It’s widely understood that 5G networks will support many applications. On the consumer side, it is expected to offer speedier communications bringing about enhanced mobile broadband, UHD and 3D video streaming, consumer cloud services, autonomous vehicles, tele-medicine and education, augmented and virtual reality. On the infrastructure side, it will support Internet of Things, Machine-to-Machine and Enterprise communications, healthcare facilities, smart city urban infrastructure and more.
That said, it’s clear that in order to achieve the required broad connectivity, technologies in addition to terrestrial International Mobile Telecommunications (IMT) are required. Further, each service and application has its specific requirements. While ubiquitous connectivity requires national/global coverage, autonomous vehicles need low-latency and highly reliable connectivity. Augmented and Virtual Reality (AR/VR) applications will need high bandwidth, low-latency communications for safety and user experience, while massive Machine Type Communications (mMTC) will require millions of low data rate connections, amounting to significant bandwidth needs. The list of requirements goes on.
The recent World Radio Conference (WRC-19) held in Egypt this past November went some way to recognizing all of these requirements and began carving out territory for many of these purposes. In addition to identifying 17.25 GHz of spectrum for terrestrial 5G, the little-known radio planning summit promoted satellite innovation in the Ka-band approving a further 4 GHz of spectrum to support broadband connectivity to airplanes, ships and land mobile platforms, such as trains. It also classified more than 6 GHz of spectrum for the provision of fixed broadband services via high altitude platform stations (HAPS), and an additional Radio Local Area Network (RLAN) spectrum for Wi-Fi. Looking ahead to WRC-23, the International Telecommunication Union (ITU) will seek additional spectrum for terrestrial IMT and broadband communications on the move, over satellite and HAPS.
However, the fact that the WRC identified spectrum bands for particular services and applications is only part of the regulatory requirement to make 5G a success. National regulatory authorities must take on the work of making frequencies available in the most appropriate way while protecting other services. The mid-band and mmWave frequencies identified for 5G do not travel as far as the lower band frequencies used for 2G, 3G and 4G services. This means that building nationwide 5G networks using these frequencies is likely to be prohibitively expensive for even the wealthiest of mobile network operators. On the plus side, these higher frequencies are much better suited to providing excellent quality, localized, high-density, high bandwidth broadband.
This move toward localization of service means that the tendency of national administrations to auction off blocks of spectrum for nationwide licensees will no longer be appropriate. Moreover, it’s certainly not appropriate to auction off bands not identified for IMT in the ITU Radio Regulations. Such action damages investment in new communications infrastructures and jeopardises the ability of non-terrestrial infrastructures such as high throughput satellites to deliver broadband connectivity in parts of the world not touched by terrestrial 5G networks. Administrations will need to adopt regulatory frameworks promoting spectrum sharing among co-frequency services and applications. This can be accommodated through approaches based on geographic and time usage and through different licensing approaches, blanket national licenses, unlicensed access to bolstering innovation, or tiered licensed access. A tiered licensed approach could, for example, include licensed protection of incumbents alongside secondary licensees accessing spectrum on an opportunistic localised or time dependent basis, alongside unlicensed users. In reality, the demands of different services and applications will mean that different licensing solutions are appropriate for different frequency bands. We consider that the adoption of technology to solve spectrum management challenges is fundamental to relieving spectrum congestion in high-demand frequency bands. The adoption of technologies, including geolocation databases and spectrum sensing, to better understand the frequency environment will be fundamental to this. In turn, this could challenge existing laws relating to data privacy, and could raise questions such as whether these databases should be public or private.
Once they’ve established how spectrum may be shared from a radio interference and policy management point of view, administrations will then need to consider how to implement the appropriate spectrum management plan. Is machine learning or artificial intelligence appropriate as part of the solution? Are new radio equipment standards necessary prior to implementation? What economies of scale are required? To which other bands are these solutions appropriate, and in which other countries? Clearly, to meet the challenges of 5G, tech/telecom companies and government alike will have to do some collaborative and innovative thinking to reap the benefits this platform promises.