Noise Figure Performance Research Articles

CMOS Tunable Channel-selection LNA Employing Active Feedback Technique and Gain-boosted N-path Bandpass Filter for Advanced Cellular Applications

In this paper, a CMOS tunable channel-selection low-noise amplifier (LNA) that employs an active feedback technique and gain-boosted N-path bandpass filter (BPF) is presented for advanced cellular applications. The proposed LNA achieves broadband input power matching and low noise figure (NF) performance by using the active feedback technique. The gain-boosted N-path BPF is also used to implement a high-Q RF fourth-order bandpass filtering. Simulated in a 65-㎚ CMOS process, the proposed LNA achieves a maximum voltage gain of 17 ㏈, minimum NF of 2.16 ㏈, maximum out-of-band blocker rejection ratio of 21 ㏈ at 80 ㎒ offset frequency. The center-frequency tuning range of the LNA is 0.1 ‒ 4 ㎓, which includes all FDD bands in the 3G/4G/5G sub-6 ㎓ cellular standards. It draws a DC bias current of 21 ㎃ from a supply voltage of 1.2 V. The active die area is 0.8 ㎟.

Investigation of C-band pumping for extended L-band EDFAs

In this study, we present systematic numerical and experimental analysis of high-power C-band light pumping in extended L-band erbium-doped fiber amplifiers (EDFAs). We investigate, for the first time to the best of our knowledge, how C-band light sources can be used as pump sources to extend the bandwidth of L-band EDFAs beyond 1610 nm. Results show that, when using a C-band light source as the sole pump, efficient amplification is obtained over the extended L-band, but at the expense of a higher noise figure. However, the advantage of C-band pumping in terms of power conversion efficiency can be exploited when using a two-stage EDFA, with a first stage pumped by 1480 nm to maintain good noise figure performance and a high-power C-band light source (up to several hundred mW) as the pump source for the second stage. Thus, a 20 dB gain covering 1570–1618 nm with a maximum noise figure of 5.7 dB is demonstrated.

Cell Weighting and Gate Inductive Peaking Techniques for Wideband Noise Suppression in Distributed Amplifiers

This paper presents a noise suppression method to improve noise figure (NF) performance of a distributed amplifier (DA) in a broad frequency range. In the proposed DA topology, the input matching resistor, which remarkably increases the NF, is replaced by a common-gate (CG) stage. The noise contribution of the CG terminating network is suppressed by employing a cell weighting plan in conjunction with a gate-inductive peaking technique. Analytical methods and a circuit implementation in a $0.18~\mu \text{m}$ CMOS technology verify that the proposed noise suppression technique provides lower NF compared with the conventional DA (CDA) topology in the entire frequency range. This is in contrast to other reported low-noise DA topologies which decrease the noise contribution of the input matching element only in a limited range of frequencies. The measurement results demonstrate an average gain of 9.5 dB in 1-20 GHz frequency range, a noise figure less than 4 dB and 5.8 dB up to 16 GHz and 20 GHz, respectively, and an average IIP3 of 5.3 dBm. The chip occupies an area of 1.2 mm2, including pads, and consumes a current of 24 mA from 1.8 V supply.

Gain and noise figure performance of an EDFA pumped at 980 nm or 1480 nm for DOFSs

Erbium-doped fiber amplifier (EDFA) is widely recognized as one of the most important optical amplifiers in distributed optical fiber sensors which has been able to compensate the loss of the optical power. The 980 nm and 1480 nm are the most efficient pumps for amplifying the 1550 nm signal in EDFA. This paper presented an investigation for influences of pump power, signal power and EDF length on the gain and noise figure in an EDFA which pumped by 980 nm and 1480 nm pumps. It is also a comparison for each of the created gain and noise figure at the forward, backward, and bi-directional pumping configurations, separately. The effects of Rayleigh scattering, temperature, homogeneous interactions between Erbium ions and the amplified spontaneous emission are considered. As a result, Bi-directional pumping shows the highest gain value, and forward pumping presents the lowest noise figure for each pump. Moreover, backward pumping creates the highest noise figure and forward pumping generated the lowest gain value. The gain of 980 nm pump is higher than 1480 nm up to 5-m EDF length and for longer lengths, it becomes vice versa. The 980 nm pump caused a lower noise figure than 1480 nm for EDF length up to 10 m and it becomes vice versa for longer lengths.

