2022年9月JSSC期刊：高采样率接收器与射频前端技术

JSSC

需积分: 10 181 浏览量更新于2024-06-28 收藏 111.28MB PDF 举报

身份认证购VIP最低享 7 折!

领优惠券(最高得80元）

"JSSC 2022.09期刊包含了多篇关于高级电子技术的文章，专注于高速采样、噪声消除、频率锁定、低功耗接收器以及射频前端设计等领域。" 在2022年9月的《集成电路科学与工程》(JSSC)期刊中，发表了多篇关于电子电路和系统的创新研究论文，这些论文涵盖了从射频(RF)到微波频率范围内的关键技术。以下是几篇重点文章的概述： 1. **时间交织切换发射跟随器用于扩展前端采样率至200 GS/s** 这篇文章由P. Thomas等人撰写，讨论了一种新型的时间交织切换发射跟随器技术，该技术能够显著提升前端采样速率，达到200 GS/s的高速数据处理能力。这种技术对于高速信号处理和高速数据传输系统具有重要意义。 2. **一种被动宽带噪声消除混合器优先架构** H. Bialek等人提出了一种创新的架构，它能够共享天线接口，同时实现干扰容忍的唤醒接收器和低噪声主接收器。这种架构通过噪声消除技术增强了系统在复杂无线环境中的性能。 3. **全集成490-GHz CMOS接收器采用基于双锁定接收器的FLL** K.-S. Choi等人的工作展示了如何采用双锁定接收器为基础的频率锁定环(FLL)来实现490 GHz的高频率接收器。这在毫米波通信和高分辨率成像应用中具有潜在的应用价值。 4. **适用于蜂窝和WiFi应用的0.4-6GHz接收器** H. Razavi和B. Razavi设计了一个涵盖0.4到6GHz频率范围的接收器，旨在满足移动通信和无线局域网的需求。该接收器具有宽频率覆盖和高效能的特点。 5. **164-μW 915-MHz次采样相位跟踪零中频接收器** Y. Guo等人提出的低功耗接收器专为短距离应用设计，其数据传输速率达到5 Mb/s。这一设计结合了次采样技术和相位跟踪技术，实现了低功耗与高性能的平衡。 6. **电流复用正交RF接收前端Blixator电路** M. Barzgari等人介绍的Blixator电路是一种电流复用的正交RF接收前端，特别适合于低功耗应用。该设计通过电流复用来优化能源效率，同时保持良好的射频性能。 7. **W波段可扩展的2×2相位阵列发射机** 文章中还提到了一个W波段的可扩展2×2相位阵列发射机设计，这在高频率无线通信和雷达系统中具有潜在的应用，展示了阵列天线技术的最新进展。这些研究展示了当前集成电路领域的前沿技术，包括高速采样、噪声抑制、高效能接收器设计以及射频阵列技术，这些都将对未来无线通信、雷达探测和数据传输等领域产生深远影响。

资源详情

资源推荐

2610 IEEE JOURNAL OF SOLID-STATE CIRCUITS, VOL. 57, NO. 9, SEPTEMBER 2022

Markus Grözing (Member, IEEE) received the

Dipl.-Ing. and Dr.-Ing. degrees from the University

of Stuttgart, Stuttgart, Germany, in 2000 and 2007,

respectively.

During his doctoral study, his research was focused

on phase noise and jitter in ring oscillators and

mixed-signal CMOS circuit design. Since 2007,

he has been leading the Integrated Circuit Design

Group, Institute of Electrical and Optical Communi-

cations Engineering, University of Stuttgart. He has

authored or coauthored more than 100 scientiﬁc

papers. His current research topics are analog and mixed-signal circuits and

neural networks, high-speed transmitters and receivers for serial links, and

gigasample-range analog-to-digital and digital-to-analog converters in CMOS

and BiCMOS technologies.

