With the rapid development of the communications industry, especially mobile communications, the low-end frequencies of the radio spectrum have become saturated. The use of various modulation methods or multiple access technologies to expand the capacity of the communication system and improve the utilization of the spectrum cannot meet the needs of future communication development. Therefore, the realization of high-speed, broadband wireless communication is bound to develop new spectrum resources in the microwave high frequency band . Due to its short wavelength and wide frequency band, millimeter wave can effectively solve many problems faced by high-speed broadband wireless access, so it has broad application prospects in short-distance wireless communication.
Various semiconductor devices are the hardware foundation of information and communication technology (ICT) . Creative research and development of emerging semiconductor technologies and circuits that meet the application of millimeter-wave wireless communication is the main technology driver for improving the capacity of communication systems and solving key problems in building a new generation of communication systems. This article follows the technological innovation and development of millimeter wave semiconductor devices, starting from the system architecture of key technologies such as phased array, semiconductor materials and processes, device design and packaging testing, and analyzes and summarizes the fifth generation ( 5G ), sixth generation (6G) ) The development trend of mobile communication technology millimeter wave system and device technology. Taking the MIDAS program of DARPA of the United States as an example, the research frontier and progress of military millimeter-wave device technology are explained.
The main problem faced by communication technology in the information age is to solve the contradiction between massive data generation and insufficient communication capacity. It is estimated that by 2032, about 45 trillion sensors will collect huge amounts of analog information from the physical world every year (equivalent to
). The existing communication network system is far from meeting the demand for information transmission capacity in the future. Solving the huge gap between the low data transmission rate and the high information generation rate of the existing wireless communication system has become a key point in the development of wireless communication technology. One of the main current solutions is to introduce a new wireless standard every few years to define a new protocol, that is, to use more complex modulation schemes to increase data throughput. However, increasing the modulation complexity to a point where it no longer significantly increases throughput, creating new solutions has become a priority.
The famous Shannon-Hartley (Shannon-Hartley) theorem points out that the capacity of a communication system is a linear function of the bandwidth. In order to transmit more data quickly, another more long-term method to improve system throughput can theoretically be adopted, that is, to extend the modulated signal (FBW) to a wider frequency range to increase its bandwidth. The current wireless communication Development is mainly to follow this line of thought. At present, the licensed operating frequency band of the cellular network is mainly below the 6GHz frequency, and the available spectrum is subject to certain restrictions. To comply with the requirements of the trend, it is inevitable to expand the operating spectrum to higher frequency bands. Millimeter-wave frequency band (generally refers to 30-300GHz electromagnetic wave frequency band) wireless communication has a wide spectrum and strong effective line-of-sight communication capabilities, which can greatly improve bandwidth, data transmission rate, and reduce end-to-end delay, achieving a significant increase in communication capacity. Therefore, it has received more and more attention. More millimeter-wave high-end frequency bands (generally referring to >6GHz frequency bands) have been used, and the millimeter-wave industry chain has also developed rapidly.
It is generally believed that the fifth-generation mobile communication technology (5G) will deploy the millimeter wave frequency band of the electromagnetic spectrum, and the sixth-generation mobile communication technology (6G) will develop and utilize the terahertz (0.1-10THz) frequency band. 5G and 6G backhaul data capacity increases from 10Gbit/s to 100Gbit/s, only by compressing data in higher modulation formats to work in mmWave mode, as shown in Figure 1, where higher operating bandwidth is available . The main technical way to achieve this goal is to carry out semiconductor technology innovation and develop semiconductor devices, materials and structures that work in the millimeter wave and terahertz frequency bands. Some key technologies that have been used in military applications for many years have become ideal technologies for 5G telecommunications. For example, phased array technology is a key technology with good development prospects. 5G telecommunications is working to realize the benefits brought by the defense industry's use of phased array antennas , overcoming the shortcomings of mmWave signals that are easily blocked by buildings or obstacles. Military applications face a more complex communication environment, usually separated by tens of kilometers or even hundreds of nautical miles. System capacity and data transmission rate are also key indicators pursued by military communication systems. The development of millimeter-wave phased arrays in the fields of military communications, radar, and electromagnetic spectrum warfare is of great significance. The virtuous circle established by military and civilian 5G communications is conducive to the formation of a situation of mutual utilization and promotion.
