Standards work on 5G began in late 2015, and the first commercial networks probably won’t launch until 2020 at the earliest. But it’s not too early to begin pondering what 5G could mean for verticals such as health care, manufacturing, smart cities and automotive.
One reason is because some of these industries make technological decisions several years out. Automakers, for example, will need to decide in the next year or two whether to equip their 2021 models with LTE-Advanced Pro or add support for 5G, too. Another reason is because understanding 5G’s capabilities today – even at a high level – enables businesses and governments to start developing applications that can take advantage of the technology’s high speeds, low latency and other key features.
As they collaborate on 5G standards, cellular vendors and mobile operators should pay close attention to those users’ visions and requirements according to a white paper commissioned by the European Commission and produced by the 5GPP (more information at https://5g-ppp.eu) . If 5G falls short in key areas such as latency, reliability and quality-of-service mechanisms, the cellular industry risks losing some of those users – and their money – to alternatives such as Wi-Fi. A prime example is HaLow, formerly known as 802.11ah, which Maravedis believes is potentially a very disruptive technology.
The ITU, 3GPP and other organizations developing 5G have set several goals for the new technology, including:
- Guaranteed speeds of at least 50 Mb/s per user, which is ideal for applications such as video surveillance and in-vehicle infotainment. But it’s probably not enough if a user is actually multiple users, such as a 5G modem in a car that’s supporting multiple occupants and the vehicle’s navigation, safety and diagnostics systems.
- The ability to maintain a connection with a device that’s moving on the ground at 500 km/h or more, enabling 5G to support applications such as broadband Internet access for high-speed rail passengers. Even on the Autobahn, cars rarely move faster than 150 km/h, so setting the baseline at 500 km/h ensures sufficient headroom for virtually all vehicular applications.
- Support for at least 0.75 Tb/s of traffic in a geographic area the size of a stadium, which in theory could reduce the need for alternatives such as Wi-Fi. But in reality, mobile operators almost certainly will continue to offload a lot of 5G traffic to Wi-Fi as they do today with 4G due to the fact that licensed spectrum is, and always will be, limited and expensive.
- The ability to support 1 million or more devices per square kilometer, an amount that’s possible in a dense urban area packed with smartphones, tablets and IoT devices. This capability would help 5G compete against a variety of alternatives, such as Wi-Fi and ZigBee, although ultimately the choice comes down to each technology’s modem and service costs. If 5G debuts in 2020, it would take at least until late that decade for its chipset costs to decline to the point that it can compete against incumbents – including 4G – in the highly price-sensitive IoT market.
999 percent reliability, which maintains telecom’s long tradition of setting five-nines as the baseline for many services. But this won’t be sufficient for some mission-critical services, such as self-driving cars and telemedicine, which may require up to 99.99999 percent reliability.
- The ability to pinpoint a device’s location to an area 1 m or smaller, a capability that could enable 5G to compete with Wi-Fi and Bluetooth for beacon-type applications. But it might not be enough for automotive applications, where 0.3 m precision sometimes is required. Like 4G, 5G will use carrier aggregation and small cells, which together create barriers to precision location indoors because combining signals from multiple sites means a device is in a much larger area than if it were connected to only one. Some vendors are working to address this problem with 4G, and 5G could leverage that work to enable high precision.
- 5 ms or less of end-to-end latency, which is sufficient for the vast majority of consumer, business and IoT applications. One factor that affects latency is whether a network is used. The latest versions of LTE support direct communications between devices, such as for public safety users in places where the cellular network is down. 5G will support device-to-device communications, where the absence of network-induced latency could be useful for industrial applications that require latencies as low as 100 μs.
4G has begun leveraging telecom-wide trends and technologies such as Network Functions Virtualization
(NFV) and Software Defined Networking (SDN). These enable mobile operators to leverage the cloud and replace cellular-specific infrastructure with off-the-shelf IT gear such as servers. All of these real-world experiences will help create 5G technologies that can dynamically allocate computing and storage resources to meet each application’s unique requirements for performance, reliability and other metrics, as well as each operator’s business model. For example, some mobile operators are already considering having data center providers host their RAN, EPC or both to reduce their overhead costs. 5G could make that model even more attractive.