Dr. Grözing has served as a Reviewer for the IEEE T

RANSACTIONS ON

MICROWAVE THEORY AND TECHNIQUES (T–MTT), the IEEE JOURNAL OF

SOLID-STATE CIRCUITS (JSSC), the IEEE TRANSACTIONS ON CIRCUITS

AND

SYSTEMS—II: EXPRESS BRIEFS (TCAS–II), the IEEE TRANSAC-

TIONS ON VERY LARGE SCALE INTEGRATION (VLSI) SYSTEMS,andthe

IEEE S

OLID-STATE CIRCUITS LETTERS (SSC-L).

Manfred Berroth (Senior Member, IEEE) received

the Dipl.-Ing. degree in electrical engineering

from the University of the Federal Armed Forces,

Munich, Germany, in 1979, and the Dr.-Ing. degree

from Ruhr-Universität Bochum, Bochum, Germany,

in 1991.

From 1980 to 1986, he was engaged in radar air

trafﬁc services and was the head of a communication

center. In 1987, he joined the Fraunhofer Institute

for Applied Solid State Physics (IAF), Freiburg im

Breisgau, Germany, where he was involved with

device modeling of heterojunction ﬁeld-effect transistors and design of analog,

digital, and mixed-signal circuits in the gigahertz-frequency range. In 1991,

he became the Head of the Department for Development of Devices and

Circuits, and in 1995, the Deputy Director of IAF. Since October 1996,

he has been holding the Chair for Electrical and Optical Communications

Engineering, University of Stuttgart, Stuttgart, Germany. He has authored or

coauthored more than 200 papers and has given more than 20 presentations

at international conferences. His research interests are modeling and charac-

terization of electronic and opto-electronic devices and integrated circuits.

Dr. Berroth has been a Reviewer for international publications, such as the

Electronic Letters (IEE), the IEEE T

RANSACTIONS ON MICROWAVE THE-

ORY AND TECHNIQUES (T-MTT), the IEEE TRANSACTIONS ON CIRCUITS

AND

SYSTEMS—I: REGULAR PAPERS (TCAS–I), the IEEE TRANSACTIONS

ON CIRCUITS AND SYSTEMS—II: EXPRESS BRIEFS (TCAS–II), and the

IEEE T

RANSACTIONS ON ELECTRON DEVICES (T–ED).

IEEE J OURNAL OF SOLID-STATE CIRCUITS, VOL. 57, NO. 9, SEPTEMBER 2022 2611

A Passive Wideband Noise-Canceling Mixer-First

Architecture With Shared Antenna Interface for

Interferer-Tolerant Wake-Up Receivers and

Low-Noise Primary Receivers

Hayden Bialek , Graduate Student Member, IEEE, Ali Binaie , Graduate Student Member, IEEE,

Sohail Ahasan

, Graduate Student Member, IEEE, Kamala Raghavan Sadagopan , Member, IEEE,

Matthew L. Johnston , Senior Member, IEEE, Harish Krishnaswamy , Senior Member, IEEE,

and Arun Natarajan

, Senior Member, IEEE

Abstract—Wake-up receivers (WuRXs) present an opportu-

nity to reduce the average power consumption of Internet-of-

Things (IoT) transceivers; however, achieving sensitivity and

interferer tolerance, providing wideband matching, and shar-

ing an antenna interface present a signiﬁcant challenge for

existing architectures. This article presents a primary/WuRX

which utilizes a quadrature hybrid coupler-based N-path mixer

ﬁrst architecture to simultaneously achieve low noise, wideband

matching, and a shared antenna interface. The passive-mixer

ﬁrst approach and a two-code modulated multi-tone signaling

scheme provide interferer tolerance in the WuRX. This article

analyzes gain/power tradeoffs in the proposed architecture in

the context of noise impact with multi-tone WuRX signaling.

The proposed architecture is implemented in 65-nm CMOS and

occupies 2.25 mm

. The primary RX achieves 3.8-dB NF and

0.75-dBm out-of-band P1dB with 440-µW power consumption.

The WuRX achieves −86-dBm sensitivity for 10-kb/s data rate

andupto−40-dB signal-to-interferer ratio (SIR) with 171-µW

power consumption.