Figure 1. Millimeter wave technology trend of commercial wireless data service rate increasing tenfold every decade
1 Millimeter Wave Technology Trends
Judging from the current international 5G technology development, frequency bands below 6GHz in major countries have been fully commercialized. In order to achieve higher data capacity or higher bandwidth in wireless communication, the main focus is to develop and utilize the millimeter wave high frequency band above 24.25 GHz. At the 2019 World Radiocommunication Conference (WRC-19), based on the framework of international standardization organizations such as the International Telecommunication Union (ITU) and the Third Generation Partnership Project (3GPP), representatives of various countries reached a consensus on the use of 5G millimeter wave spectrum: Global The millimeter wave frequency identification with 14.75GHz bandwidth of 24.25-27.5GHz, 37-43.5GHz, and 66-71GHz is used for the future development of 5G and the International Mobile Communication System (IMT). The WRC-19 resolution plans a large number of millimeter-wave frequencies with continuous bandwidth for 5G technology, as shown in Figure 2. This has laid the foundation for the development and maturity of 5G/6G related industry chains, and the pace of global 5G system deployment and commercialization is accelerating.
The key challenge for the semiconductor industry is to develop and provide technologies that can empower 5G and 6G information transmission networks, increasing information transmission throughput, coverage space and transmission distance. These requirements will be translated into requirements for performance indicators such as semiconductor device RF and baseband bandwidth, operating frequency, power consumption, gain, noise figure, linearity, and transmit power. In addition to the 26GHz/28GHz/39GHz frequency bands that are most likely to be deployed first, in recent years, the industry has given more and more attention to semiconductor technologies that work in the V-band (57-66GHz), E-band (71-86GHz) and W-band (75-110GHz). Much attention. Frequency bands above 90GHz and up to 300GHz have also been developed. The 6G network communication frequency band will be extended upward to the terahertz frequency band and extended to three-dimensional space, which can connect satellites, aircraft, ships and land-based infrastructure to achieve truly global coverage of intelligent communication.
Figure 2. Spectrum distribution corresponding to 5G key scenarios
The new generation of millimeter-wave wireless communication system technology mainly includes massive MIMO system architecture working in the millimeter-wave frequency band, beamforming chips , base station (BS) and user terminal (UT) antennas, system measurement and calibration technology, and wireless channel characterization. The communication base station is the most critical infrastructure in the mobile communication network. Figure 3 shows a schematic diagram of the core network (CN) 5G millimeter wave base station baseband unit-active antenna unit (BBU-AAU) architecture. The base station mainly completes the conversion between the new air interface (NR) baseband signal and the radio frequency signal and the transceiver processing function of the NR radio frequency signal.
Figure 3. Schematic diagram of a 5G millimeter wave base station architecture
When transmitting a signal, the baseband signal from the 5G baseband unit is processed by the transmission link (TX) such as up-conversion, D/A conversion, radio frequency modulation, filtering, and signal amplification, and then transmitted by the switch and antenna unit. When receiving the signal, the 5G radio frequency unit receives the radio frequency signal through the antenna unit, and after processing by the receiving link (RX) such as low noise amplification, filtering and demodulation, it performs A/D conversion and down conversion, converts it into a baseband signal and sends it to the 5G baseband unit.
2 Millimeter wave beamforming system composition
According to the phase of each antenna element, 5G wireless beamforming is divided into three types by system architecture: analog beamforming, all-digital beamforming and hybrid beamforming.
