The era of 5G is well and truly upon us. The next cycle of mobile technology is being heralded as the breakthrough that consumers, corporations, and now industries have yearned for since 3G became a reality. The deployment of 5G networks will probably follow a similar path to prior mobile generations, but with one important difference: the lack of true low-band spectrum for wide-area coverage that can deliver a step change in average radio transmission speeds to end users and devices. How CSPs refarm low-band spectrum assets to make way for 5G will be a key consideration in ensuring that 5G becomes a global reality for all.
Bandwidth needed for fast 5G speeds available only in poor-coverage high-band spectrum
Mobile technologies have typically been deployed in low bands for coverage and high bands for capacity, as shown in Figure 1.
Figure 1: Initial primary spectrum bands for each generation of mobile technology
With each generation of mobile technology, achieving higher transmission speeds requires increasing spectrum in terms of both new bands (e.g., mmWave at 26–28GHz) and greater bandwidths (e.g., 100MHz). With 3G and 4G, spectrum bands for coverage (below 1GHz) and spectrum bands for capacity (above 2GHz) were chosen to complement each other and so facilitated a smooth rollout, enabling wide-area and in-building coverage, along with hot-spot-type capability for denser areas.
For 5G, the definition of low and high bands has shifted, given the limited availability of new spectrum bands below 3GHz. This redefinition presents significant challenges for 5G in achieving the coverage and capacity typically seen in prior mobile technology rollouts. At spectrum bands above 3GHz, coverage is significantly reduced, given their poorer radio propagation characteristics than bands below 1GHz. For mmWave spectrum (>26GHz), radio propagation is even more attenuated, driving a requirement for network densification to deliver suitable coverage and capacity at these new high frequencies. The implications for achieving ubiquitous 5G coverage – even a "thin layer" – are therefore significant, namely increased network capex due to densification, challenges with small site acquisition, the ability to load existing tower infrastructure with 5G equipment, and management of customer experience with spotty 5G coverage. To solve these problems, CSPs will look to refarm existing low-band spectrum (e.g., 700/900MHz and 1800MHz). However, the challenge is that 3G and 4G services are currently deployed in these bands.
What should CSPs and regulators do?
As a consequence, a massive planning effort will be required by CSPs to accelerate and manage the shutdown of 2G/3G to clear sufficient true low-band spectrum for ubiquitous 5G coverage. For 5G, each CSP will need to consider, in their own markets, changing consumer behavior and preferences, rate of usage growth, spectrum band use by technology and by geography, upcoming spectrum auctions, and device cost and ecosystem. A standard playbook is unlikely to exist. However, CSPs can take steps to ensure a sustained leadership position in 5G, rather than merely a "launch" leadership:
- Accelerate 2G/3G shutdown in markets where 4G is relatively mature. Shutdown planning should be finalized by end-2019, with a two-year grace period announced to manage customer migration and so enable sufficient true low-band spectrum to be freed up for 5G wide-area coverage by 2022. Less mature markets will have more time.
- The 3G bands 2100MHz and 2600MHz offer significant potential and global appeal for 5G, as more than 10MHz (ideally 20MHz) is required to meaningfully deploy a technology for coverage and capacity.
- The 900MHz band is likely to be unsuitable, as CSPs typically do not possess sufficient spectrum. The 800MHz band has potential in some markets.
- Keep 700MHz for 4G, given the global dominance of that technology.
- Consider 1800MHz for 5G once the majority of traffic moves from 4G to 5G.
- Enable CSPs to test 5G radio technology in various bands by facilitating appropriate test licenses in relevant areas. (Updated propagation models for 5G coverage and capacity are critical to its success.)
- Facilitate clearing of globally harmonized 5G bands well ahead of market needs (>3 years) to provide certainty to CSPs who need to identify strong business cases, seek appropriate funding from internal governance processes, and award contracts to vendors and service partners to approach 5G rollout with vigor.
- Ensure spectrum licensing is technology-agnostic, to enable CSPs to refarm spectrum bands as needed. (In several markets, spectrum has been licensed for a specific technology.)
Spectrum Requirements for 5G, TE0007-001111 (February 2017)
"Spectrum policy should be the top priority for regulators in the race to 5G,"GLB005-000080 (August 2018)