| Ultra-Wideband impulse radio (UWB-IR) is a promising technology for wireless communications due to its carrier-free feature. Compared with narrowband radio, UWB-IR has the potential of low complexity and low power consumption, which makes it a desirable radio for low power wireless applications such as wireless sensor networks and RFID. In this dissertation, we investigated how to improve energy efficiency from system level to circuit level for environmental monitoring applications. By taking advantage of the low occurrence rate of event of interest in environmental monitoring applications, we proposed an adaptive architecture that allows the system to work at three modes, high speed, normal, and power saving, under different QoS scenarios with the help of a cross-layer adaptive control block. At the circuit level, we proposed a delay locked loop based adaptive dock generator that can generate docks with three different frequencies to support the above three modes by using a frequency control vector. The switching time from one frequency to another is less than 1 microsecond. To meet the FCC's spectral mask, a Gaussian Pulse Approximation (GPA) approach was proposed to approximate the Gaussian pulse with 96% accuracy by using the transient response of CMOS inverters. The pulse width and amplitude can be adaptively tuned to compensate for the PVT variation and the approximation error through a pulse shaping control vector. On the receiver side, the resolution of the ADC can be scaled down to save power through an ADC resolution control vector. An adaptive current source based comparator was proposed to reduce the INL caused by the voltage and temperature variation. Finally an adaptive sampling dock generator was proposed to maintain the synchronization of the sampling and compensate for the jitter of the received signal. Simulation results indicate that the proposed approaches could reduce the power consumption significantly with minor overhead. |