2.1 Analog and digital beamforming architectures
Phased array-based analog beamforming performs phase shifting in the analog domain. The analog beamforming system is divided into three modules: digital module, bit-to-millimeter wave module and beamforming module. According to the position of the analog phase shift, the phase shift can be divided into intermediate frequency phase shift, local oscillator phase shift and radio frequency phase shift. Phase shifting is achieved using digitally controlled phase shifters (such as 6-phase shifters) or static analog beamforming structures (such as Butler matrices, Blass matrices, and Lenses). Figure 4(a) shows an architecture for an RF beamforming receiver, where the signals from the antenna elements are weighted and combined to produce a beam that is then processed by the mixer and the rest of the signal chain. This is the traditional implementation of phased arrays. The advantage is that the implementation cost is low and the deployment is simple, but the disadvantage is that it is difficult to generate multiple beams.
The principle of digital beamforming is: after the signal of each element unit is digitized independently, the phase shift is completely implemented in the digital circuit . The structure is shown in Figure 3(b), and it is fed to the antenna array through the transceiver array. The all-digital phased array is the most promising architecture. Each antenna unit is connected to an independent high-speed, high-precision A/D converter (ADC) or D/A converter (DAC). If you choose a low-resolution ADC/DAC, you can significantly reduce power consumption. All signal streams in a digital phase shifting system are digitized, so digital beamforming has the advantages of fast beam management and beamforming, can create multiple beams at the same time or search in all directions, and is robust to obstacles.
Figure 4 Schematic comparison of analog and digital beamforming architectures
2.2 Hybrid beamforming architecture
Hybrid beamforming (HBF) is a combination of analog and digital beamforming technologies, and is an intermediate solution for the combination of the two. As shown in Figure 5, it is a solution that weighs cost/hardware complexity and system performance. One option is to divide the array into smaller subarrays and perform analog beamforming within the subarrays. Each subarray can be thought of as a superelement with a certain directional radiation pattern. Digital beamforming is then performed using the signals from the subarrays, producing narrow beams of high gain corresponding to the full aperture of the array. Hybrid beamforming is currently the mainstream solution in 5G wireless communication systems.
Figure 5. Schematic diagram of hybrid beamforming with multiple analog beams
2.3 Functional modules and electronic components
Figure 6 shows the antenna module configuration for 5G and the semiconductor technology solution that each functional block should adopt. There are various functional blocks and semiconductor technology solutions for building phased arrays. The signal received by the rightmost antenna array is amplified by the front-end low-noise amplifier . Their signal bits are then aligned and combined within an RF beamformer. The synthesized signal is converted from RF frequency to IF frequency. Then it is converted into a digital signal by ADC for signal processing. On the other hand, the signal generated by the digital part is converted to analog by DAC and converted to radio frequency. They are then split into phase-adjusted signals by an RF beamformer, amplified by a front-end power amplifier and transmitted from the array antenna.
Radio frequency integrated circuits (RFICs) are an important class of components. Tokyo Institute of Technology and NEC Corporation have developed a RFIC consisting of 4 transmit/receive circuits (TRX), fabricated using 65nm bulk silicon CMOS technology. This RFIC has the function of converting the signal from IF-RF to the front end, and modifies the phase of the RF signal by changing the phase of the local oscillator signal, so that the IC can be miniaturized. Samsung has developed an RFIC, which has 16 parallel transmission channels that convert signals from IF-RF to RF front-end, and is fabricated with 28nm bulk silicon CMOS technology, as shown in Figure 7. IBM and Ericsson have jointly developed a RFIC using SiGe BiCMOS technology, which has the function of converting from intermediate frequency-radio frequency (IF-RF) to the front end as shown in Figure 6. This RFIC uses a real-time delay circuit as a phase shifter, it has 32TRX and good beamforming performance. MixComm has developed a 8TRX RFIC, using 45nmPD-SOI technology, the circuit has RF beamformer and front-end functions. The output of the power amplifier on the SOI is increased by vertical stacking to compensate the output power drop caused by the miniaturization of the gate size of the transistor . RFICs with good high-frequency characteristics for 5G millimeter waves have been developed using GaAs and GaN materials. But it can only be used to develop analog circuits such as power amplifiers and low noise amplifiers. It is currently not possible to create digital circuits with these techniques.
