Qualcomm has been instrumental in developing and delivering the technologies that make mmWave 5G, both mobile and fixed, commercially viable. Now the company is looking to the next step of mmWave which includes using machine learning to help operators simplify network planning and deployment, reducing latency and smooth inter-cell handovers by moving mobility management from Layer 3 to Layer 1 and 2, and how operators can see system-level capacity gains by sharing high-band spectrum. Qualcomm Senior Director of Engineering Ozge Koymen shares his outlook on the continued evolution of mmWave.

At MWC 2022, Qualcomm showcases the many noteworthy benefits of 5G mmWave technology, including extremely low latency and rapid speeds that can deliver massive capacity to support dense mobile user connections. This is achieved by utilizing a wider spectrum, with frequency bands above 24GHz, so that users can experience the true speeds high-performance 5G systems offer.

At MWC 2022, Qualcomm showcases the many noteworthy benefits of 5G mmWave technology, including extremely low latency and rapid speeds that can deliver massive capacity to support dense mobile user connections. This is achieved by utilizing a wider spectrum, with frequency bands above 24GHz, so that users can experience the true speeds high-performance 5G systems offer.

At MWC 2022, Qualcomm showcases new technologies focused on AI-enabled Air Interface. In this video, we guide you through demonstrations of two over-the-air prototypes that utilize machine learning to bring new levels of air interface performance and efficiency.

For the first over-the-air demonstration, we use our test network and a test device based on the Qualcomm Snapdragon 5G modem-RF system. There are neural networks running on both the device and network. We compare the performance of different neural network implementations to learn how machine learning can enable more efficient Massive MIMO Systems.

Here, we learned that machine learning can bring better performance and efficiency to the 5G system by exploiting a new data-driven design approach.

For the next demonstration, our team uses our mmWave OTA test network in San Diego. We implement machine learning-based beam management and use a test device to compare that algorithm’s performance and network KPIs against the traditional method of beam management.