Spectrum is becoming key for enabling 5G ambitions. As the next generation of mobile connectivity, 5G promises to support new types of applications and services with the help of low but also extremely high frequencies, including millimeter waves (mmwaves) to provide very high throughputs. These networks will take advantage of ten times as much spectrum as they use today.
Impact on Architecture
The extension of spectrum range has an impact on the network architecture. mmwave cells will employ shorter ranges of around 100-to-200 meters which will require extreme densification to provide high coverage. 3G networks reached densities of fourto- five base stations per km², 4G networks eight-to-ten per km², while 5G networks could reach densities of 40-to-50 per km². Each base station could provide transport links up to 25 Gbit/s, with up to 100 Gbit/s backhaul links. Often, new optical transport networks will need to be built to accommodate these requirements.
In addition, mmwaves are extremely fade sensitive; their effective range depends on building materials, weather, foliage, mast height and whether line-of-sight is available.
Consequently, it is reasonable to believe that very high throughputs will be limited to urban areas where a high density of base stations can be deployed, along with a high capacity transport network. Fortunately, small cells, with their reduced costs and self-backhaul technology, should help in deploying such dense networks.
In semi-urban areas, very high 5G throughput will rely on the availability of a powerful optical backbone. In any case, 5G will not replace LTE; both technologies will co-exist until at least the end of 2020 and most probably beyond. LTE Advanced also brings multiple benefits including carrier aggregation, 1 Gbps peak throughputs and higher-order multiple-input and multiple-output (MIMO). LTE-Unlicensed (LTE-U), LTE- Licenses Assisted Access (LAA) and Multifire bring the additional benefit of being able to operate in unlicensed spectrum.
As for remote and difficult to reach areas, several configurations which involve satellite backhaul could be envisaged. The first one could deploy a network of small cells to cover several villages with an aggregation of the traffic at central points for backhauling. The second option would leverage a macro cell base station with very wide coverage but with lesser throughput than those small cells and their mmwaves.
According to the ITU use case model, latency requirements for 5G are very diverse.
Mission critical applications require millisecond latency, while broadband-to-the-home has a higher latency tolerance. Software updates, for example, can be transmitted with an even higher latency over satellite and be much more cost-effective than unicast. In many cases, Multi-access Edge Computing (MEC) platforms will help process content at the very edge of the mobile network and will reduce latency significantly.
As a result, different connectivity offerings can address the multiple latency use cases proposed by 5G: mmwaves, fiber/ microwave and newer satellite constellations (lower constellations, high throughput capability). The table below summarizes the connectivity choices for different requirements.
Future broadband needs will expand in all corners of the world, from connected objects in deserts to smart cities and connectivity in objects on-the-move – for example, trains, planes and ships – to name a few. Future networks will, therefore, be much more ubiquitous than today, leveraging different combinations of transport networks including satellite, 5G and LTE, small cells and macrocells to accommodate the connectivity demands of in-city, semi-urban and remote and hard-to-reach areas.