Signal processing approach towards designing ultra-low power transceivers

We propose an ultra-low power transceiver used as a wake-up radio, to detect a beacon signal transmitted from an access point and subsequently enable the rest of the transceiver chain. However, such ultra low-power radios suffer from poor sensitivity due to noise. The objective of this paper is to model such RF samplers and understand the signal processing in Rx chains. Based on this model, we propose optimal signal shaping techniques and achieve significant improvements noise sensitivity. Finally, we derive the performance bounds of such transceivers and show the convergence of optimal signal shaping techniques with that of ideal transceivers under Gaussian noise.