Active Positive Sloped Equalizer for X-Band SiGe BiCMOS Phased Array Applications

This brief presents an active equalizer circuit with positive gain slope at X-Band (8–12 GHz). Compared to passive examples, the active equalizer realized better filter and impedance characteristics in frequency of interest with increased functionality for a single amplification stage. It achieved close to 10 dB of peak gain, a +1.13 dB/GHz gain slope with 2.8 dB noise figure (NF) by utilizing cascode topology. The design reaches a −1.5 dBm input-referred compression point (input-P1dB) while consuming 46 mW of power. To the best of authors’ knowledge, the presented work achieves the best on-chip gain, a gain slope and NF performance in the literature as an equalizer that utilizes SiGe BiCMOS technology.

Ultra-Low Noise Amplifier for X-Band SiGe BiCMOS Phased Array Applications

This brief presents, a low noise amplifier (LNA) with sub-1dB noise figure (NF) at X-Band. Different than generally known noise-and-power match technique, the presented LNA is designed by considering impedance between the base-collector terminals of an HBT in common-emitter configuration. The presented amplifier stage can achieve sub-1 dB NF performance with ~10 dB gain. The LNA dissipates 19.8 mW of dc power and has 0.77 dB NF while occupying 0.4 mm2. It succeeds 1.5 dBm of input-referred compression point. To the best of authors’ knowledge, the presented work achieves the best NF performance in the literature of any LNA that utilizes SiGe technology.

Design and Analysis of a Broadband Current-Mode CMOS Direct-Conversion Receiver Frond-End Circuit

A broadband (0.8–5[Formula: see text]GHz) CMOS current-mode direct-conversion receiver has been integrated in a 0.18-[Formula: see text]m CMOS process. The proposed receiver front-end features a broadband active-balun low-noise transconductance amplifier (LNTA) driving a current-mode passive mixer terminated by a low-input-impedance transimpedance amplifier (TIA). The receiver chain has improved robustness to out-of-band interference, conversion gain and outstanding linearity. With the technique of noise and distortion cancellation which performs a better input impedance matching, we employ a broadband common-gate–common-source (CG–CS) LNTA and a current mirror to improve both gain and noise figure (NF) performance. Compared to the 50% duty-cycle switching stage, the 25% duty-cycle I–Q switching stage is implemented by using serial switches driven by 50% quadrature local oscillator (LO) signals separately, which improves the down-conversion gain by 3[Formula: see text]dB and lowers the noise figure. The transimpedance amplifier employs the [Formula: see text]-boosting technique to realize low input impedance and high transimpedance gain. The core circuit (RF and baseband signal path) consumes 26[Formula: see text]mW, and the prototype receiver achieves approximately 33–34.5-dB conversion gain, 8.1–9.35-dB NF and 7.5–9.8-dBm IIP3 from 0.8[Formula: see text]GHz to 5[Formula: see text]GHz.

Noise Performance Improvement of Broadband Discrete Raman Amplifiers Using Dual Stage Distributed Pumping Architecture

We demonstrate a novel design of dual stage backward pumped broadband discrete Raman amplifier with distributed pumping architecture which improves the short wavelength noise performance compared with conventional backward pumped single stage design. Overall noise figure (NF) performance has been improved by 2.8 dB over 70 nm bandwidth without introducing forward pumping. Nonlinear phase shift has been numerically calculated and compared with single stage design to evaluate the overall performance improvement. Finally, we report that our proposed dual stage design with ∼14 dB net gain improves low wavelength NF by 3.3 dB providing 1.2 dB Q2 factor improvement and 1134 km reach extension versus a conventional single stage design for 18 × 120 Gb/s PM-QPSK transmission.

Configurable GPS/GNSS Antenna Module Resistant to RFI Saturation

With GPS/Global Navigation Satellite System (GNSS) receivers exposed to greater levels of Radio Frequency Interference (RFI), a potential problem is the saturation of the receiver front end (FE). This problem is further complicated in typical multifrequency receivers by the interfrequency saturation effect. Specifically, any inband RFI targeted to induce FE saturation at only one specific GPS/GNSS frequency, if not properly handled, would potentially impact the reception of the other frequencies. This paper presents the design of an antenna module to detect, identify, and isolate potential RFI to prevent FE saturation including that due to the interfrequency effect. Analysis showed that any specific antenna module configuration with fixed internal components must sacrifice noise figure (NF) performance to increase robustness to RFI saturation, and vice versa. To provide a compact solution to this dilemma and the interfrequency saturation issue, two dynamically configurable antenna modules based on the concept of network topology were proposed for typical dual-frequency GPS/GNSS receivers. Such a solution can adapt to different RFI conditions by operating in corresponding modes resulting in a better NF versus robustness tradeoff. The proposed antenna-module design was validated by experiments with live GPS signals under controlled RFI conditions.

