Still, mmWave spectrum is sure to play a key role in fulfilling 5G's promise. Verizon, for instance, has plans to extend its existing mmWave network (its "Ultra Wideband" 5G network) with another 14,000 millimeterWave sites this year. And while deployments may have stalled, Qualcomm Technologies is developing new ways to overcome the challenges associated with millimeter wave.
Qualcomm on Monday showcased its latest over-the-air (OTA) prototypes and simulations to improve 5G mmWave performance and efficiency. It also demonstrated new applications for 5G mmWave, including "boundless XR" -- more immersive mobile virtual and augmented reality experiences that can support multiple users across diverse applications.
Repeaters: Extending mmWave's reach
Qualcomm's new OTA prototypes include a series of repeaters -- devices that extend a millimeter wave signal. Millimeter wave 5G offers high-frequency, high-speed service. However, it only covers short distances without any interference -- once the signal hits a building, for instance, it typically dies. (By comparison, sub-6 GHz 5G offers more coverage at lower speeds.)
At Qualcomm Technologies' outdoor OTA test network in San Diego, the company tested three types of repeaters: a simple, bidirectional repeater; a TDD-aware repeater, which amplifies either the uplink or downlink; and a smart repeater, which has TDD-awareness and dynamic beam management.
"Repeaters can absolutely be a cost-effective way of improving coverage, mobility, robustness," Ozge Koymen, senior director of technology at Qualcomm, told reporters. "Each kind of repeater really has its own advantages and disadvantages as different cost point and they all play a role in improving millimeter wave coverage."
In a typical non-line-of-sight scenario, such as in an urban canyon where a mobile user is walking between tall buildings, a smart repeater can offer more coverage.
For line-of-sight coverage extension, all the repeaters help, Koymen said. "Where the smart repeater shines is when the person moves away from the core site, there's some mobility away from the repeater," he said.
For targeted "out-to-in" coverage -- when a user is moving from outdoors to a targeted, indoor location such as a cafe -- a single repeater is sufficient. For full coverage of a complex indoor space, such as a walled office, a smart repeater has the best performance.
Machine learning: Optimizing mmWave deployments
Qualcomm has also demonstrated ways to adopt machine learning to optimize 5G mmWave deployments and improve their robustness.
One way is to use ML-based beam prediction. This can increase a system's usable capacity by reducing the communication overhead, and it can extend the device battery life.
Qualcomm is also using machine learning and optimization techniques to plan network topology -- in other words, to determine which kind of nodes to use, as well as what kind of connectivity. To accomplish this, Qualcomm creates a digital twin of the targeted deployment area, using a variety of data sources to capture the local environment, including foiliage that could interfere with signals.
With a digital twin, Qualcomm can abstract the problem into a network graph to find the right balance of performance and cost-efficiency. This helps determine whether to use IABs, repeaters or other nodes. It can help determine which kind of connectivity is optimal, such as whether drawing fiber to a node is worth it or if it'd be more cost-effective to use out-of-band wireless.
Qualcomm is also demonstrating how 5G NR-Light in Release 17 reduces device complexity and power consumption, making it suitable for IoT use cases like wide-area surveillance cameras. 5G NR (New Radio) is a global standard for a 5G wireless air interface that delivers faster and more responsive mobile broadband. The first 5G NR standard was adopted in 2017, driven by Qualcomm innovations. In its latest demonstrations, the company showed how NR-Light devices can be added to a factory floor without using significantly more network resources.
Boundless XR: Using mmWave for immersive experiences
Qualcomm also demonstrated its "boundless XR" platform, which is currently in commercial trials. Boundless XR uses 5G and distributed computing for immersive VR and AR experiences. Image processing happens both on the device and in the edge cloud.
The use cases include real-time interactive communication, shopping, working in a mobile VR office and multi-player gaming with photorealistic graphics.
For example, a VR device would send 6 degrees of freedom data to a edge cloud server over 5G. The edge cloud processes this data and renders an image. The compressed image would be sent to the device over the high-bandwidth, low-latency 5G connection. Qualcomm has worked on achieving on-device optimization and has partnered with Ericsson for system-level optimizations to keep the 5G roundtrip time as low as possible.
The Boundless XR optimizations manage to maintain a solid user experience even in scenarios that should create latency spikes, such as when one user is blocking another user's 5G mmWave signal. Qualcomm's research in this area has laid the groundwork for XR experiences that leverage mmWave 5G, even when multiple users are involved.