Silicon Labs targets ultra-low power Bluetooth connectivity with latest SoCs

2 mins read

Silicon Labs has introduced two new SoCs, the BG22L and BG24L, for Bluetooth LE connectivity, with the BG24L Bluetooth 6.0 SoC optimised for Channel Sounding and AI/ML applications.

Silicon Labs introduces two new SoCs, the BG22L and BG24L Credit: SB4 GM - adobe.stock.com

The BG22L combines security, processing power, and connectivity for high-volume, cost-sensitive, and low-power applications. The BG24L SoC includes the Silicon Labs accelerator for AI/ML applications and support for Bluetooth Channel Sounding for asset tracking and geofencing, even in the most crowded areas like packed warehouses and multi-family housing units.

Commenting Ross Sabolcik, senior vice president of the Industrial and Commercial business unit at Silicon Labs said, "The BG22L and BG24L expand our portfolio to match those customer needs and concerns by providing an optimised set of Bluetooth features with our signature IoT capabilities like high RF sensitivity, low power, robust security, and powerful compute."

Channel Sounding was one of the most exciting features when the Bluetooth SIG announced Bluetooth 6.0 in September 2024. It allows devices like the BG24L SoC to use two-way connections to determine the distance between themselves within sub-meter accuracy. This opens a range of applications from item finding to access control.

Another use case for channel sounding is wireless access control. Bluetooth 6.0-equipped devices will be able to communicate with vehicles, package lockers, and other things that generally require an interaction or passcode. By sensing that a device is in close proximity, the locker or vehicle can open automatically simply by recognising the signal of the approaching, securely designated Bluetooth device.

With a 78 MHz ARM Cortex M33 processor, 768 kB of Flash, and 96 kB of RAM in a 5 mm x 5 mm QFN40 packaging, the BG24L is intended to be a cost-effective SoC for high-volume Bluetooth Channel Sounding.

For AI and machine learning applications, the BG24L also includes Silicon Labs' proprietary Matrix Vector Processor (MVP) AI/ML accelerator. With the embedded MVP accelerator, the BG24L provides up to 8x faster inferencing performance using only 1/6th of the power compared to performing the same calculations on the Cortex M, helping to extend battery life. Bu using AI/ML development tools from Silicon Labs the BG24L provides an accessible IoT platform for developing edge AI applications.

While these devices support Channel Sounding and AI/ML, there are also opportunities to leverage the core strengths of Bluetooth, with billions of devices using it for simple sensors, asset tracking, beacons, and streamlined Wi-Fi setup for appliances.

The BG22L is competitively designed for these common wireless applications where secure Bluetooth connectivity is needed. BG22L SoCs deliver a balance of low cost, low power consumption, high reliability, and superior performance. It includes ultra-low power receive mode along with a Precision Low-Frequency RC Oscillator (PLFRCO) that eliminates the need for an external oscillator with no performance loss, which conserves space within the host device while saving on costs and bills of materials for the manufacturer.

These enhancements enable BG22L to operate for up to ten years on a single coin cell battery.