5G is developing rapidly. From vague notions of the next generation of revolutionary wireless technology, to elusive goals, to increasingly sophisticated use cases and technical standards, 5G is evolving at a rapid pace, with achievable goals and implementable standards already in place. In the process, 5G will exhibit very different characteristics from current cellular networks.
Early on, people often talked about the quantitative changes 5G would bring: gbps-grade bandwidth, huge surface density in competitive urban markets, excellent energy efficiency, and so on. LTE still has its limitations. And the 5G vision is considered elusive. Many application architects see 5G as a fast, highly available and reliable network that will help them out.
Want to expand the fixed broadband access market without any fiber or copper wires? 5G could help, with speeds comparable to those of fiber optics. Want to experience mobile augmented reality without using a 5kg helmet? It's a piece of cake: with seamless, always-available high-bandwidth connectivity, 5G will help do all the heavy computing in the cloud. Want a self-driving connected car with no supercomputer in the trunk? With only a 5G modem, the cloud offers an unparalleled self-driving experience (figure 1). Want to connect sensors and brakes of the Internet of things (IoT) system directly to the Internet? As people wish
These objectives cover different areas and are not unattainable when time and resources are available. Needless to say, standard-setters need to keep expectations within reasonable bounds if progress is to be made.
The itu's international mobile telecommunications (IMT) vision 2020 statement reduces the wish list to three representative use cases: enhanced mobile broadband, large-scale machine-like communications, and ultra-reliable low-latency communications. These three use cases will help fulfill many of the expectations of the 5G space.
Of the three use cases, enhanced mobile broadband is probably closest to what most people envision for the next generation of mobile phones. In this use case, a static user in a place equipped with advanced technology can obtain data rates of up to 20 Gbps, a mobile user can obtain sufficient actual bandwidth (from 80 to 200 Mbps, depending on the place), stream 3D or uhd video, and interact closely with cloud applications in scenarios such as games or augmented reality.
Large-scale machine-class communication provides an entirely different scenario. The clients here are not servers or people, but iot devices in smart cities, factories, buildings or homes. In this scenario, the raw data rate is less important; The machine either provides relatively little information -- just a few sensor readings per second -- or is preprocessed locally to dramatically reduce bandwidth (like a smart surveillance camera). The key here is not data speed, but connection densities - up to a million connected devices per square kilometer - and energy efficiency - equivalent to 100 times that of 4G networks.
The third scenario, ultra-reliable low-latency communication, is an incredibly new use case that supports industrial automation, mission-critical connectivity, and autonomous vehicles to avoid dropped calls while driving. Not only do these applications require relatively high data rates -- such as cars sending large streams of information from their cameras and lidar to the cloud for analysis -- but they also require two attributes that have little to do with wireless networks: millisecond latency and functional reliability.