40 dB gain all fiber bismuth-doped amplifier operating in the O-band

In this Letter, we investigate and compare the gain and noise figure characteristics of bismuth (Bi)-doped fiber amplifiers configured in both single and double signal pass implementations. A maximum gain of 25dB and a noise figure of 4dB is measured at 1360nm in the single pass configuration for -23 dBm input signal power, whereas in the double pass configuration the gain of the amplifier is improved significantly by 14dB allowing us to achieve a gain of 39dB with a noise figure of 5dB. To the best of our knowledge, this is the highest gain reported to date using Bi-doped fiber as a gain medium. Furthermore, we also study the gain and noise figure dependency on pump power, signal power, and pump wavelength for the double pass amplifier configuration. We observed similar gain and noise figure performance in the double pass configuration to that of the single pass configuration but with the benefit of less pump power and a shorter length of the Bi-doped fiber.

Sub-1-dB and Wideband SiGe BiCMOS Low-Noise Amplifiers for <inline-formula> <tex-math notation="LaTeX">$X$ </tex-math> </inline-formula>-Band Applications

In this paper, a design methodology for SiGe HBT-based low-noise amplifiers (LNAs) is proposed that can be utilized for both sub-1-dB noise figure (NF) and wide bandwidth as an alternative to the conventional simultaneous input noise and power matching technique. This paper focuses on the removal of the series base inductor and the inclusion of an output matching network as a design parameter that can be used to achieve both sub-1dB NF and wide bandwidth. The effects of the mentioned design parameters on NF and bandwidth are described and analyzed, including the comparison of results with the conventional design technique. To demonstrate the validity of the analysis and impact of utilizing base-to-collector capacitance on LNA performance metrics, two different LNAs implemented in a 0.13- $\mu \text{m}$ SiGe technology. The first LNA achieves a lower than 1-dB NF with 26-dB gain at 8.5 GHz, and the second LNA exhibits wideband operation, with a better than 10-dB return losses in the range of 8–35 GHz and lower than 3-dB NF from 6 to 20 GHz. The sub-1-dB LNA reaches −17.3 dBm of input-referred compression point (IP1dB), with a power consumption of 62 mW and an area of 0.795 mm2. The wideband LNA achieves 17.6 dB of peak gain with a 48.5-mW power consumption and 0.69-mm2 chip area. To the best of our knowledge, this paper achieves the best NF performance in the literature utilizing a SiGe technology, while the wideband LNA exhibits the best operational bandwidth, together with a reasonable low-noise performance.

A SiGe HBT $D$ -Band LNA With Butterworth Response and Noise Reduction Technique

This letter presents a wideband high-gain four-stage cascode $D$ -band low noise amplifier (LNA) implemented in a 0.13- $\mu \text{m}$ SiGe BiCMOS technology. A shunt inductor that is used at the intermediate node of the cascode topology to reduce the noise contribution of the common base transistor is analyzed and employed for the first time for the SiGe HBT technology. Furthermore, the staggered-tuning technique based on the Butterworth distribution is utilized to have a wideband flat-gain characteristic. The designed LNA has a measured 32.6-dB peak gain at 144.5 GHz with a 3-dB bandwidth of 52 GHz. The measured noise figure (NF) is lower than 6.1 dB across the whole $D$ -band, and its minimum value is 4.8 dB. The total dc power consumption is 28 mW. To the authors’ best knowledge, these results demonstrate the lowest NF performance and the widest 3-dB bandwidth among the $D$ -band LNAs implemented in the silicon-based technologies. The effective chip area excluding the pads is 0.6 mm2, and the total IC occupies an area of 1 mm2.

