Thanks to AI, modems will be able to improve the performance of 4G and 5G, but it is on mmWave millimeter waves that the amplification will be the most interesting.
If, on smartphones, AI is often used to improve the appearance and quality of photos, it will soon be able to boost cellular performance and the coverage of 5G mobiles. This is at least Qualcomm’s bet, which claims that the AI algorithms integrated into its modems will be able to amplify signal coverage and thus extend the range of all radio devices connected to the device, including 4G and 5G equipment. The use of AI will be included in the Snapdragon X70 modem which should be delivered this year to be found in smartphones in 2023.
These announcements were made during Qualcomm’s 5G Summit held May 9-11 in San Diego, California. Most of the improvements are in short-range, high-speed millimeter wave (mmWave) technology. 5G combines two technologies working in parallel: Sub-6GHz wireless signals, which can cover tens of kilometers at a rate of a few hundred megabits per second, and the powerful but limited mmWave technology, which only covers a few tens meters at speeds of more than one gigabit per second of throughput. Operator T-Mobile also uses an intermediate band which bridges the gap. But what interests most users are mmWave’s big capabilities, and its many real-world applications, for example thousands of people taking photos and videos of an outdoor concert, then sharing them on social networks. However, due to its limited range, Qualcomm and carriers want to improve mmWave’s performance. And AI can provide an answer.
Improve smartphone/relay antenna connections
Ignacio Contreras, Qualcomm’s director of product marketing, told reporters that AI-powered inference can extend mmWave’s reach by about 20 percent, though the technique can be used to improve other cellular technologies as well. In summary, Qualcomm teaches the modem what to do before it is faced with the real-world scenario. The mobiles always communicate with a relay antenna to which they send information on the context so that it optimizes the transmission of the data. The modem supported by the AI can therefore be finely configured before it is put into action. In the real world, the smartphone modem will communicate not only current conditions, but also its predictions about the state of the network, so as to improve the performance of all devices in the system. With the mmWave, the smartphone could indicate to the mobile that it would be better to switch to a new beam if it is available.
Smart Transmit 3.0 technology automatically adapts the power when a smartphone downloads data over the network. (Credit Qualcomm)
“Thanks to the artificial intelligence processing embedded in the system model, the device will not only be able to inform about the current conditions, but it will also be able to predict the conditions at the next moment,” added Mr. Contreras. . At the same conference that ends today, Qualcomm will also demonstrate standalone mmWave technology. Today, every mmWave deployment requires sub-6GHz spectrum anchoring to perform control functions. By freeing millimeter waves from this control, 5G should be able to be used in more applications, including fixed wireless broadband, as offered by T-Mobile. Finally, Qualcomm has announced that it will launch a technology called Smart Transmit 3.0, which can intelligently manage power when a smartphone downloads data over the network. Smart Transmit 3.0 power management algorithms will now include information sent over WiFi and Bluetooth.