna elements redistribute input signal energy as electromagnetic radiation
- Antenna arrays adjust the superposition of the electromagnetic wave fronts
- Baseband processors control antenna arrays, for example by pointing beams towards users.
The benefit is clear; their company’s R&D trials demonstrate over double the spectral efficiency in certain active antenna systems compared to 4x4 MIMO, and about six times the efficiency of 64QAM Peak. However, their actual performance depends heavily on their operating environment. When challenged by the audience on whether Massive MIMO is appropriate for all use cases, it was agreed that in certain geographies – such as those with ultra dense urban populations – Massive MIMO did not deliver performance significant enough to warrant the extra cost.
Three has planned 6,000 Massive MIMO sites across the UK which they believe will carry 80% of Three’s traffic and offer a 3-5x increase in capacity. They are delivering this in the UK in partnership with Huawei, with whom they already had a contract for delivering 4T4R for their 4G network. Such a roll-out does come with challenges; the company is working out how it can strengthen masts and increase the amount of power to sites. But early trials are seeing stats such as 9ms latency in eMBB applications. They are excited by the significance of this for consumer gaming and other high bandwidth entertainment or communication applications.
Chronos Technology pointed out to delegates that one of the complexities of Massive MIMO was that the electronics and power consumption had moved to the top of the mast where oscillators are more exposed to extremes of temperature compared to typical cabinet housing. While the delivery of microsecond accurate frequency, phasing and timing is fundamental to the efficient and reliable operation of 5GNR, the new network architecture is presenting challenges.
Non-Terrestrial & Hybrid Networks
The opportunities and challenges of Non-terrestrial networks (NTN) in 5G were laid out clearly by platinum sponsors CGI, who are building network management systems for satellite operators, at the start of the session after the morning coffee. For example, on top of the clear benefit of connecting rural and hard to reach areas, the estimated revenue potential of connecting those without cellular coverage is $20-30Bn per year; while that is only about a 2-3% increase in current global cellular revenues, it’s a significant amount. Non-terrestrial networks could improve the security of the network by offering flexible redundant routing and security overlays; and it can improve reliability by offering a “drop-in” solution to expand the capacity of existing cell sites as needed (for example, for live events).
Latency had been considered a challenge, but with LEO offering a 20ms round trip time (excluding switching and terrestrial routing) compared to the typical 4G round trip time of 30-40ms, LEO satellites could still meet user’s needs. With intelligent traffic based routing and pooling, both LEOs and GEOs could meet the bandwidth needs of 5G. However, the price of installation is still large, even when compared to the fibre provisioning cost of up t $50-100k/km in certain locations, and the break-even point for very high throughout satellites are $4-5k/Gbps/month compared to $0.5-1k/Gbps/month for a 1Gb fibre lease in the UK.
From CGI’s perspective, the technical capability for hybrid networks is emerging, the challenge is deploying it at a cost-effective level. For this, further technology development is needed.
Yinan Qi from Samsung Electronics followed with a breakneck tour through what the 3GPP is doing to support satellite communications in 5G. Rel.15 held the first study item on NR to support NTN. It studied the channel models, defined the deployment scenarios as well as the related system parameters and identify any key impact area s that may need further evaluations. In Rel.16, the study item of funding solutions for new radio to support non-terrestrial networks has been approved. In detail, this will explore physical layer control procedures, uplink timing and making retransmission mechanisms at the physical layer more delay-tolerant. For Layer 2 and above, and the RAN architecture it will study propagation delay, handovers between non0terrestrial space-borne vehicles that move at high speed but other predictable paths and dual connectivity.
Avanti Communications was the highlight of the track. This satellite operator enables MNOs to offer effective 2G, 3G and 4G LTE coverage in areas where backhaul is unavailable. It is trusted by the UK Government to deliver on the emergency services network program with EE and is involved in an array of 5G programmes. One of these SaT5G, is well into its project life and is tasked with building compelling business cases and commercial models for eMBB, primarily backhaul, and researching the key technology enablers for the integration of satellite into 5G. The four use cases it has settled on is edge delivery and offload for multimedia content, 5G fixed backhaul, 5G to premises and moving platform backhaul. Simon Watts from Avanti closed with some personal comments on where he thinks the industry is going. In his view, in two to three years we should see the edge delivery of cached content, streamed content, video, A/VR, games etc over GEOs. Further ahead (five-ten years) we will see LEOs providing satellite 5G RAN direct to consumer equipment.