Principle and noise performance of optical phase arithmetic devices using four wave mixing

The existing theoretical equations cannot provide an excellent guidance for developing four-wave mixing (FWM)-based optical logic devices, though the experiments have been done in several researches. The optimization of noise figure performances of such devices should be further investigated. In the paper, the universal analytic expressions for the amplitude and phase of the idler in degenerate or non-degenerate FWM process under pump depletion are derived in detail from the nonlinear coupled-mode equations for guiding optical waves propagation in highly nonlinear fiber. The universal analytic expressions are obtained by the first-and the third-kind of elliptic integrals. By using equivalent infinitesimal to calculate the limit of phase sensitive amplification, we find out the initial phase relationship between the idler and the input guided wave for phase-independent amplification, which is crucially important for explaining the operating principles of the FWM-based adder and subtracter. As an example, the configuration of non-degenerate FWM-based hybrid arithmetic device with three logic functions of A+B-C, A+C-B, and B+C-A for QPSK signals is presented, and then the noise transfer characteristics in terms of signal-to-noise ratio (SNR) and error vector magnitude (EVM) are taken into account by adjusting the fiber length, input wavelength, and optical power. The calculation results show as follows. 1) This kind of arithmetic device has a noise figure of about 1.1 dB and an input SNR of more than 24 dB is necessary for the symbol error rate of 10-3 without forward error correction, corresponding to an output EVM of 23.2%. 2) The length of highly nonlinear fiber used in the hybrid arithmetic device may be taken flexibly, provided that the variation of FWM conversion efficiency is controlled in a range of 1 dB relative to the maximum, with an EVM fluctuation of less than for the idlers. 3) The hybrid arithmetic device has an operating optical bandwidth of about 16 nm for the SNR degradation of 1.3 dB. 4) The output EVM increases with the increase of input power, and the allowable input power should be no more than 100 mW for an input SNR of 28 dB, noting that the larger the input SNR, the higher the allowable input power is.

Noise Figure Degradation in Balanced Amplifiers

Balanced amplifiers suffer from the noise figure performance in comparison with a stand-alone amplifier even with an ideal input divider. The noise parameters degrade further with an imperfect divider. We present exact and approximate analytical results for the noise figure and noise parameters of the balanced amplifier in divider topology. Measurement results are also presented as a verification.

Design of Low Noise, Flat Gain CMOS-based Ultra-wideband Low Noise Amplifier for Cognitive Radio Application

ABSTRACT Cognitive radios are smart Radio Frequency (RF) communication system that senses the RF spectrum which allocates a less crowded spectrum to the user abstaining the interference of other spectrum users while confirming to Federal Communications Commission regulation. In this work, the design of an ultra-wideband low noise amplifier (50 MHz–10 GHz) for cognitive radio application adapting noise cancelling and gm -boosting techniques is presented. With the integrated noise cancelling and gm -boosting technique, the proposed low noise amplifier is able to achieve an ultra-wideband frequency response in reference to the gain and noise figure performance. Simulation result shows that the input and output matching are better than −10 dB and −7.1 dB with a gain of 11.5 ± 1.5 dB, noise figure of 3.5 ± 0.5 dB and a RF input third order intercept point ranging from −10.4 to −5.3 dBm in a frequency span from 50 MHz to 10 GHz on a complementary metal-oxide semiconductor (CMOS) 0.13 μm platform. The designed architecture occupies 1.19 mm2 chip area and consumes 24.2 mW of DC power from 1 V of DC supply headroom.

60 GHz-Band Low-Noise Amplifier

This paper presents the design of a 60[Formula: see text]GHz-band LNA intended for the 63.72–65.88[Formula: see text]GHz frequency range (channel-4 of the 60[Formula: see text]GHz band). The LNA is designed in a 65-nm CMOS technology and the design methodology is based on a constant-current-density biasing scheme. Prior to designing the LNA, a detailed investigation into the transistor and passives performances at millimeter-wave (MMW) frequencies is carried out. It is shown that biasing the transistors for an optimum noise figure performance does not degrade their power gain significantly. Furthermore, three potential inductive transmission line candidates, based on coplanar waveguide (CPW) and microstrip line (MSL) structures, have been considered to realize the MMW interconnects. Electromagnetic (EM) simulations have been performed to design and compare the performances of these inductive lines. It is shown that the inductive quality factor of a CPW-based inductive transmission line ([Formula: see text] is more than 3.4 times higher than its MSL counterpart @ 65[Formula: see text]GHz. A CPW structure, with an optimized ground-equalizing metal strip density to achieve the highest inductive quality factor, is therefore a preferred choice for the design of MMW interconnects, compared to an MSL. The LNA achieves a measured forward gain of [Formula: see text][Formula: see text]dB with good input and output impedance matching of better than [Formula: see text][Formula: see text]dB in the desired frequency range. Covering a chip area of 1256[Formula: see text][Formula: see text]m[Formula: see text]m including the pads, the LNA dissipates a power of only 16.2[Formula: see text]mW.

