2019 Brooklyn 5G Summit Looks at What’s Next

The sixth annual Brooklyn 5G Summit (B5GS), organized by both Nokia and the New York University (NYU) Wireless research center at NYU’s Tandon School of Engineering, has evolved into one of the most preeminent conferences on 5G since its inception six years ago. This year’s Brooklyn 5G Summit saw about 300 physical attendees and a further 1,600 people viewing a live stream of the event thanks to IEEE.tv. Today NYU Wireless counts on a range of commercial partner companies, including Nokia, Qualcomm, AT&T, Sprint, Ericsson, Huawei, InterDigital, Crown Castle, and others.

Building on the themes of past summits, the theme for this year’s B5GS was “The First Year of the 5G Era” and saw sessions centered around key issues such as 5G New Radio (5G-NR) and core and network slicing as well as emerging technologies such as machine learning, Terahertz (THz) communication, nonterrestrial networks, ultrareliable low latency, and private networks. These advanced technologies are active research topics in academia right now, and based on previous history, it will take 14–15 years to become commercial in beyond-5G or 6G communications in the 2035 timeframe. We may well be witnessing the year zero of 6G.

The agenda was very rich, and we have decided to focus on the key takeaways centered around the roadmap for 5G and its long-range extension out to 2035.

The benefits of today’s Rel15 include ultralean design, multi-antenna support, forward compatibility, wide spectrum range, and low latency. Rel16 covers enhancements not only for 5G-NR but also for Long-Term Evolution (LTE). LTE evolves in parallel and Rel16 is planned to deploy commercially in 2021–2022; it will enhance 5G with the addition of work items or features, including Integrated Access and Backhaul (IAB) that is useful for those situations where the number of access points exceeds the number of fiber drops. IAB will be most useful in outdoor, small-cell relay nodes targeting Fixed Wireless Access (FWA) and enhanced Mobile Broadband (eMBB) deployment scenarios.

Also planned for Rel16 is New Radio-Unlicensed (NR-U), an evolution of MulteFire for 5G in unlicensed spectrum with support for various deployment use cases, including carrier aggregation NR and NR-U, dual connectivity LTE/NR and NR-U, and standalone NR-U. Vehicle-to-everything (V2X) (3GPP V2X Phase 3) with NR-V2X addresses new use cases such as platooning, advanced driving, and sensor sharing. Time-Sensitive Networking (TSN)—an essential capability for deterministic latency in industrial networks (a major emerging use case)—is also planned for Rel16 as industrial-grade NR for the Industrial Internet of Things (IIoT).Tropospheric ducting, which occurs in certain weather conditions, can cause base-station-to-base-station interference over large distances and will be tackled in Rel16 by the NR Remote Interference Management (RIM) and cross-link interference framework.

Topics under discussion for Rel17 (planned to deploy commercially starting in 2022–2023) include operation in spectrum above 52.6 Gigahertz (GHz) for eMBB and FWA; unlicensed operation in 5–6 GHz and 60 GHz; further enhancements for IAB; Multiple Input, Multiple Output (MIMO) antennas for distributed massive MIMO; delay tolerance for drone operations; public safety; nonterrestrial access; Self-Organizing Networks/Minimization of Drive Test (SON/MDT); and techniques for low cost and low power in 5G Machine Type Communication (MTC). Sideline or Device-to-Device (D2D) communication without using the base station will also be considered for Rel17.

For beyond Rel17, AI/ML and THz communications are among the topics being researched today. At the B5GS, Stanford University discussed AI/ML in communications systems and concluded that it is a powerful tool in the system designers’ toolbox; they cautioned that these techniques must be carefully deployed. AI/ML will be useful where channel parameters do not exist or where the complexity of existing techniques is high. Significant performance gains can be had when AI/ML is carefully deployed. For example, in the Radio Access Network (RAN), AI/ML is superior to conventional channel models where the channel state information is not perfect, which is the case in the real world. AI/ML also compensates for channel variations. This research has practical applications in the area of massive MIMO, and we outline this in our executive foresight (The Massive MIMO Innovation Pipeline) that includes a discussion on intelligent massive MIMO.

Judging by previous history, academia starts researching technologies 14 or 15 years before a mobile generation’s commercial deployment. In consecutive sessions, Gerhard Fettweis, a professor at TU Dresden, and Ted Rappaport, the founder and director of NYU Wireless, discussed the potential of communications using the THz spectrum. This research has begun thanks to the recent U.S. Federal Communications Commission’s Spectrum Horizons decision for experiments from 95 GHz to 3 THz in which 21.2 GHz of unlicensed spectrum is available. NYU expects THz communications will find applications in spatial cognition, sensing, imaging, communication, and positioning. Spatial cognition will find use in robotic control and drone fleet control. Sensing is another promising field for THz communications, with uses in air quality detection, personal health monitoring, gesture detection, touchless smartphones, and explosive detection and gas sensing. Imaging applications will include such uses as “see in the dark” (or mmWave) cameras, high definition video resolution radars, and security body scans. In communications, wireless fiber for backhaul, intradevice, or chip-to-chip radio communications and connectivity in data centers are proposed. Finally, centimeter-level positioning will become possible with THz communications.

This is highly speculative, but if past trends pan out (given that the research validates the promises of THz communications), we might expect THz communications to become a commercial reality by 2035 thanks to the fundamental research starting today.

Considering that 5G is just starting to deploy commercially and that the full set of features (or true 5G) won’t become available until Rel16 or Rel17 in 2022–2023, it might seem premature to talk about beyond 5G (or even mention 6G) at this early stage. But that’s exactly what the researchers at Nokia Bell Labs, NYU, and TU Dresden are doing given the anticipated 14- or 15-year research-and-development lead time needed for 2035 deployment.

The sixth B5GS declared 2019 as year zero for 6G.