Did you know that the smartphone in your hand is actually a powerful radio transceiver. It allows you to use your mobile phone to make calls, send text messages, surf the Internet, watch videos, and play games; it can also help you use drones to shoot beautiful scenery and control smart homes... These are all wonderful applications of wireless communication, bringing human life Unlimited convenience and fun.

Wireless communication is a technology that uses radio frequency signals as a medium for signal transmission. To complete wireless communication, it is first necessary to convert the signals into radio frequency signals.

"Radio Frequency Transceiver (RF Transceiver)" is such a device, which is responsible for converting baseband signals and analog signals into radio frequency signals and giving them to amplifiers and antennas for output. It is also responsible for restoring the received radio frequency signals to baseband signals and analog signals, so that these information can be turned into visible video and audible sounds. Radio frequency transceiver is the basic module in wireless communication, which is the necessary component of mobile phone, satellite communication, radar and other wireless communication equipment.

The history of radio frequency transceivers can be traced back to the end of the 19th century, when people used electromagnetic waves for wireless telegraph communication. With the advancement of science and technology, RF transceivers are constantly evolving and innovating. From the earliest transistor transceiver, to the later integrated circuit transceiver, and then to the current multi-frequency multi-mode transceiver, the performance, function and scale of RF transceiver have been greatly improved.

With the advent of 5G, mobile phone systems have become more and more complex, and higher requirements have been placed on RF transceivers. How does the RF Transceiver chip in 5G mobile phone evolve step by step, and what is the future evolution trend? Next, we will introduce the basic principle and structure of RF transceiver in detail, and discuss its continuous evolution in 2G to 5G communication.

Introduction to RF Transceiver

The term "RF transceiver" is translated from Radio Frequency Transceiver. The word Transceiver is a combination of transmitter (Transmitter) and receiver (Receiver). It can be seen from the composition of the word that the function of the Transceiver is to complete the transmission and reception of signals.

In industry applications, in order to distinguish from a transceiver device, a radio frequency transceiver is generally directly referred to as a Transceiver, and sometimes abbreviated as an XCVR. In some SoC chip manufacturers, the Transceiver chip is also called RFIC because it is a radio frequency chip.

Although the Transceiver is also responsible for the transmission and reception of signals, its function is different from the "RF Front-end" which also has the transmission and reception functions. The RF front end generally refers to a part of a path and a signal strength for processing a radio frequency signal after an antenna, and includes four basic modules: a power amplifier, a low noise amplifier, a switch, and a filter. The Transceiver is responsible for converting the analog signal and the radio frequency signal to each other: the analog signal is converted into the radio frequency signal when transmitting, and the radio frequency signal is converted into the analog signal when receiving.

The relationship between the Transceiver and the RF front end is like a pair of brothers. Brother Transceiver has a clear mind and is the head of the family. He constantly places information on the appropriate RF channel; then takes down the signal from the appropriate RF channel and converts it into useful information. The younger brother's radio frequency front end is in good physical condition and can transmit the radio frequency signal converted by the elder brother with strong strength. At the same time, very small radio frequency signals can be carefully amplified and handed over to the elder brother for processing.

Brother and brother must cooperate closely to complete the perfect signal transceiver. If you need to design and use a good RF front-end chip, you must understand the working principle of the Transceiver chip.

Why RF transmission

The main function of the Transceiver is to complete the transmission of analog signals to RF signals. Before understanding the Transceiver, the first question to be answered is: Why do you want to perform RF transmission?

Radio frequency transmission is the process of wireless transmission of information in the form of radio frequency signals by using the characteristics of electromagnetic waves propagating in the air or other media. The frequency range of the RF signal is typically 3kHz to 300GHz.

RF transmission has the following advantages:

RF transmission can overcome the physical limitations of wired transmission to achieve long-distance, barrier-free, mobile and flexible communication

RF transmission enables multi-user, multi-scenario communication, using a variety of modulation, multiplexing and coding techniques to improve communication efficiency and quality.

RF transmission can achieve more complex networking, using a variety of antenna technologies, such as directional antennas, smart antennas, phased array antennas, etc., to achieve signal transmission and reception direction control and optimization

RF transmission can achieve stronger confidentiality than wired transmission, using encryption, spread spectrum and frequency hopping technology to improve communication security.