Design Exploration of mm-Wave Integrated Transceivers for Short-Range Mobile Communications Towards 5G

This paper presents a design exploration, at both system and circuit levels, of integrated transceivers for the upcoming fifth generation (5G) of wireless communications. First, a system level model for 5G communications is carried out to derive transceiver design specifications. Being 5G still in pre-standardization phase, a few currently used standards (ECMA-387, IEEE 802.15.3c, and LTE-A) are taken into account as the reference for the signal format. Following a top-down flow, this work presents the design in 65[Formula: see text]nm CMOS SOI and bulk technologies of the key blocks of a fully integrated transceiver: low noise amplifier (LNA), power amplifier (PA) and on-chip antenna. Different circuit topologies are presented and compared allowing for different trade-offs between gain, power consumption, noise figure, output power, linearity, integration cost and link performance. The best configuration of antenna and LNA co-design results in a peak gain higher than 27[Formula: see text]dB, a noise figure below 5[Formula: see text]dB and a power consumption of 35[Formula: see text]mW. A linear PA design is presented to face the high Peak to Average Power Ratio (PAPR) of multi-carrier transmissions envisaged for 5G, featuring a 1[Formula: see text]dB compression point output power (OP1dB) of 8.2[Formula: see text]dBm. The delivered output power in the linear region can be increased up to 13.2[Formula: see text]dBm by combining four basic PA blocks through a Wilkinson power combiner/divider circuit. The proposed circuits are shown to enable future 5G connections, operating in a mm-wave spectrum range (spanning 9[Formula: see text]GHz, from 57[Formula: see text]GHz to 66[Formula: see text]GHz), with a data-rate of several Gb/s in a short-range scenario, spanning from few centimeters to tens of meters.

The effect of ASE reinjection configuration through FBGs on the gain and noise figure performance of L-Band EDFA

In this study, we present the gain and noise figure performance improvement in L-band erbium-doped fiber amplifier (L-EDFA) provided by amplified spontaneous emission (ASE) reinjection through different configurations of 1533nm band FBGs. The experimental results are compared with a single-stage bidirectionally pumped conventional L-EDFA design. It is shown that when the forward and/or the backward ASE noise is partly reinjected to L-EDFA using a double/single 1533nm fiber Bragg gratings (FBG), the gain and noise figure performance of L-EDFA increases depending on the FBG configuration. The best gain and NF performance in our L-EDFA was achieved by reinjection of forward and backward ASE through FBG1 and FBG2 leading to an 4.5dB increase in gain and 1dB decrease in NF at 1585nm and −30dBm input signal power. The results show that both FBGs must be used at the same time to improve gain and NF performance in L-band EDFAs.

Advances in Ferrite Redundancy Switching for Ka-Band Receiver Applications

Previously limited to simple redundancy switching schemes, such as 2-for-1, a ferrite-based equivalent of the traditional 10-for-8 mechanical redundancy switch (R-switch) was developed. Standby redundancy of active components, such as low-noise amplifiers (LNAs), is essential for space missions, and the demand for Ka-band communications satellites is rapidly increasing. Advances in the reduction in size, mass, and insertion loss of higher complexity ferrite circulator switch networks have made ferrite switches a feasible alternative to mechanical switches for redundancy switching applications. This work set out to understand how these advances could be applied to a higher level of redundancy switching, such as 10-for-8, in order to meet the growing needs of the Ka-band satellite market. A prototype redundant LNA assembly was designed and manufactured. The results herein show that the reduced-size ferrite switch networks provide considerable bandwidth and insertion loss improvements over traditional ferrite switch networks. The 10-for-8 redundant LNA assembly, based on ferrite switching and developed by Honeywell (formerly EMS Technologies) provided excellent noise-figure performance of 2.1 dB at 25 °C in Ka-band. More significantly, the high level of integration in the ferrite switch rings and in the interface between the ferrite switches and LNAs results in size and mass savings of 50% over the mechanical R-switch approach, constructed of individual components.

Performance Analysis of an EDFA Utilizing a Partially Doped Core Fiber (PDCF)

AbstractThe effect of transversal design in Erbium-doped fiber amplifiers’ gain and noise figure performance is illustrated in this work. In this work, we investigate experimentally a single pass 980 nm pumped EDFA with partially doped Erbium core fiber (PDCF), which has the core partially doped with Erbium ions. Later, the enumerated results for PDCF are compared with a standard fully doped EDF, having similar Erbium ion doping concentration. The PDCF Amplifier gain and noise figure performance is studied against different pump power and signal power at different operating wavelengths. The noise figure indicates improvement due to reduced spontaneous emission from un-doped region of the core.