Index Terms— Interferer-tolerant, Internet-of-Things (IoT),

industrial, scientiﬁc, and medical (ISM), low-noise, low-power,

N-path, RF, receiver, wake-up.

I. INT RODUCTION

HE Internet of Things (IoT) has emerged as a pervasive

inﬂuence in our daily lives [1], [2]. Widespread IoT

adoption requires always-

ON, low-power devices that can

communicate wirelessly [3]. In order to achieve low latency

and low power, a transceiver (TRX) can have an always-

Manuscript received 2 June 2021; revised 21 September 2021 and

23 November 2021; accepted 22 December 2021. Date of publication 1 March

2022; date of current version 26 August 2022. This article was appro ved by

Guest Editor Nagendra Krishnapura. This work was supported in part by

the NIH under Award R01EB028104 and in part by the NSF under Grant

1554720. (Corresponding author: Hayden Bialek.)

Hayden Bialek, Kamala Raghavan Sadagopan, Matthew L. Johnston, and

Arun Natarajan are with the School of Electrical Engineering and Com-

puter Science, Oregon State University, Corvallis, OR 97333 USA (e-mail:

bialekh@oregonstate.edu).

Ali Binaie, Sohail Ahasan, and Harish Krishnaswamy are with the Electrical

Engineering Department, Columbia University, New York, NY 10032 USA.

Color versions of one or more ﬁgures in this article are available at

https://doi.org/10.1109/JSSC.2022.3148088.

Digital Object Identiﬁer 10.1109/JSSC.2022.3148088

Fig. 1. Low-power RXs for always-ON applications require ultra-low-power

interferer-tolerant wake-up RX that can share antenna interfaces with primary

RX. Operation across wide frequency range is desirable.

active low-power wake-up receiver (WuRX) and a duty-cycled

primary receiver (RX) [4] where the WuRX has low latency

and low power and is sensitive (see Fig. 1). In addition,

given operation in the sub-GHz industrial, scientiﬁc, and

medical (ISM) band, it is desirable for the WuRX to be

operational across allocations in different g eographies while

achieving tolerance to interference from existin g applications.

Finally, since many IoT devices are area constrained, a shared

antenna interface for the Wu RX radio and primary TRX is

desirable.

See https://www.ieee.org/publications/rights/index.html for more information.

2612 IEEE JOURNAL OF SOLID-STATE CIRCUITS, VOL. 57, NO. 9, SEPTEMBER 2022

Fig. 2. Proposed QHC-based N-path mixer-ﬁrst architecture with noise canceling and bottom-plate switching is utilized for both primary RX and WuRX.

Prior work in low-power wake-up radios can be classi-

ﬁed based on power consumption. Near-zero power radios

(sub μW) utilize resonant networks along with energy-detector

ﬁrst topologies [6]–[8]. In such WuRX, the sensitivity tends to

be low (>−70 dBm), and the high-Q resonant front ends are

used to provide voltage boost limit operating frequency range.

Uncertain-IF architectures downconvert with an unlocked local

oscillator (LO), leading to IF that is not precisely known [9].

Since g ain is achieved at IF as opposed to RF, high gain

is feasible prior to an envelope detector (ED) even with

low power (∼100 μW), improving sensitivity to <−80 dBm

[9]–[11]. Such mixer-ﬁrst schemes also enable LO-deﬁned

RF frequency tunability. While the uncertain-IF approach

eliminates power-hungry PLLs, it also requires wide RF

matching to account for LO uncertainty. Other techniques

proposed for low power consumption while maintaining

sensitivity/probability of false alarm include co-design of

antenna/rectiﬁer [8], frequency hopping [12], and duty

cycling [13]. Interferer tolerance in WuRX is required for

practical deployment in crowded ISM bands, and adaptive

RX [4] and multi-tone interferer-tolerant signaling [14] have

been proposed to improve interferer tolerance (∼signal-to-

interferer ratio (SIR) of −26 dB with power consumption

below 150 μW).

