High frequency noise modeling for RF CMOS and low noise amplifier design

Abstract

Continuous advances in a CMOS process technology have reduced the minimum channel length of MOS device, thereby improving the microwave performances greatly. So, by using the MOS technology, many researches have been carried out to realize system-on-chip, which integrates digital modules and RF modules. Although there are many progresses in I-V and C-V modeling in MOSFETs, existing MOSFET compact models used in modern circuit simulators are unfortunately not suitable for predicting high frequency thermal noise behaviors. The noise behavior in short-channel MOSFETs is not well understood yet and so even very controversial. It has been often reported that the drain thermal noise generated in short-channel MOSFETs is higher than predicted by the long-channel noise model. The significant excess noise was reported in the saturation regime. The excess factor was found to be enhanced up to 8 in n-channel devices with 0.7 ㎛ gate length. And there are many literatures to try to explain such severe increase of the drain thermal noise for short-channel MOSFETs by introducing hot electron effect in velocity saturation region. More recently, much smaller enhancement was reported in device with 0.65 ㎛ gate length. But the excess factor dependency on gate bias has been quite different among other groups. For the induced gate noise, experimental data have seldom been reported. Recently, the reported induced gate noise shows an even more dramatic enhancement factor as large as 30 for a 0.25 ㎛ gate-length n-channel MOSFET. Therefore, more detailed experiments and modeling are urgently needed because the reported noise enhancements would seriously limit the viability of RF CMOS technology. We think the following prerequisites should be solved in the noise modeling. Firstly, the noise measurement accuracy has to be confirmed to obtain a meaningful noise analysis. Since the noise experimental results critically depends on the measurement system, the measurement accuracy s...