Figure 6 Simplified block diagram of mmWave phased array using hybrid beamforming
Figure 7.16 Signal link RFIC chip layout
3 Research progress of millimeter wave semiconductor technology
The development of systematic solutions that effectively improve information transmission and processing efficiency through semiconductor technology innovation requires collaborative research on materials, processes, system and circuit design , packaging and testing, and software.
3.1 Material technology
Semiconductor process platforms that promote the development and innovation of future communication technologies include RFSOI, FinFET, and optoelectronic technologies based on SOI/SiGe. Materials technology is at the center of innovation in semiconductor technology. Only by continuously innovating from basic materials can millimeter-wave circuits continuously increase the operating frequency and meet the requirements of different application scenarios, as shown in Figure 8. Mainstream analog IC/RFIC semiconductor materials include the following.
1) III-V compounds. At present, GaAsp HEMT and InGaP HBT are mainly used to make circuits. Wide bandgap compound materials represented by GaN are on the rise. The thermal conductivity of GaN is comparable to that of Si, but its breakdown voltage is very high, and it has higher electron mobility, higher power gain, lower noise and higher power efficiency than Si, which is very suitable for making millimeter wave Front-end circuits such as power amplifiers, low noise amplifiers, and low phase noise oscillators of the system.
Figure 8. Development roadmap of major semiconductor materials and devices
2) Si-based materials. At present, CMOS and SiGe/BiCMOS are mainly used, which are easy to achieve high integration and high cost performance, and are the preferred materials for making low-power devices. Fully depleted SOI (FDSOI) process compatible with CMOS planar bulk silicon process is a very promising technology to provide high frequency operation at low voltage. The literature proposes a mmWave beamforming system based on FDSOI devices to achieve SOC integration with high overall power efficiency. SiGe BiCMOS integrates high-performance bipolar transistors and CMOS devices in a single chip, achieving performance that can only be achieved by more expensive processes such as GaAs.
3) Multi-material heterogeneous integration. There are two different integration methods of III-V and Si co-integration technology, that is, CMOS-compatible GaN process and silicon-based III-V group wafer-level integration technology, both of which are carried out on 200 mm silicon wafers. The former utilizes the existing CMOS infrastructure and uses III-V chips and Si chips to form the final system, while the latter adopts a process compatible with the existing Si foundry process to integrate III-V devices and Si devices together. in one chip. Both are compelling research directions.
3.2 Technology
Many foundries have turned to optical lithography , which is more cost-effective than electron beam lithography, to develop new process technologies to compete in 5G chip processes; or integrate new functions into a single process node to reduce costs and gain price advantages. Figure 9 shows the evolution history of Si technology applied in 5G. In emerging 5G millimeter wave cellular applications in the 28GHz and 39GHz frequency bands, there are two compelling silicon-based technologies -- 28nm RFCMOS and 130nm/90nm SiGeBiCMOS. A number of documents have introduced the 28nm node CMOS technology and the radio frequency characteristics of the devices in this node technology in detail.
This planar technology uses gen-4nFET strained structures and immersion lithography. As far as the gate processing technology is concerned, there are Poly/SiON and high-k metal gate (HKMG) processing methods, namely the HKMG process, which can provide better Ion and gm while reducing the gate resistance Rg. The research also shows that 28nm bulk silicon CMOS technology can realize advanced SOC integrated transceiver and 25Gbit/s 60GHz broadband digital power amplifier for 802.11ac, and achieve reasonable RF front-end performance.
Figure 9 Evolution history of Si technology applied in 5G
SiGe BiCMOS is used in WiFi front-ends, long-range automotive radars, optical ICs, and 5G millimeter-wave base stations. Under BiCMOS technology, SiGe HBTs are typically added to larger size CMOS nodes, and other aspects of the technology, such as HBTs, wiring and substrate losses, are carefully optimized to maximize application benefits. For example, 350nm SiGeBiCMOS is still sufficient to meet the very challenging WiFi front-end power amplifier (PA) requirements. SiGe BiCMOS will be used as one of the important basic technologies for millimeter wave applications above 100GHz in the future.