Radio frequency transmission is the foundation and core of wireless communication technology, which brings great convenience and value to human society. In order to achieve radio frequency transmission, an important step is to convert daily images, sound, video and other signals into radio frequency signals. The process of converting analog signals to RF signals is the main function of RF transceivers.

Historically important "Transceiver" circuits

Late 19th century: embryonic appearance

Since Maxwell put forward the theory of electromagnetic waves in 1864, human beings have been imagining where this invisible and intangible magical object can be used. Around 1895, Marconi, Popov, Tesla and others realized that electromagnetic waves could be used to realize the wireless of wired communication, and designed the prototype of radio transmitter and receiver. In 1896, Italian genius radio engineer Marconi won the world's first radio patent, which also opened the door to the rapid development of radio communication.

In the RF conversion circuit designed by Marconi, the transmitter uses the Morse code key as the input to generate intermittent current pulses, and the pulse signal is connected to the high-frequency oscillator, thus the Morse code can complete the modulation of the high-frequency signal, and the modulated signal is emitted into space through the antenna. In the receiver, Marconi uses a metal powder detector. Through detection, the wireless signal can be converted into an audible sound signal and output through headphones. As a result, Marconi completed the first "Transceiver" circuit in human history. This patent has also become one of the important basis for Marconi to win the Nobel Prize in physics.

Subsequently, Marconi made improvements to this architecture, adding a tuning circuit, which can change the oscillation frequency of the circuit, making it more convenient for the transmission and reception of radio signals. Marconi used the radio system he invented to achieve communication across the English Channel and across the Atlantic.

If Marconi's invention is just to lead humans to simply understand the function of radio, in 1907, the American engineer De Forrest (De Forest) invented the vacuum triode, which makes broadcasting, telephone and communication possible on a global scale. De Forrest found that on the basis of the vacuum diode, adding a gate can control the diode current. According to this characteristic, De Forrest invented the amplifier, oscillator and other circuits to make the amplification and oscillation of wireless signals possible, thus helping the realization of radio broadcasting and remote telephone.

Heterodyne and Superheterodyne: Towards Modernity

The English name of the heterodyne transceiver is Heterodyne, which is a great invention in the history of human Transceiver.

Heterodyne is a technique of mixing signals of two frequencies to create a new frequency signal. The two input signals are mixed by a nonlinear device (such as a vacuum tube, transistor or diode), such as two signals with frequencies f1 and f2. After mixing, signals with two new frequencies f1, f2 and f1-f2 will be generated. This phenomenon is called mixing processing, and the nonlinear device used to achieve mixing is called a mixer.

Through frequency mixing, the original electromagnetic wave propagating in space can be transformed into a lower frequency range signal that can be heard by human ears, and then the information can be received through a simple detector.

In 1901, Reginald Fessenden (Reginald Fessenden) demonstrated this architecture of the transceiver, although at this time the transistor has not yet been invented, the operating frequency of the oscillator is not stable, but this architecture for modern Transceiver laid a solid foundation.

After inventing this structure, Fessenden took inspiration from the Greek words "Hetero-" (different, different) and "dyn-" (power, ability), and named this structure Heterodyne, which translates as "heterodyne" in Chinese ".

After discovering the "heterodyne" phenomenon, the engineers continued to explore. Engineers have found that using higher frequency electromagnetic wave transmission is helpful for some application scenarios, but it is extremely difficult to design amplifiers that operate at high frequencies. The engineer thought, since the original idea of heterodyne is to move the signal of sound frequency to the frequency of high-frequency electromagnetic wave through mixing, when amplifying the high-frequency signal, can it also be amplified at a relatively low RF frequency first, and then move the amplified signal to high frequency through frequency shifting? Wouldn't this save the need for high-frequency and high-linearity amplifiers?

The above design concept is the American engineer Edward Armstrong (Edwin Howard Armstrong) and others put forward the idea in 1918. In the process of frequency shifting, the fixed radio frequency preset in the middle is called "intermediate frequency" (intermediate frequency, IF,Intermediate Frequency). Because this frequency is beyond the audible range of sound, it is "ultrasonic" (supersonic), so It is named "superheterodyne" (Super-Heterodyne).

