The urban-rural digital divide: can 5G close the gap?
March 6, 2018
By the end of 2017, the number of fibre connections in New Zealand reached 506,075 with an uptake rate of just over 40% of homes passed. By the end of the Ultra-Fast Broadband Extension (UFB2) in 2022, 87% of New Zealand homes and businesses are expected to have access to fibre-based broadband. With download speeds of up to 1Gbit/s, customers can enjoy triple-play services on multiple devices per household. However, for the remaining 13% of homes in rural New Zealand the luxury of tapping in to an unlimited ultra-high speed data pipe will be out of reach, or come at a significantly higher cost.
4G vs wireline services: quality and price
As part of the publicly funded Rural Broadband Initiative (RBI), Crown Infrastructure Holdings has partnered with many regional providers and national mobile network providers to deliver a fixed wireless 4G service, which provides download speeds of up to 40Mbit/s. Currently, video streaming is the main driver for bandwidth demand and is expected to represent about 81% of consumer Internet traffic by 2021, according to Cisco’s Visual Networking Index published in June 2017. Web browsing and email traffic, on the other hand, make up 10-15% of Internet traffic. A high definition (HD) video such as Netflix uses about 4Mbit/s per stream, which ideally requires dedicated bandwidth. Web browsing, however, is sporadic in nature and due to the relatively small sizes of web sites, multiple web users sharing a link are unlikely to affect each other’s sessions. From that perspective, a well-designed 4G fixed wireless network appears to sufficiently cater for two or three HD streams and multiple web sessions per household. However, the critical fact about fixed wireless access is that the cell capacity is shared among multiple users, which could potentially degrade the quality of a connection in peak traffic times.
Experience from urban New Zealand shows that there is a considerable difference in quality between fixed wireless and wired broadband. While wired broadband services in urban New Zealand suffer minimal speed degradation during busy traffic, the average drop in the speed of a 4G fixed wireless connection is about 25% in peak traffic times (Exhibit 1). Some 4G connections can experience a degradation of up to 40%. With the increasing demand for video streaming, the currently deployed 4G systems will soon reach their capacity limits.
Other key performance indicators are the latency and video buffering events (Exhibit 2). Here, 4G fixed wireless scores worse than ADSL. While these differences will not be noticeable to an average user, the implication is that 4G fixed wireless will not be sufficient to support emerging applications such as augmented reality and tele-health care with haptic feedback. Other limitations of 4G subscriptions are the imposed data caps and higher price compared to similar capacity urban wireline services, such as VDSL. Rural 4G fixed wireless services are usually available in 30-200GB/month plans. These data caps are commercial constructs designed to avoid oversubscribing the shared broadband resources. A rural 4G fixed wireless 200GB/month plan is about twice as expensive as an unlimited VDSL service. Worth noting here is that the average monthly data usage of a New Zealand household connected by copper or fibre exceeded 150GB by the end of 2017, with a projected annual growth rate of 20-25%. Such differences and trends are expected to further deepen the urban-rural divide.
Can 5G address the rural-urban broadband gap?
5G promises to provide speeds 10-100 times higher than 4G, latencies lower than 5ms and the ability to support a massive number of connected devices for machine-to-machine communications (M2M) and the Internet-of-Things (IoT). Some of the important features of 5G are the ability to utilise a wide range of spectrum (from 600MHz to 100GHz) and different channel sizes, and hence the ability to address a variety of applications and deployment scenarios. However, massive capacity gains of 5G rely on using millimetre waves (mmWaves), massive Multi-Input Multi-Output (MIMO) techniques combined with antenna beamforming, and small cell deployments (Exhibit 3). As mmWaves travel only short distances, have low penetration power and require line of sight, 5G is mainly intended for densely populated urban areas with cell radii of a few hundred metres.
Exhibit 3: Massive MIMO employs multiple transmitters and antenna elements to transmit simultaneous data streams on the same radio channel. This is backed by antenna beam forming technology to intelligently target users, reduce interference and extend reach. [Source: Network Strategies]
The 5G standard is expected to be finalised by 2020/2021. However, many companies aspire to take a lead in the 5G market before that date. Starry, a US based company, has already launched an urban wireless broadband service using pre-standard 5G. Using 37.0-38.6GHz spectrum, the company provides an uncapped 200Mbit/s service at a price that compares favourably to equivalent fibre plans. In a rural context, a similar scenario might only be viable in small towns with closely located premises where a 5G access point can be mounted on an utility pole or a building. A fibre cable is still required to backhaul the 5G traffic, creating the need for pushing fibre deeper into rural areas. On the subscriber side, outdoor units will likely to be necessary to receive the mmWave signals. However, the economics of such a deployment are yet to be explored. Whether it will be more cost effective to bridge the last couple of hundred metres with a 5G wireless link instead of fibre will depend on 5G hardware costs.
That said, the technologies associated with 5G still have some promise for providing better broadband in sparsely populated areas. Developments in antenna beamforming, massive MIMO techniques, and energy efficiency of base and relay stations will help in extending the reach of higher frequency bands, improving cell capacity and reducing capital expenditure. However, the gains compared to advanced standards of 4G, such as LTE-A and LTE-A Pro, would be limited and it is unlikely there will be a significant difference between 4G and 5G over the 600-700MHz band. The main reason is that most of the 5G enabling technologies are already part of the evolving 4G standards. LTE-A already employs 4×4 MIMO, high rate modulation schemes and carrier aggregation over multiple frequency bands to deliver Gigabit class speeds, as demonstrated by recent trials in rural Australia. 8×8 MIMO, which is also supported by LTE-A , has the potential to further double capacity.
There are compelling societal arguments for improving rural coverage, including better public safety, e-health applications, e-learning and counteracting the overcrowding of urban centres by making these hinterland areas more attractive for residents. In many parts of the world, moderately priced and high speed rural broadband connectivity remains a challenge. For rural towns in New Zealand to experience the full features of 5G, a deeper penetration of fibre is necessary to place 5G access points within the reach of mmWaves. In sparsely populated rural areas, such as farms, 4G and its evolving versions are likely to dominate in the long term, providing increased speeds over time. However, the digital divide problem is not just about speed but also about quality, competition and price. The considerable 4G/wireline price gap is unlikely to be addressed by technology alone, creating a growing need for social and regulatory initiatives.