In commercial applications, the potential of each semiconductor technology in terms of performance and integration levels must be balanced against the maturity of the process and the potential return on investment (ROI) of the chipset being developed in the context of various application market sizes. For this reason, the choice of technology should be a compromise between performance, system complexity and cost metrics, as shown in Table 1.
Table 1 Comparison of mmWave semiconductor technologies in terms of performance, integration level and development cost
3.3 IC Design Technology
With the introduction of new technology, IC design is also evolving. IC designers combine certain functions into a product by providing new functions in a single process node, or develop higher performance than before from the core transistor. These trends ultimately lead to chips that are more integrated and easier to deploy, as shown in Figure 10. Two key challenges in mmWave phased array system design are transmitter power efficiency and overall system thermal power budget. If the system is designed in Si technology, the difference between the systems will mainly be caused by the maximum power output (P sat ) of the power amplifier operating point, the power efficiency (PAE ) and the loss between the antenna and the transmitter/receiver.
Figure 10 Schematic diagram of the main circuit of the millimeter wave semiconductor system
Beamforming architecture and chip partitioning will be determined by equivalent isotropic radiated power (EIRP), frequency band and DC power consumed in the system. There is a trade-off between the area scaling of the SOC and the thermal power density dissipated. Among all PA devices, GaN stands out with the highest transmit power, the highest PAE (as shown in Figure 11), the widest bandwidth, the largest power density and the highest reliability. The main challenge is the vector magnitude degradation (EVM) caused by the difference in the filling and releasing time constants of lattice-mismatched charges in the well. In addition, it has not been proven whether these devices can operate at frequencies above 120GHz. Future research will focus on extending the operating frequency of GaN devices and PAs to 200GHz with power up to 40dBm@30% PAE. The challenge will intensify as we move to higher mmWave frequency bands.
Figure 11 Comparison of the added power efficiency of several technologies operating at mmWave frequencies
Key semiconductor device elements to focus on for mmWave device design include high-frequency technology, low-loss back-end processes that optimize on-chip RF passives, advanced modeling and simulation capabilities that account for the effects of wiring dependencies, and transmission-line-based devices and cross -coupling . It is generally believed that taking fmax or 1/3 of fT as the upper frequency limit of work can tolerate changes in process, voltage and temperature while maintaining sufficient gain. For Si CMOS devices, due to various limitations caused by gate and interconnect resistance, shrinking below 20nm has little effect on improving device performance, and fmax may reach a peak value of 450GHz at around 20nm. CMOS circuits are suitable for small-signal RF applications and have been proven to support mmWave technology in the 100GHz range. Compared with CMOS, SiGe has higher frequency, higher breakdown voltage and higher output power at high temperature. IHP has demonstrated that the fmax of the DOT750 device reaches 700GHz, and the performance of the device is much higher than that of CMOS.
3.4 Packaging and testing
Over the past few years, RF applications have driven the advanced electronic packaging market covering different industries. With the emergence of automotive radar, high-end smartphones , WiGig devices, etc., the RF packaging market is expected to grow in various segments. The RF advanced packaging market is expected to reach $34 billion by 2025. Wafer-level packaging (WLP), 3D through-silicon vias (TSV), system-in-package (SiP), and electromagnetic interference (EMI) shielding are key factors for RF devices requiring small size, high-speed operation, and heterogeneous integration. Optimizing SiP-based packaging technology at mmWave frequency bands is one of the major challenges facing communication ICs. New packaging and manufacturing techniques to manage thermal or electrical performance need to be investigated. Smart hands have strict requirements on power consumption, size and integration, while base stations have relatively less stringent requirements. Cost efficiency is also a crucial factor.
Millimeter-wave 5G requires new packaging technologies for large antennas that enable highly miniaturized wideband (>400MHz) arrays for massive MIMO. An important European project of common interest (IPCEI) is developing a wafer-level package (FOWLP) antenna package module for mmWave 5G base station applications. Figure 12 shows a schematic diagram of the package, showing a dual-die stacked FOWLP RDL (RDL is wire redistribution layer) chip post-assembly process. Two copper layers (antenna 2 and antenna 1) respectively include an integrated antenna array and its ground plane, while two packaging layers (die 2 and die 1) serve as the antenna substrate and board layer respectively. The analog front-end IC (AFE IC) made by Globalfoundries 22FDX technology is integrated into the board layer, connected to the antenna through the through hole of the stack layer, and connected to the system board through the RDL. The package size is 10mm x 10mm, and the integrated antenna array includes a 2 x 2 patch antenna. The device operates in dual frequency bands of 28GHz and 39GHz with a minimum impedance bandwidth of 400MHz for both frequency bands.