Compared with the high-frequency amplified transceiver, the superheterodyne architecture has high sensitivity, high selectivity and stability, and can adapt to the reception needs of high-frequency and weak signals in remote communication. In the past 100 years, superheterodyne structures have been widely used in wireless communication systems.

Zero intermediate frequency: the scheme is simplified and the difficulty is increased

The idea of zero intermediate frequency is no longer through the IF frequency, but directly convert the RF signal into a baseband signal in the frequency range of 0Hz. Because it is equivalent to setting the IF frequency to 0 in a superheterodyne structure, it is called a zero intermediate frequency scheme (Zero IF), also known as a direct conversion scheme (Direct Conversion), and a zero-difference scheme (Homodyne).

Zero IF scheme has its unique advantages, such:

Zero IF scheme can simplify the design, do not need to move the signal to the IF first

Zero-IF scheme for image rejection in superheterodyne schemes

The zero-IF scheme does not require circuits such as IF filters, which is convenient for single-chip integration.

The zero IF scheme no longer needs to go through an IF conversion, which looks very simple, but it will bring many problems to the actual design:

Without intermediate frequency pre-processing, the baseband output level will fluctuate widely due to the difference in received signal strength.

The local oscillator frequency is the same as the RF frequency, which may cause signal leakage interference

DC offset may occur when the mixed signal is near 0 frequency

The local oscillator needs to have precise phase locking in order to accurately move the RF frequency to near zero frequency.

Because of the above challenges, the zero IF architecture was proposed in 1924 and has not been widely promoted. In 1932, engineers used the method of comparing the local oscillator with the radio frequency to correct the frequency of the local oscillator so that the local oscillator frequency could be locked with the radio frequency. This circuit became the prototype of today's phase-locked loop (PLL).

Other problems of zero intermediate frequency were gradually solved after the invention of the world's first integrated circuit in 1958. The development of integrated circuits makes the phase-locked loop circuit to achieve more complex functions, high dynamic range, high compensation characteristics of the circuit makes the circuit can deal with a large range of space fluctuations of radio frequency signals. At the same time, the characteristics of the zero-IF scheme to facilitate single-chip integration make it complement the rapid development of integrated circuits. At present, the zero intermediate frequency scheme is widely used in mobile phones, avionics and software defined radio systems.

From 2G to 5G:Transceiver Evolving

After a hundred years of development, the RF Transceiver from the original can only transmit/receive a spark, the development to support the global frequency band, multi-function, multi-mode complex chip system. After entering the 21st century, communication protocols are still evolving, which also promotes the continuous evolution of Transceiver technology.

2G:CMOS emergence, monolithic integration

The main application of 2G cellular standard (GSM as an example) is voice communication. After 1990, 2G began to be commercialized on a large scale in the world.

The popularity of 2G mobile phones has been accompanied by the rapid development of integrated circuits. With the evolution of Moore's law, the feature size of CMOS process has been reduced to 1um before and after 1995. However, CMOS devices with a characteristic length of 0.6um can already be used to design 2.4GHz RF circuits, and 0.35um devices can even make 5GHz circuits possible [2].

The ability to build a single RF module is not enough to demonstrate the advantages of CMOS technology in RF applications, and it is the possibility of large-scale monolithic integration that attracts attention to CMOS technology. CMOS process was originally prepared for the digital process, and can also do part of the analog circuit, if even the RF can be overcome down, you can achieve complex analog-to-digital, RF hybrid circuit, while achieving single-chip integration. Because of this feature, the 2G Transceiver realized by CMOS process became a research hotspot at that time [2][3].

The CMOS implementation of fully integrated GSM Transceiver did not go smoothly. Early GSM Transceiver used BJT technology and required a large number of external devices [4]. Subsequently, some single-band GSM Transceiver designed in CMOS process were designed [5][6], and then gradually began to design multi-band fully integrated CMOS Transceiver chips. The article [7] shows the design of a fully integrated 4-band GSM Transceiver designed with 0.25um CMOS. The design adopts a receiver architecture with direct conversion and a transmitter structure with offset local oscillator, integrated PLL, VCO, mixer, intermediate frequency filter and amplifier, with a chip area of 3.2x 3.3mm.

3G:FDD transceiver at the same time, monolithic integration greater challenges

The communication standard represented by the 3G era is WCDMA,WCDMA is a FDD frequency division utilization system, the transmitter and receiver work at different frequencies at the same time, which poses a greater challenge to the design of monolithic integrated Transceiver.

