Millimeter-Wave Digitally Intensive Frequency Generation in CMOS

Wanghua Wu,Robert Bogdan Staszewski,John R. Long

• Verlag: Academic Press
• Jahr: 2015
• ISBN:
  9780128023617
• Thema: Technik & Maschinenbau
• Thema: Signale & Signalverarbeitung

This book describes the digitally intensive time-domain architectures and techniques applied to millimeter-wave frequency synthesis, with the objective of improving performance and reducing the cost of implementation. Coverage includes system architecture, system level modeling, critical building block design, and digital calibration techniques, making it highly suitable for those who want to learn about mm-wave frequency generation for communication and radar applications, integrated circuit implementation, and time-domain circuit and system techniques.

• Highlights the challenges of frequency synthesis at mm-wave band using CMOS technology
• Compares the various approaches for mm-wave frequency generation (pros and cons)
• Introduces the digitally intensive synthesizer approach and its advantages
• Discusses the proper partitioning of the digitally intensive mm-wave frequency synthesizer into mm-wave, RF, analog, digital and software components
• Provides detailed design techniques from system level to circuit level
• Addresses system modeling, simulation techniques, design-for-test, and layout issues
• Demonstrates the use of time-domain techniques for high-performance mm-wave frequency synthesis

Chapter 1

Introduction

The demand for higher integration of circuits into a single chip and to lower production cost has driven millimeter-wave (mm-wave) electronics into CMOS. The excellent switching characteristic of MOS transistors and the high logic-gate density of sub-micron CMOS technologies motivate the digitization of mm-wave systems for improved RF performance. This chapter highlights the motivation for digitization and briefly introduces the concept of digitally intensive frequency synthesis. The design challenges of mm-wave frequency generation are also described.

Keywords

Millimeter-wave; wireless communication; FMCW radar; frequency synthesis; digitally intensive; CMOS

Contents

1.1 Motivation 2

1.1.1 Advantages of Millimeter-Wave Radios 2

1.1.2 Deep-Submicron CMOS 5

1.1.3 Digitally Intensive Approach 7

1.2 Design Challenges 9

1.2.1 Toward All-Digital PLL in mm-Wave Regime 9

1.2.2 Wide Tuning Range and Fine Frequency Resolution 11

1.2.3 Linear Wideband FM 12

References 13

Wireless communication has evolved remarkably since Guglielmo Marconi demonstrated the transmission and reception of Morse-coded messages across the Atlantic Ocean in the early twentieth century. Since then, new wireless communication methods and services have been continuously adopted that revolutionize our lives. Today, cellular, mobile, and wireless local-area networks (WLANs), afforded by breakthroughs in semiconductor technologies and their capability of mass production, are in use worldwide. They enable us to share images of our cherished moments with family and friends anywhere, and at anytime. The current trends toward portable wireless devices with ultra-high-speed (e.g., gigabit per second) connectivity will soon allow us to go online via our notebooks, cell phones, and tablets, simultaneously emailing, chatting with friends, web browsing and downloading movies and music in a fraction of the time it takes today. These devices will have to meet aggressive performance specifications in a sufficiently small and low-cost product at low power dissipation. This has prompted frantic research into new radio frequency (RF)-integrated circuits, system architectures, and design approaches.

This book explores the feasibility, advantages, design, and testing of digitally intensive frequency synthesis in the millimeter-wave (mm-wave) frequency range. An all-digital phase-locked loop (ADPLL)-based transmitter demonstrator fabricated in a production bulk CMOS process is described, which operates in the 60-GHz band, and achieves fractional frequency generation and wideband frequency modulation (FM). This digitally intensive design has the potential for low cost in volume production. It is also amenable to scaling in future technology nodes as opposed to other analog-intensive implementations. The silicon area and power consumption of such transmitters may be reduced further in future by harnessing the power of digital signal processing (DSP).

1.1 Motivation

To achieve gigabit per second (i.e., Gb/s) transfer rates, Wi-Fi technology (IEEE 802.11ac in the 5-GHz band) [1] has been developed in recent years. Multistation WLAN throughput of at least 1 Gb/s, and a single link throughput of at least 500 Mb/s is specified. It employs RF bandwidths of up to 160 MHz, multiple-input and multiple-output (MIMO) array transmitter/receiver streams (up to 8), multi-user MIMO, and up to 256-QAM (quadrature amplitude modulation) schemes in order to achieve that level of performance. The mm-wave frequency bands, by contrast, are less crowded than the low-gigahertz radio communication bands and, more attractively, have wider license-free RF bandwidth available (e.g., 7 GHz bandwidth in the 60-GHz band). This will enable the gigabit-per-second short-range communication for consumer multimedia products and support the development of emerging short-range wireless networking in many important areas, for example, commerce, manufacturing, transport, etc., and thus provide significant growth potential in new internet applications in price-sensitive communication markets.