Figure 12. Dual-mode FOWLP package structure diagram
mmWave testing includes testing of systems, RF circuits, digital circuits, new materials (including some developed at the most advanced process nodes), new packaging methods, antenna arrays, SiPs, antenna-in-packages (AiPs), and mmWave-exclusive Over-the-air (OTA) testing. Millimeter wave testing has just started, with high circuit complexity and many process access points, testing, detection and measurement will take longer. The performance of MIMO is measured in real or isolated environments. Measurements of propagation channel characteristics in real indoor or outdoor environments are used to obtain the impulse response of each specific MIMO channel. It provides complete knowledge about the system under test, but only for one specific scenario. The second type of MIMO measurement is performed in an isolated environment, the OTA test. OTA testing is a key element of mmWave. The literature describes different OTA methods and proposes OTA test solutions for 28GHz and 39GHz phased array antennas and their ICs.
4 Frontiers of Military Millimeter Wave Digital Phased Array Research
4.1 Technical route
Military applications have a strong demand for mmWave technology. Key millimeter-wave technologies such as phased array antennas in communications and radar were first developed in the military field, and have now transformed into the mainstream technology of 5G communications. The two promote each other to achieve a virtuous circle. Military millimeter wave technology focuses on digital phased array cutting-edge technology research. From the application level, there are roughly the following three related directions
1) Longer-distance broadband transmission. The MIDAS program of the US Defense Advanced Research Projects Agency (DARPA) is developing a 18-50GHz highly integrated unit-level digital phased array for communication and remote sensing. The ultimate goal of MIDAS is to develop antenna apertures that enable secure communications networks between rapidly moving tactical platforms, transmitting data over greater distances with greater speed and bandwidth. Northrop Grumman and DARPA verified the wireless data transmission capability of 100Gbit/s over a distance of 20 kilometers in the "100Gbit/s radio frequency backbone network" project. The links operate at mmWave frequencies (71-76GHz and 81-86GHz) with bandwidths up to 5GHz.
2) Higher resolution and miniaturization. Public information shows that foreign fighter radars mostly work in the X-band (8GHz to 12GHz), and the radar systems for deploying and aiming missiles usually work in the Ka-band (33GHz to 37GHz). Higher resolution and smaller size antennas contribute to improved performance. 94GHz band missiles are under development.
3) Extend frequency coverage to high frequencies. Traditional EW systems operate between 2GHz and 18GHz, covering S-band, C-band, X-band and Ku-band. As the detection distance increases, so will the listening electronics. Since 5G devices operating at frequencies of 28 GHz and 39 GHz may overlap with the Ka frequency band used for missile guidance, in order to reduce channel conflicts, the electronic warfare system proposes a new extended frequency coverage—from 24 GHz to 44 GHz. The increase in bandwidth and frequency will facilitate the development of more and higher performance military electronic equipment.
4.2 Key objectives
The purpose of the DARPA MIDAS program is to develop a unit-level digital beamforming array at millimeter wave frequencies to achieve a frequency-agile multi-beam network, reduce network discovery time, and increase network throughput. The project will bring together research advances in advanced RF and mixed-signal CMOS ASIC design, compound semiconductor devices and heterogeneous integration to develop thin digital phased arrays for aerospace and defense applications. Technical field 1 (TA1) (planned development time 2018-2021): develop broadband millimeter wave digital "tiles". In addition to developing this core building block in TA1, T/R components including low-noise amplifiers (LNAs), PAs, T/R switches and radiating elements were developed, along with packaging and thermal management infrastructure for digital beam Shaped computing resources. The project completion goal is a prototype system of more than 256 units. Technical field 2 (TA2) (planned development time from 2018 to 2022): use the "tile" developed by the TA1 team to develop a broadband millimeter-wave aperture. Technical field 3 (TA3) (planned development time 2018-2021): basic research on millimeter-wave arrays, solving basic technological innovations in digital and hybrid beamforming.