In the FDD system, the receiving sensitivity of the receiver is affected by the following four conditions: the noise figure of the receiver; Tx noise in the Rx receiving band; mixing noise of Tx large signal; IM2 product of Tx signal. Three of the effects of these conditions are directly related to the isolation between the transmitter and the receiver.

In the design of 3G Transceiver, the method of enhancing LNA IIP2 and adding notch network can be used to solve the blocking problem and improve the receiving performance of the transceiver. The article [8] shows a single chip integrated WCDMA/HSDPA Transceiver with a 0.18um design. The article uses digital signal processing and tunable filters to eliminate external components, thereby achieving a highly integrated and high transceiver rejection WCDMA transceiver.

4G: Band fragmentation, CA increase

4G and smart phones appeared almost at the same time. In order to meet the demand of smart phones for high data rate cellular communication, more and more frequency bands were opened up. Operators also compete fiercely on frequency resources, resulting in multiple narrow frequency bands that are discontinuous and fragmented for each operator. In 4G phones, there may be as many as 40 bands to support.

The increase of frequency bands brings great challenges to Transceiver design. In the design, sufficient multiplexing must be considered to maintain the number of sub-modules within a reasonable range.

Another bigger challenge Transceiver 4G is the support of CA(Carrier Aggregation, carrier aggregation). CA requires multiple RF paths to work at the same time, and coupling between these simultaneous signals is inevitable. In the design, the RF lanes need to be grouped efficiently. As shown in the following figure, in the LTE receiver system supporting 3CC, Ch1 is a 2.1GHz signal, Ch2 is a 2.3GHz signal, and Ch3 is a 700MHz signal. Since Ch1 is the triple frequency of Ch3 signal, they need to be allocated to different mixer groups [9].

5G:MIMO/EN-DC, more channels

The arrival of 5G makes the rate of wireless communication increase again, and Transceiver need to realize the transceiver function of Gb/s throughput. For this reason, large-scale MIMO and CA up to 200MHz are introduced into 5G NR system. In addition, coupled with the demand for LTE NR dual connectivity (EN-DC), the design difficulty of 5G NR Transceiver has greatly increased.

In the article [10], MediaTek provide a 5G Transceiver system designed by 12nm CMOS process. The system supports up to 2 inter-band uplink CA,6 inter-band downlink CA, 4x 4 MIMO and NR 200MHz CA. In order to achieve the above functions, the Transceiver integrates 20 Rx paths, covering the frequency range of 600MHz to 6GHz. Even after the internal LNA multiplexing technology, the internal LNA still reached 28. Transceiver also use a lot of digital circuitry to achieve 200MHz bandwidth support. Up to 5 Gb/s throughput at NR 200MHz/4x 4 MIMO/256QAM.

Summary

With the evolution of the protocol, the communication capability of the terminal is required to be higher and higher. The function of the radio frequency Transceiver has changed from receiving and transmitting only one electric spark at the earliest, to transmitting several gigabytes of data per second. These requirements pose a significant challenge to Transceiver design.

The development of CMOS RF integrated circuits makes all this possible. Based on the evolving CMOS process, different circuit modules can be integrated on a single chip, and then multiple frequency bands, multiple modes, and even multiple channels can be integrated.

The complex implementation of Transceiver also makes the threshold of Transceiver design constantly higher. In the early 3G era, there were a number of third-party companies that designed RF Transceiver, and by 5G, the RF Transceiver of mobile terminals had been monopolized by head SoC platform companies, such as Qualcomm and MediaTek.

With the complexity of Transceiver functions, higher requirements are put forward for the use of Transceiver and the cooperation of other circuits in RF front-end. In the latest 5G Transceiver user manuals of Qualcomm and MediaTek, in addition to introducing the basic performance, a large amount of space is used to introduce the precautions for Transceiver use under CA and EN-DC, and the precautions for RF front-end cooperation.

In the future, RF Transceiver will face more frequency bands, higher bandwidth, lower power consumption, higher integration and other requirements. In order to meet these requirements, RF Transceiver may adopt more advanced technology, more flexible architecture, more intelligent control and other technological innovations, but also put forward higher requirements for the application and use of Transceiver.