Summary Bullets:
• At the ZTE 5G Summit in Bangkok (Thailand) in November 2023, ZTE continued its focus on RIS with a next-generation prototype that consumes less power than its predecessor.
• ZTE also introduced ZTE RAN Composer, a solution that inserts general-purpose computing into the baseband unit to aid AI-based RAN optimization applications.
At the 2023 ZTE 5G Summit in Bangkok (Thailand), ZTE presented a range of views on 5G’s continuing and future evolution, including presentations from operators and analysts in addition to ZTE representatives. The company also highlighted two forward-looking technologies that stood out: reconfigurable intelligent surfaces (RIS) and what ZTE has branded as ‘RAN Composer.’
RIS
For a few years now, ZTE has differentiated itself by promoting RIS, a technology often associated with 6G that can be programmed to reflect radio signals to fill coverage gaps in intelligent ways. Using RIS may be less expensive than using small cells because RIS consumes less power and doesn’t require backhaul. At its Bangkok (Thailand) event, ZTE demonstrated an RIS prototype – a flattish square about the size of an ordinary framed wall painting, with more than 4,000 antennas and 24 dB of reflection gain – reflecting a millimeter-wave signal around a corner. In part, ZTE was demonstrating the progress it is making, whereas an earlier version of its RIS consumed 100 W of power, the one demonstrated last month (‘Dynamic RIS 2.0’), consumes just 20 W. It can also perform what ZTE calls “dynamic beam-sweeping and tracking” to maximize its performance.
Today, operators may see ZTE’s RIS as less relevant to them than solutions that are more commercially mature. Going forward, a key question for RIS offerings will be how widely they are useful and whether that justifies ZTE’s early investment and focus on them. But for now, ZTE has proactively established strong authority in this area.
ZTE RAN Composer
ZTE RAN Composer allows operators to insert a general-purpose server card into a baseband unit to support RAN-enhancing applications – especially those including AI (ZTE emphasizes ‘native AI,’ suggesting an essential role for AI in the fundamental solution design).
One use that ZTE imagines for its RAN Composer is service-aware RAN-resource allocation. The solution would analyze network traffic patterns in real time to determine which kinds of services each flow of traffic represents. With that service awareness, the solution would then choose the best RAN resources to allocate to each service’s traffic based on each service’s specific requirements. The result, in theory, would be a more efficient use of the radio spectrum and networks and improved user experiences. As ZTE put it in a white paper on this topic, “Say you have two layers of co-coverage – one at 1800 MHz and the other at 2100 MHz… You can wake up the minimum amount of co-coverage needed based on the service to be supported. If you have co-coverage at cells 4, 5, 6 and 7, you can wake up the 2100 MHz band only in Cell 4, if that’s all you need to support the service.”
Because ZTE RAN Composer could potentially address multiple applications, it might find its best fit in enterprise use cases, including private networks. ZTE is already focused on enabling other virtual network functions in scaled-down enterprise-centric solutions, including core functions, and the use of multiple virtual functions (e.g., RAN, core, etc.) could represent an efficient use of infrastructure.
Still, some operators may ask why ZTE is proposing general-purpose computing be narrowly applied to AI-based functions via an inserted card that could limit its scale and flexibility rather than offer a broader vRAN solution that supports ZTE RAN Composer’s purposes while also keeping ZTE competitive with top rivals and challengers that are driving vRAN evolution. ZTE RAN Composer could demonstrate that ZTE is investing in general-purpose computing and AI in the RAN, but, at the same time, this offering could increase pressure on the vendor to embrace vRAN more fully than it is right now.