1. THz Spectrum Heats Up Talk of 6G

Last year we discussed the promise of huge swaths of bandwidth in the frequency frontier that ranges between 100 GHz and 1 THz, along with a blend of new possibilities waiting up there, not only in terms of bit rates, but of high-resolution positioning and imaging as well. Having published a post devoted to this issue just weeks ago, we continue to expect that this topic will become one of the big ones of 2020 and we moved it up to pole position in the list. But we recognize that this will be a process and that it will take time (and research!). The first step in this process is, in fact, the deployment of mmWave 5G components, and the ongoing trials are providing an unprecedented wealth of data on how signals propagate at these frequencies, especially outdoors. In many cases, the non-line-of-sight coverage is proving to be more robust than anticipated thanks to reflections, which is welcome news, not only for 5G itself, but further for the prospects of what lies beyond. Let’s see where we are a year from now…

2 . AI Reaches the Boonies with Federated Learning

Our prediction last year was that AI would start to roll down the hype curve, but we were wrong. Last year was the year of edge computing for everything and the hot topic of 2019, Machine Learning, was not immune to this influence. Much good work was done towards an understanding of how machine learning could be distributed to reduce the latency of inference as well as reducing the bandwidth of data moving back to the cloud for training. In addition, some folks have proposed that federated learning can occur with only encrypted data moving to the cloud so that we can learn even when we don’t know what it is we are learning! So with 2019 being the year of training at the edge of the network, we expect that 2020 will see federated learning move solidly into the handset.

3. Meta-materials, Meta-surfaces, Meta-resonators, Meta-anything

A theme that has quietly but manifestly moved into a rather central position in the discussions on the evolution of wireless communications is that of meta-materials. Broadly speaking, a metamaterial is an artificially structured material that allows controlling and influencing the flow of electromagnetic waves (or possibly of other types of waves). Metamaterials are typically arranged in repeating patterns, at scales smaller than the wavelength of the waves they seek to mold, and their properties emanate precisely from these patterns. Efforts are underway to design antennas based on meta-materials, say for massive MIMO, that could incorporate some of the signal processing functionalities directly on the antenna structures, and antennas based on meta-resonators that could be more compact and broadband than regular antennas. An even more intriguing development is that of intelligent reflecting surfaces, which could be deployed in strategic locations and reflect signals on a dynamically adjustable fashion. Taking the idea to the limit, think off a world where multipath propagation was a controllable phenomenon rather than an imposition from nature. It’s too soon to tell whether these meta-surfaces can be a transformational development, or merely an anecdotal re-incarnation of good old relays. Only time and research will tell.

4. The Cellphone is Dead. All Hail IoT!

This was our prediction last year and we are calling it. OK, maybe the cell phone isn’t exactly dead but the Financial Times ran an article in October with the headline “Apple’s wearables and services drive revenue as iPhone sags”, so if not dead it is showing its old age. Certainly 2019 was the year of IoT more than it was the foldable screen (Though that was also pretty cool) and we expect that IoT will be a hot topic for 2020, especially in the areas of Factory 4.0, V2X and enterprise. In 2019 the 3GPP standards process was all over trying to define use cases for these new, and potentially massive, applications. Note that in these cases the IoT isn’t the slave to a phone but an ecosystem all by itself with some very unique requirements that are now driving URLLC. Expect to hear a lot more as this starts to roll out in big manufacturers like Mercedes, Bosch and Foxconn.

5. Cellular is Dead. All Hail IT!

OK, we just said that to attract your attention. But 2019 was the year of IT insurgency into traditional cellular operator arenas, and we would like to point out that we did call 2019 the year of IT convergence. We expect this to continue. In particular, we expect that Edge compute will continue to be a hot topic and that the impact of ORAN’s attempts to commoditize the RAN will become clearer in 2020. After all of the hype of 2019 we are tempted to predict that 2020 will be a year of adjusting for reality in the IT convergence at the network edge.

6. The Year the Telecom World Cracked

Are we touching a third rail here? 2019 was the year that the Global ecosystem that telecoms had sat on top of for decades started to crack and governments got into the telecom regulation business in a way that we haven’t see in a long time, if ever. You know what we are talking about. This process will continue into 2020 and, needless to say, there will be winners and losers. Your humble editorial board will stick to the technology side of this and stay away from the politics and business side. But it is hard to imagine this is not a hot topic in 2020 for technology.

7. Come Fly with Me, and My Network

Driven by an emerging use of flying platforms such as unmanned aerial vehicles (UAVs), drones and unmanned balloons in future network applications and the challenges that the 6G networks exhibit, last year we noticed loads of research and development activities that demonstrate the evolution of the flying platforms as a novel architectural enabler for radio access network (RAN) and their integration with the future cellular access and backhaul/fronthaul networks. It seems that these platforms could be used as potential way to offer high data rate, high reliability and ultra-low latency access and backhaul/fronthaul to future wireless networks. Such large scale and flexibly deployable platforms and frameworks may guarantee the global information and communication requirements in future smart and resilient cities and solve the ubiquitous connectivity problems in many challenging network environments, e.g., coverage or capacity enhancements for remote or sparsely populated areas, social gathering and disaster affected scenarios, etc and responding to United Nation’s sustainable development goals and societal challenges.

8. Massive MIMO Implementation Becomes a Career Plan

Yet again typing “massive MIMO implementation” into IEEE explore once again reveals that the rise in paper density has pretty much rooflined since 2017 at around 100 papers a year. We promise not to mention it again.

9. Everyone is Slicing and Virtualizing

Globecom in December was awash with papers on RAN slicing and we expect this trend will continue to ramp for a while because the availability of open source frameworks and data is now allowing universities to finally apply some of that theory in a meaningful way. Some folk believe that this will be a trend that will fundamentally alter the landscape of the telecoms industry. Some of us older folk remember CloudRAN and NFV and get a similar feeling about this time around. But either way we will call 2020 as the year that everyone got comfortable with slicing anything that moves in the telecom network