Ultra-Low Power, Low Noise Mac Protocol Compliant Non-Pll Based Wearable Transmitter For Dynamic Open Spectrum Sharing

The current paper proposes an ultra low power, low emission media access control (MAC) compliant frequency shift keying neural transmitter to operate within wireless body area networks (WBANs). The design has been tested using a 10s single channel sleep related EEG recording. An ultra low power, low noise chopper stabilized neural amplifier (0.494 mu Vrms input referred noise) has been used to amplify the incoming EEG signal. The EEG signal is converted into a binary bit stream by a 12 bit analog to digital converter followed by a parallel to serial converter. Frequency modulation is achieved by utilizing a ramp to encode the bit stream derived from the EEG signal in to two distinct frequencies. The negotiated frequency of operation for dynamic frequency shifting and frequency encoding is achieved changing the gradient of the ramp thorough load cap and charging current in a charge pump. The work has been implemented on 0.18 mu m AMS technology and results presented.