5G: a catalyst for network transformation
OpenRAN, site management and an operator’s perspective on network transformation were covered in the track after lunch.
Parallel Wireless opened with coverage on why they believe OpenRAN is the future of the networks. The ongoing disaggregation of mobile network architectures started with 3GPP’s separation of the control and user plane in Rel-14 and the emergence of the service-based architecture. With access networks moving from proprietary hardware and software in both the remote radio unit and the baseband unit to off-the-shelf solutions, it makes sense for the interfaces to similarly move from being proprietary to being open. In this way, a network operator could replace hardware with that of any other OEM and have it work seamlessly with the software, or replace the software with that of any vendor with similar results. It increases flexibility for network operators and lowers the costs of deployment by avoiding vendor lock-in. RAN is 60% of Capex and Opex, so is the most important cost factor to focus on in telecom. A significant milestone in the OpenRAN journey came in June this year when Orange revealed a breakthrough, having successfully completed an implementation test of the X2 interface between equipment from Nokia and Ericsson.
Cost reduction was a key feature of Sitetracker’s talk as well. They support organisations to successfully deploy and monitor critical infrastructure. From the point of view of their customers, who include Verizon and T-Mobile, project volume, variety and velocity are accelerating in telecoms; macro sites are moving aside for small cells, indoor DAS and stealth sites. This means an increased focus on operational efficiency as network providers deploy more assets with fewer resources. For Verizon, Sitetracker managed to save 300hrs per month through their reporting and forecasting tool.
For Vodafone, network transformation is a many-headed beast that must be handled alongside changing consumer demand, emergent technology enablers and operational / cultural changes throughout the ecosystem. For them, the future relies on network sharing, digital operations (zero touch operations and predictive solutions), programmable networks and global service platforms. Their teams are working to move from a siloed, monolithic architecture and manual processors to something more akin to a Network as a Service, with automation across customer service lifecycle management, network services lifecycle management, resources use optimisation and more. In terms of new applications, they are hoping to deploy cost effective low latency capability on existing 4G bands and on the new 5G NR.
Getting ready for the beyond 5G-era
The key takeaway from this session was that it feels like we are already prepared for the beyond-5G era. At least, there are sufficient concrete plans and concept ideas to take telecommunications through the next 5-10 years. And in fact, as Alan Carlton from InterDigital pointed out, the Beyond-5G era will start rather soon in 2020 with Rel.17 and Rel.18. These will look at NR-Lite and coverage extensions, 5G NR for >52.6GHz, AI/ML network enablement and – with great significance for the earlier session – initial support of non-terrestrial networks.
But, what is technology without an application? Beyond-5G communications are being envisioned as useful for holographic communications, autonomous systems and robots, self-sustaining networks and extreme industrial internet of everything applications. The key technologies that will play into the evolution are extreme edge computing, the move to above 150GHz, AI / ML at every layer of the stack and fusion with other technology like blockchain and quantum. The technology areas that need big innovation to pull off this change are circuits and devices, radio transceivers, radio systems, network protocols and data intelligence.
Dean Bubley from Disruptive Analysis cast a different light on what the telecommunications industry might need to consider for 6G. Rather than more and faster, he suggested that we needed to consider wider societal movements such as increased awareness of climate change and factor power reduction into the next generation as a priority. Still on the theme of climate change, he recognised that the industry will increasingly struggle to deliver reliable coverage if weather patterns continue to change. Looking well beyond 5G, spectrum is an area that could be revolutionised. The evolution has already been from basic spectrum sharing to more dynamic sharing; the next step could be offering enterprises network-as-a-service: simply purchase the equipment you need, and you get the spectrum required to operate it as well.
CW sends our thanks to our speakers who shared their ideas and experience; to our sponsors without whom the event wouldn’t happen; and the tireless CW Technology & Engineering Conference Committee who pulled together and chaired such an enjoyable day.
If you would like to be a part of CW Technology & Engineering Conference 2020, register your interest.