MathWorks 5G Trends and Predictions for 2020
Ken Karnofsky, Senior Strategist, MathWorks

Deep Learning in Communications Remains in a Nascent Stage
In 2020, there will be increased interest surrounding the integration of deep learning into communications systems. Yet despite a heightened desire from the wireless industry to apply deep learning or AI, this will remain in the exploratory and research phase for the time being.

Over time, there will be a few areas where deep learning can be applied, including:

• Recognizing signals in locations or spectrum bands where there is a high saturation of other signals, such as urban areas or with military applications.
• Building adaptive algorithms that can respond to changes in the environment, such as digital predistortion to linearize the power amplifier in a mobile phone efficiently based on channel conditions
• Enabling communications networks to provide better service given fluctuating network traffic and peak loads.

Advanced Connectivity Is Being Developed in Parallel with 5G
WiFi will continue to evolve as we can soon expect to see more advanced versions of higher network capacity and greater download speeds. WiFi 6 is being deployed around the same time as 5G – providing options for achieving higher bandwidth in large commercial areas, such as factories and sports stadiums. 

Bluetooth is fairly ubiquitous among consumer devices, but 2020 will lead to Bluetooth-enabled devices that use location tracking information for indoor local positioning applications. This will allow robots or machines with Bluetooth capabilities to be tracked on a factory floor, for example, and easily be able to transfer data or control information.

IoT connectivity, whether using cellular, Bluetooth, or WiFi will demand greater levels of power management. For example, narrowband Internet of Things (NB-IoT) solutions will be transferring data over long distances and low data rates from potentially remote locations – all while needing to be done in low power mode.

Expect a Greater Need for End-to-End Technology Integration
Currently, the entire signal chain, from RF to baseband, can be implemented in a single programmable device or module. However, most engineering teams do not possess the incumbent expertise to design these devices and integrate them into a complete system.

This is a major challenge for R&D teams as they work to prototype and validate new 5G design concepts. To mitigate these challenges, companies will need to increasingly equip their wireless engineering teams with software tools that connect algorithm design, system simulation, beamforming architectures and system partitioning, over-the-air (OTA) testing, prototyping and implementation – saving development and verification time as a result.

Software tools that enable an integrated engineering team and linked workflow will be necessary as design teams tackle advanced 5G networks and clear hurdles such as standards compliance and technical complexities driven by multi-antenna systems and highly coupled digital and RF front ends.

See the next page for Ericsson 2030 predictions