FinFET is a multiple-gate silicon transistor structure that nowadays is attracting an extensive attention to progress further into the nanometer era by going beyond the downscaling limit of the conventional planar CMOS technology. Although the interest for this architecture has been mainly devoted to digital applications, the analysis at high frequency is crucial for targeting a successful mixed integration of analog and digital circuits. In view of that, the purpose of this review paper is to provide a clear and exhaustive understanding of the state of art, challenges, and future trends of the FinFET technology from a microwave modeling perspective. Inspired by the traditional modeling techniques for conventional MOSFETs, different strategies have been proposed over the last years to model the FinFET behavior at high frequencies. With the aim to support the development of this technology, a comparative study of the achieved results is carried out to gain both a useful feedback to investigate the microwave FinFET performance as well as a valuable modeling know-how. To accomplish a comprehensive review, all aspects of microwave modeling going from linear (also noise) to non-linear high-frequency models are addressed.

A comprehensive review on microwave FinFET modeling for progressing beyond the state of art

A measurement-based quasi-static nonlinear field-effect transistor (FET) model relying on an artificial neural network (ANN) approach and using real-time active load-pull (RTALP) measurement data for the model extraction is presented for an SOS-MOSFET. The efficient phase sweeping of the RTALP drastically reduces the number of large-signal measurements needed for the model development and verification while maintaining the same intrinsic voltage coverage as in conventional passive or active load-pull systems. Memory effects associated with the parasitic bipolar junction transistor (BJT) in the SOS-MOSFET are accounted for by using a physical circuit topology together with the simultaneous ANN extraction of: 1) the intrinsic FET current-voltage characteristics; 2) the intrinsic charges of the FET; and 3) the BJT dc characteristics, all from the same modulated large-signal RF data. The verification of the model using load-lines, output power, power efficiency, and load-pull, which is performed using two additional independent RTALP measurements, demonstrates that a reasonably accurate large-signal RF device model accounting for memory effects can be extracted from a single 10.5-ms RTALP measurement with a physically based ANN model.

/pdf/artificial-neural-network-model-of-sos-mosfets-based-on-48ff7oj729.pdf

Artificial Neural Network Model of SOS-MOSFETs Based on Dynamic Large-Signal Measurements

Microwave FinFET modeling based on artificial neural networks including lossy silicon substrate

An analytical extraction of the non-quasi-static nonlinear lookup table FET model at mm-wave frequencies is demonstrated in this study. Frequency dispersion present at mm-wave frequencies is compensated for by introduction of higher order dynamics of nonlinear constituent components. Nonlinear charge- and current-sources are used as the constituent components to represent the nonlinear device where the higher order charge- and current-sources are introduced to compensate for the non-quasi-static effects. The presented approach is validated using SOI Multi-Fin MOSFET transistors, the multiple gate MOSFET structures which have been introduced as an alternative candidate for bulk planar CMOS technology in the nanometer regime. An accurate multibias linear model is employed as the keystone for building the non-quasi-static nonlinear model. Large-signal measurements are used for model validation. Significant improvements are observed over the quasi-static nonlinear model which is demonstrated by excellent agreement between measurements and nonlinear model presented here.

Technology-Independent Non-Quasi-Static Table-Based Nonlinear Model Generation

A novel hybrid large-signal model of GaAs pseudomorphic HEMTs (pHEMTs) is proposed for monolithic microwave integrated circuit design. This new model is based upon accurate electromagnetic (EM) description and creative multi-path artificial neural network (ANN) optimization. To precisely describe the EM effect in the high-frequency range, the extrinsic part of this model includes both lumped and distributed components. In order to re-grid the discrete data, the bias-dependent intrinsic elements are determined by ANNs rather than traditional interpolations. The dispersion effect is represented by nonlinear sources with the multi-path-dependent integration technique, which is described by processed multi-bias S-parameters. This proposed approach can be applicable to different bias conditions, which is also verified by different types of GaAs pHEMTs with good agreement. In addition, a class-AB Ka-band power amplifier and a Ka-band switch using a 0.15- $\mu {\text {m}}$ GaAs pHEMT process were designed based on the novel hybrid model for further practical verification.