In the following sections, the advantages and challenges of mm-wave transceiver design in CMOS technology will be examined. The focus is on mm-wave frequency synthesis.

1.1.1 Advantages of Millimeter-Wave Radios

The mm-wave frequency band is defined as 30–300 GHz with a wavelength between 1 and 10 mm in the air [2]. There are various aspects of mm-wave bands that make it attractive for short-range applications. One major advantage is the bandwidth available to carry information. To keep operating costs low, regulatory licensed bands should be avoided, thus calling for the exploitation of the unlicensed or the industrial, scientific, and medical (ISM) radio bands. Figure 1.1 plots the available bandwidth (indicated in GHz at the top of each column) for ISM and unlicensed bands below 100 GHz in the United States [3]. Below 25 GHz, the RF spectrum is congested due to frequency slots reserved for military, civil, and personal communication services. For reference, most commercial products operate in bands below 10 GHz, for example, the global system for mobile communications operates at 900 and 1,800 MHz (in Europe), and 850 and 1,900 MHz (in the United States), and ultra-wideband (UWB) radios are permitted to operate from 3.1 to 10.6 GHz [4]. Less than 1 GHz of bandwidth in total has been allocated for the license-free ISM bands at 2.45, 5.8, and 24 GHz. On the contrary, there is 7 GHz of bandwidth in the 60-GHz spectrum band allocated for license-free use, which is the largest ever allocated by the Federal Communications Commission (FCC) in the United States below 100 GHz. With such wide bandwidth available, mm-wave wireless links can achieve capacities as high as 7 Gb/s full duplex, which is unlikely to be matched by any of the RF wireless technologies at lower frequencies. The FCC has also recently approved another unlicensed band (92–95 GHz) to meet the growing demand for point-to-point high-bandwidth communication links [5].

Figure 1.1 Bandwidth allocation for the ISM and unlicensed bands below 100 GHz by the FCC (in the United States) [3].

For a given antenna size, the beamwidth can be made finer by increasing the frequency. Another benefit of the mm-wave radio is a narrower beam due to the shorter wavelength (λ=c/f_c, where c is the speed of light and f_c is the carrier frequency), which allows for deployment of multiple, independent links in close proximity. The main limitation of mm-wave radio is the physical range. Due to absorption by atmospheric oxygen and water vapor, signal strength drops off rapidly with distance compared to other bands. Figure 1.2 illustrates the general trend of increasing the attenuation of radio waves with frequency (due only to atmospheric losses; free space path loss is not accounted for) [6]. Atmospheric absorption by oxygen causes more than 15 dB/km of attenuation. The loss of a link budget at 60 GHz is therefore unacceptable for long-distance communication (e.g., >1 km), but can be used to an advantage in short-range indoor communications because the limited range and narrow beamwidths prevent interference between neighboring links. These attributes have led to greatly reduced regulatory burdens for mm-wave communications.

Figure 1.2 Average atmospheric attenuation of radio waves propagating through free space versus frequency [6].

Due to its potential for short-range, gigabit-per-second communications, several standards in the 60-GHz band have been established in recent years. The IEEE 802.15.3c standard was approved in 2009 for wireless personal-area network [7]. A similar standard for Europe (ECMA-387 [8]) was published in 2008. The WirelessHD consortium has released a specification version 1.0a for regulating the transmission of high-definition video in this unlicensed band [9]. Most recently, the IEEE 802.11ad standard (known as WiGig) [10] was adopted in 2013. It provides data rates up to 7 Gb/s, or more than 10× the maximum speed previously supported by the IEEE 802.11 standard. IEEE 802.11ad also adds a “fast session transfer” feature, which enables wireless devices to seamlessly transition between the 60-GHz frequency band and legacy bands at 2.4 and 5 GHz in order to optimize link performance and range criteria.

In addition to the gigabit-per-second communication, the 60-GHz unlicensed band also holds promise for integrating wireless sensors. Frequency-modulated continuous-wave (FMCW) radars may be utilized for presence detection and ranging at 60-GHz applications, where high-frequency resolution is required [11]. This is also the intended application for the realized ADPLL frequency synthesizer that is fully described in this book. As an example of such FMCW application is a gesture recognition system for cars, where the driver gestures (e.g., nodding the head) without taking the eyes off the road when interfacing with applications such as...