4.3 Research progress
Northrop Grumman and Jariet Technology Co., Ltd. have cooperated to develop the 18-50GHz scalable digital phased array in the MIDAS project. Other participating companies include Qorvo, Micross, TowerSemiconductor and Protolabs. The division of labor of MIDAS participating companies is shown in Figure 13. NorthropGrumman plans to use bare chip 3D stacking and TSV vertical interconnection to achieve heterogeneous integration. The stacked device consists of a 3D printed radiator, two silicon feed plates that create a balanced structure for the radiator, a GaAs T/R MMIC layer, a SiGe RFIC layer, and a CMOS tile layer. Data and power enter the CMOS from the TSV in silicon, and the output signal enters the SiGe BiCMOS RFIC. The RFIC provides bias distribution, control, test and calibration RF distribution for the GaAs MMIC. The GaAs T/RMMIC layer is an 8-channel quarter-wafer interconnected by flip-chip bonding to the SiGe IC, which is the active link between the GaAs T/RMMIC layer and the CMOS tile. connector. In the second phase, the TA1 digital tile, designed by Jariet, will replace the CMOS tile and be integrated with the rest of the array. Multiple alloy properties, bonding and temperature characteristics also need to be investigated to ensure a stable assembly process.
TA1 mixed-signal ASIC was developed by Jariet, using Global Foundries' 12nm FinFET (12LP) CMOS process, which has achieved a good balance between digital and analog/RF performance, and is an extremely power-saving and compact mixed-signal and digital design plan. Efficient logic is very important to implement the digital downconverter/upconverter (DDC/DUC) module, as well as the digital beamforming function. To implement a quarter "test tile" scheme with 8 transceiver channels, a high-mid frequency range of 6 GHz is used, and the ADC sampling rate in the second Nyquist zone is 8 GS/s. The transmit path uses a DAC to maximize signal energy in the second Nyquist zone, using either fill-in (RTC) or mixed-mode waveforms. While the Phase 1 goal was to achieve a 200MHz bandwidth, Jariet's data converter design achieves a Nyquist bandwidth of 4GHz, easing the difficulty of achieving the Phase 2 goal.
To date, the participating companies have developed multiple CMOS technologies using two different state-of-the-art CMOS processes, 3D printed wide-scan and broadband radiators, InPHBT/HEMT and GaAs pHEMT low-noise amplifiers and high-power amplifiers, and low-loss T/R switches. channel transceiver ASIC. Northrop Grumman used grooved antenna arrays for the TA2 aperture study, which were 3D printed using stereolithography (SLA) and then metallized with copper. Developed advanced packaging techniques using chip stacks, copper pillars, solder bumps and distribution layers to integrate all components in a package. The completed prototype will be a scalable 256-element mmWave antenna "tile" phased array, targeting the use of large phased arrays for communications in tactical defense platforms and low-orbit satellites.
Figure 13 MIDAS division of labor
Summarize
The construction of the next generation of high-speed ubiquitous, integrated interconnection, intelligent green, safe and reliable large-capacity communication network will rely on the promotion of semiconductor technology innovation and breakthrough. Millimeter wave wireless communication is an important technology to realize the long-term sustainable development of information and communication technology, and it is currently struggling to overcome unprecedented technical challenges. These challenges stem in large part from the fact that semiconductor technology development is approaching its fundamental limits, hindering the next generation of energy efficiency on which information processing, communication, storage, sensing and actuation depend. The development of my country's millimeter wave technology should form a joint effort of the application-oriented industry, academia and research circles with the support of the government, make full use of multidisciplinary research results, and carry out various researches including basic research. Collaborative development in the fields of devices, devices and software has made breakthroughs and finally achieved the desired development goals.