A Novel 4-D Artificial-Neural-Network-Based Hybrid Large-Signal Model of GaAs pHEMTs

A comparison between two different, yet analytical approaches for lookup-table large-signal model construction is presented in this study. Both approaches use the same linear (small-signal) model as the starting point to construct the large-signal model. The first nonlinear model employs the charge- and current-sources to represent the nonlinearity of the active part of the device where the extended charge- and current-sources are added to compensate for the non-quasi-static (NQS) effects at mm-wave frequencies. The second nonlinear model employs the small-signal model schematic and adapts it to be used directly for large-signal simulations. The study was performed on SOI multi-Fin MOSFET transistors. Both modeling approaches can reproduce the device behavior at microwave and mm-wave frequencies and under highly nonlinear conditions with good accuracy.

Evaluation of lookup-table non-quasi-static nonlinear models at microwave and mm-wave frequencies

The Direct RF-to-Digital $\Delta\Sigma$ receiver has emerged as an attractive solution for multi-band multi-standard wireless applications. This architecture is a direct RF to baseband digitizer with RF feedback from baseband to RF stages. The presence of the RF blocks such as the low noise amplifier and down-conversion mixer inside the loop filter significantly relaxes their linearity requirements and also improves the noise transfer function of the loop filter. However, the RF feedback poses several challenges, especially making the design methodology more complex. In contrast to classical receiver architectures, the blocks of the Direct RF-to-Digital $\Delta\Sigma$ receiver cannot be designed, analyzed, and simulated independently. Besides, conventional design methodologies for $\Delta\Sigma$ Modulators cannot be applied for this kind of receiver due to the presence of the RF blocks and the RF feedback inside the loop. This paper presents a new design methodology suited for this family of receivers. The proposed methodology covers simulation approaches, system design, and non-ideality impacts on the receiver. A design example of a multi-band multi-standard receiver for GSM/WCDMA/LTE is considered throughout the paper to demonstrate this design methodology.

System Design for Direct RF-to-Digital $\Delta\Sigma$ Receiver

A direct analytical extraction of a non-quasi-static nonlinear table-based FET model at mm-wave frequencies is demonstrated in this study. It makes use of extended charge and current sources at both gate and drain terminals to account for the frequency dispersion behavior of transistors at mm-wave frequencies. The model is validated using silicon FinFET transistors. Excellent agreement is achieved between measurements and model simulation revealing significant improvements over quasi-static nonlinear model results.

Non-quasi-static nonlinear model for FinFETs using higher-order sources

This paper focuses on the implementation of table-based models of high-frequency transistors for time-domain simulators at microwave and mm-wave frequencies. In this frequency range, the channel is not capable of responding to the excitation instantaneously therefore, a delay-time exists between the channel response and the channel excitation. This delay is represented by a complex trans-conductance in terms of circuit elements. The high-frequency models of transistors are required to have the implementation of complex trans-conductance, where the complex part accounts mathematically for the delay-time between the channel response and the channel excitation. This paper presents simple and accurate approaches to incorporate the complex trans-conductance in both small-signal and large-signal table-based models for time-domain simulators (MOS-AK International Meeting. Eindhoven, Netherlands, April 2008). Implementation approach for each model, small-signal and large-signal, is presented in separated sections. In the first step, the delay is realized by the introduction of an ideal transmission line between the channel excitation and the channel response. As transmission lines are not generally suitable for time-domain simulations, a lumped element equivalent network is introduced in the second step. The latter approach is fully compatible with time-domain simulators but frequency limitation, determined by the delay-time value itself, is introduced. Then the implementation of the complex trans-conductance in large-signal model is introduced. In terms of large-signal behavior, delay-time is important to achieve a non-quasi static model. Yet again there is limitation in terms of the frequency range that is determined by the delay value itself. The methodology is illustrated on the small-signal and the large-signal equivalent circuit of a Multi-Fin MOSFET transistor. Simulations are carried out by Cadence Spectre and Agilent ADS simulators, and comparisons are carried out between the simulation results and the measurements. Copyright © 2010 John Wiley & Sons, Ltd.

Simple and accurate approaches to implement the complex trans-conductance suited for time-domain simulators for small-signal and large-signal table-based models

Seyed Majid Homayouni

Papers

Technology-Independent Non-Quasi-Static Table-Based Nonlinear Model Generation

Evaluation of lookup-table non-quasi-static nonlinear models at microwave and mm-wave frequencies

System Design for Direct RF-to-Digital $\Delta\Sigma$ Receiver

Non-quasi-static nonlinear model for FinFETs using higher-order sources

Simple and accurate approaches to implement the complex trans-conductance suited for time-domain simulators for small-signal and large-signal table-based models