Quantization Noise Suppression in Digitally Segmented Amplifiers

In this paper, we consider the problem of out-of-band quantization noise suppression in the general family of direct digital-to-RF (DDRF) conversion circuits, where the RF carrier is amplitude modulated by a quantized representation of the baseband signal. Hence, it is desired to minimize the out-of-band quantization noise in order to meet stringent requirements such as receive-band noise levels in frequency-division duplex transceivers. In this paper, we address the problem of out-of-band quantization noise by introducing a novel signal-processing solution, which we refer to as ldquosegmented filtering (SF).rdquo We assess the capability of the proposed SF solution by means of performance analysis and results that have been obtained via circuit-level computer simulations as well as laboratory measurements. Our proposed approach has demonstrated the ability to preserve the required signal quality and power amplifier (PA) efficiency while providing more than 35-dB attenuation of the quantization noise, thus eliminating the need for substantial post-PA passband RF filtering.