In this article, we present a passive, wideband noise-

canceling primary RX/WuRX that achieves interferer tolerance

and sensitivity in both the primary and WuRXs across a

wide operating range while sharing an antenna interface.

We have expanded on the work in [15], including a complete

description of the system architecture, analysis of the signal,

and noise propagation in a multi-tone ED system, and provided

implementation details of key blocks in the proposed system.

The organizatio n of this article is given as follows. Section II

describes the wide-band shared antenna interface front end

used for both primary and WuRXs. Section III provides the

background on code-domain modulation for wake-up radios

and details the implemented two-code wake-up signaling

scheme. Section IV describes the WuRX architecture and

provides an analysis of the noise and signal propagation in

this system. Section V details circuit implementations for key

blocks in the primary and wake-up RXs. Section VI p resents

measurement r esults. Section VII presents conclusions and

avenues for future research.

II. W

IDEBAND SHARED ANTENNA

INTERFACE RX FRONT END

The work targets shared antenna interfaces for the pri-

mary TRX and WuRX while supporting operation across

∼700–900 MHz. Wideband matching is achieved using a

quadrature-hybrid coupler (QHC)-based mixer-ﬁrst primary

RX/WuRX architecture shown in Fig. 2. The antenna input

is applied to the IN port of the QHC; the THRU and CPL

ports are each connected to N-path mixer with botto m-plate

switching; and the ISO port is terminated in 50  (see

Fig. 2). The primary and WuRXs adopt an N-path mixer-ﬁrst

architecture, and both sets of mixer switches are connected to

the THRU and CPL ports of the coupler and can be activated

as required by enabling/disabling the quadrature LO driving

the mixers.

As detailed in the f ollowing, there are three critical fron-

tend features that enable low-power, low-noise performance:

wideband matching independent of switch sizing, noise can-

celing of the termination resistance, and passive voltage boost

through the QHC and bottom plate switching.

A. Wideband Matching

Matching at the ANT port in a QHC is achieved with

the ISO port termination resistor as long as the equal output

impedances are present at the THRU and CPL ports [16]–[18].

Fig. 3(a) shows the matching network provided by the QHC

with respect to a signal applied at the IN port of the QHC,

open-circuit terminations at the THRU/CPL ports, and a 50-

termination at the ISO port. Signals reﬂected from the THRU

and CPL ports appear with 180

◦

of phase difference following

the reﬂection, resulting in the reﬂections getting canceled at

the antenna interface. Reﬂections appearing at the ISO port

of the QHC are absorbed by the 50- resistor. In this work,

we utilize impedance translation in N-path mixers to achieve

an LO-deﬁned high impedance at THRU and CPL ports

[19], [20]. This provides a voltage boost due to the reﬂection

of the input signal. The proposed approach provides several

BIALEK et al.: PASSIVE WIDEBAND NOISE-CANCELING MIXER-FIRST ARCHITECTURE WITH SHARED ANTENNA INTERFACE 2613

Fig. 3. (a) Signal propagation for QHC with input signal applied at IN

and open circuits at THRU/CPL ports. (b) I

noise path for N-path implicit

capacitive stacking mixer with QHC demonstrating cancellation of termination

resistor noise [5].

beneﬁts with respect to input matching. First, since matching is

achieved using the QHC and a termination resistor, wideband

matching is feasible with a wideband QHC. Second, since

mixer switches do not need to provide small

ON-resistance

relative to 50  unlike previous works that relied on baseband

impedance translation [19]. Therefore, primary and WuRX

mixers are connected to the same node without impacting

the m atching as both mixers use relatively small switch

sizes.

B. Noise Canceling in QHC-Based Front End

The termination resistance can degrade the noise ﬁgure.

