BCIwiki (bciwiki.com) — Researchers have developed a 32-channel event-based bio-signal analog front-end chip fabricated in 180 nm CMOS, supporting dual-mode encoding with pulse frequency modulation and adaptive asynchronous delta modulation. The chip adapts its encoding data rate in real time based on the input signal envelope, achieving high data compression and paving the way for wireless transmission and online processing of neural signals in brain-computer interfaces. The preprint was posted on arXiv on July 14, 2026.
The chip integrates 32 independently programmable input channels, each configurable for either pulse frequency modulation or adaptive asynchronous delta modulation. The adaptive asynchronous delta modulator features an auto-scaling mechanism that adjusts the encoding data rate in real time based on the input signal envelope, enabling efficient information transmission at low power. The design aims to support neural signal acquisition and encoding compatible with spiking neural network processors, advancing brain-computer interface systems toward low-power, wireless operation.