In general, noise canceling has been used in the prior art to

achieve low noise with wide bandwidths by canceling noise

from the resistor/common-gate terminations using auxiliary

ampliﬁers [21], [22]. In this RX, a noise-canceling scheme is

used in both primary and WuRXs to cancel the noise from

the matchin g resistor, similar to dual-path active ampliﬁer

architectures in [23], [24]. Fig. 3(b) shows the noise cancel-

lation in the I

path. With noise V

applied to the ISO port

of the QHC, the voltage seen at the THRU and CPL ports

n,thru

= j

√

n,cpl

√

. (1)

For the I

output, the signal is sampled by the ﬁrst and

fourth phases of the LO at THRU and CPL, respectively. This

introduces a phase shift of 0

◦

and 270

◦

to the signals seen at

THRU and CPL. The noise at the output of the I

path is

nout,I +

= v

n,thru

− jv

n,cpl

+ v

n,sw

= v

n,sw

. (2)

Therefore, the noise from the termination resistor is can-

celed at the output of the N-path mixer. In the following, the

noise contribution of the RX is due to the mixer switches.

The antenna resistance, R

,isdeﬁnedas50,andthe

switch resistance R

is based on a 65-nm CMOS model.

Fig. 4. Calculated versus simulated mixer noise ﬁgure for the proposed

QHC-based N-path mixer ﬁrst architecture.

The QHC IN and ISO terminals are ter minated in 50 .

In this conﬁguration, the equivalent impedance seen from

the THRU/CPL ports is >=50 . Therefore, a conservative

estimate of the shunt resistance can be calculated assuming

that the THRU/CPL ports present a 50- impedance for all

harmonics.

The N-phase passive mixing can be modeled using an

linear time-invariant (LTI) equivalent where the harmonic

re-radiation loss is represented using the shunt resistance

[25], where

= R



4γ

1 − 4γ

 4.3R



(3)

with R



= R

+ R

and γ = (2/π

) for a four-phase N-path

mixer.

The passive voltage gain, from ANT port to mixer input,

, can be shown to be

√

(

1 + 

)

(4)

where

 =

(

+ R

− R

)

+ R

Similar to the result in [25] and assuming perfect cancella-

tion of termination resistor noise, the mixer noise ﬁgure with

the p roposed architecture can be calculated as

F = 1 +



+ R



. (5)

The calculated NF is compared to simulated NF across

mixer switch sizes in Fig. 4. Good agreement is seen between

calculated a nd simulated results. An increase in switch size

leads to lower NF but higher power consumption in the LO

path. Therefore, 8 and 1 μm are selected as switch sizes in

the primary and WuRXs to balance noise performance against

power consumption.

C. Voltage Boost

The high impedance at the THRU and CPL ports results

in a 3-dB voltage boost relative to the input signal ampli-

tude V

. In addition, a bottom-plate switching N-path mixer

2614 IEEE JOURNAL OF SOLID-STATE CIRCUITS, VOL. 57, NO. 9, SEPTEMBER 2022

architecture is adopted, similar to [5] and [26], and the implicit

capacitive stacking mixer architecture in the quadrature N-path

mixers provides an additional 6-dB voltage boost, leading to

an overall ∼9-dB voltage boost from RF to single-ended IF

in the passi ve mixer (∼15-dB differential IF).

This approach decouples the matching and noise require-

ments set on the mixer, expanding the design space. First,

the QHC matching network provides matching independent

of mixer switch sizing. Second, termination resistor matching

does not come at the cost of reduced noise performance due to

noise canceling. Therefore, the proposed QHC-based approach

allows for reductio ns in mixer switch sizes, decreasing the

system power consumption.

This architecture is utilized for both primary and WuRXs

with scaled switches to tradeoff between noise ﬁgure and

power consumption (see Fig. 2). The performance and design

of the N-path mixer utilized in the WuRX are discussed

in detail in Section IV. The primary RX was designed to

achieve sub-4-dB NF with sub-500-μW power consumption.

This perf ormance metric is achieved due to the QHC passive

voltage boost, noise canceling, and bottom-plate capacitor

switching, which allows for low noise, relative to other low

power RXs [27], [28], in spite of the mixer-ﬁrst architecture.

The dominant source of power consumption for the primary

RX is the “windmill” divide-by -2 generating th e LO drive

for mixer switches [29]. While the implementation in 65-nm

CMOS achieves excellent sensitivity with <500-μWpower,

the digital- intensive architecture will beneﬁt from scaling

to more advanced technology nodes leading to substantial

reductions in power consumption.

III. W

AKE-UP RX SIGNALING

As noted in Section I, the objective of the wake-up

RX (WuRX) is to detect the wake-up sequence speciﬁc to the

RX and enable the primary TRX. Importantly, practical WuRX

must provide interferer tolerance while achieving low power

consumption, which presents a signiﬁcant design challenge.

In prior work, a passive high-Q RF ﬁlter has been used to

ﬁlter out-of-band interferers [7], [11], [12]. However, such

an approach limits the WuRX operating frequency range and,

hence, is not suitable for wideband/multi-band schemes. In the

case of slow-varying interferers, baseband output voltage due

to interferer power can be treated as an offset, and adaptive

offset cancellation techniques can mitigate interferer impact

on energy-detection

ON–OFF-keying (OOK) RX [8], [30].

However, an average signal-power-based offset-cancellation

loop does not track transient or wideband modulated inter-

ferers, limiting its applicability in ISM bands with congested

spectrum usage.

A. Prior Art in Multi-Tone Code-Domain Wake-Up Signaling

A spread-spectrum m ulti-tone or transmitted-reference

Wu RX signaling architecture that can provide interferer tol-

erance is proposed in [14] where the transmit signal for

the WuRX is shown in Fig. 5(a). The wake-up signal, S

is constructed with three tones spaced at  f . Two tones are

modulated with a data signal d(t) and code c

[see Fig. 5(a)].

Fig. 5. (a) Multi-tone signaling scheme (transmitted-reference) pro-

posed in [14]. (b) Proposed multi-tone signaling for enhanced interferer

tolerance.

In this case, the spacing between two of the tones is

kept constant at  f , while the tones themselves are varied

across time by multiplication with a pseudorandom code.

Interestingly, when this tone signal is passed through a down-

conversion, squaring function in an ED in the WuRX, and a

bandpass ﬁlter [see Fig. 5(a)], the output appears at an IF of

f , which can be recovered with a second downconversion

to baseband. OOK or binary phase shift keying (BPSK)

modulation can be used to apply the wake-up sequence for

RX. As described in [14], this approach provides tolerance

to modulated interferers if the ED output for the interferer

(which tracks the interferer envelope) does not have signiﬁcant

spectral content at  f .

B. Proposed Multi-Tone Two-Code Wake-Up Signaling

In this work, we improve the interferer tolerance by

using three-tone signalin g that includes code-modulation of

d(t) with a code c

and a code-domain baseband ﬁl-

ter in the RX [31]. In addition , this work utilizes a

mixer-ﬁrst uncertain intermediate-frequency (UIF) architec-

ture that improves sensitivity and ach ieves high linearity

and wideband operation compared to [14]. The proposed

signaling scheme and a simpliﬁed RX chain are detailed in

Fig. 5(b).

The input signal consists of three tones; two tones are

modulated with d(t) and c

. The center tone is modulated

with only the code c

serving as an over-the-air reference for

剩余299页未读，继续阅读

netshell

粉丝: 11
资源: 185

会员权益专享

2022年9月JSSC期刊：高采样率接收器与射频前端技术

JSSC 2022.02

JSSC 2022.11

JSSC 2022.07

maven jssc

(5)在IntelliJIDEA 中使用socketTextStream监听 8888端口，获取数据。

spark streaming demo （java 1.8）

Java 3D demo

springboot如何调用串口

在IDEA中使用Spark Streaming套接字数据源实现课程实时查找

javafx 结合jSerialComm的案例

vscode怎么导入jar包

spark streaming 结合kafka 精确消费一次将结果保存到redis

jssc与jdk 版本不兼容报错

用java写spark消费kafka的消息

net.wimpi.modbus.modbus依赖

jssc jar最新版本

java程序获取温湿度传感器数据代码

android stuidio 设计一个app用来连接esp8266的热点与stm32f103c8t6通信

JAVA 监听USB扫描枪数据代码

会员权益专享

最新资源