Design method and system for heterogeneous integrated circuit of microwave system

Abstract

The invention provides a design method and system for a heterogeneous integrated circuit of a microwave system, and relates to the technical field of integrated circuits, and the method comprises the steps: building a multi-layer structure model of the heterogeneous integrated circuit of the microwave system, and the multi-layer structure model comprises a substrate layer, an active device layer, a passive device layer and an interconnection layer; based on the established multilayer structure model, constructing an electromagnetic simulation model of a heterogeneous integrated circuit of the microwave system; performing full-wave electromagnetic simulation on the whole heterogeneous integrated circuit by using a finite element method to obtain an electromagnetic coupling effect and a parasitic effect between devices; based on the optimized electromagnetic simulation model, performing performance evaluation and optimization on the heterogeneous integrated circuit of the microwave system; carrying out large-scale random simulation by utilizing a Monte Carlo method, and evaluating the performance stability of the microwave system under different working conditions; and dynamically adjusting key parameters of the heterogeneous integrated circuit by adopting a self-adaptive algorithm according to a simulation result.

Description

Technical Field

[0001] The present invention relates to integrated circuit technology, and in particular to a design method and system for heterogeneous and isomorphic integrated circuits for microwave systems. Background Art

[0002] With the rapid development of wireless communications and radar systems, the performance requirements for microwave system integrated circuits are constantly increasing. Traditional single-process integrated circuits are difficult to simultaneously meet the performance requirements of high frequency, high power, high linearity, and low noise. Therefore, heterogeneous heterogeneous integration technology has become a research hotspot in the field of microwave system integrated circuit design.

[0003] At present, microwave system heterogeneous integrated circuits mainly face the following technical challenges:

[0004] Material compatibility: Different types of semiconductor materials (such as silicon, gallium arsenide, gallium nitride, etc.) have differences in lattice constants, thermal expansion coefficients, etc., which easily lead to stress and defects during heterogeneous integration, affecting device performance and reliability. Process complexity: Heterogeneous integration involves a variety of materials and processes. How to achieve high-quality preparation of different types of devices on the same substrate while ensuring good interconnection and packaging is a huge challenge.

[0005] Electromagnetic compatibility: Electromagnetic interference and parasitic effects between different types of devices will significantly affect system performance, especially in high frequency bands, where these problems are more prominent. Thermal management: The power consumption and heat dissipation characteristics of different types of devices are different. How to achieve effective thermal management in a limited space and avoid local overheating is another challenge faced by heterogeneous heterogeneous integration.

[0006] Modeling and simulation: Heterogeneous integrated circuits involve a variety of materials and devices. Traditional single-process modeling methods are difficult to accurately describe the behavior of the entire system, and new multi-physics field coupling modeling and simulation technologies need to be developed. Reliability and stability: Due to the complexity of materials and processes, heterogeneous integrated circuits may face more reliability challenges during long-term use, such as interface failure, thermal stress cracking and other problems.

[0007] In the existing technology, there are some solutions for heterogeneous integration, such as wafer bonding technology, epitaxial transfer technology, etc. However, these methods often have disadvantages such as high cost, low yield, and limited integration, which are difficult to meet the needs of large-scale commercial applications.

[0008] On the other hand, the traditional microwave system integrated circuit design method is mainly based on a single process platform, which makes it difficult to give full play to the advantages of heterogeneous integration. For example, the common design method often designs different functional modules separately and then simply combines them. This method ignores the mutual influence between heterogeneous devices and makes it difficult to achieve the optimization of overall performance.

[0009] Therefore, there is an urgent need to develop a new design method for heterogeneous integrated circuits for microwave systems, which can comprehensively consider the characteristics of multiple materials and devices, achieve overall optimization from the device to the system level, and at the same time solve key issues such as material compatibility, electromagnetic compatibility, and thermal management, improve system performance and reliability, and reduce design and manufacturing costs. Summary of the invention

[0010] The embodiments of the present invention provide a design method and system for heterogeneous integrated circuits of a microwave system, which can solve the problems in the prior art.

[0011] According to a first aspect of the embodiments of the present invention,

[0012] Provides design methods for heterogeneous integrated circuits of microwave systems, including:

[0013] A multi-layer structure model of heterogeneous integrated circuits of microwave systems is established, and the multi-layer structure model includes a substrate layer, an active device layer, a passive device layer and an interconnection layer; a plurality of different material regions are set on the substrate layer, and each material region is used to integrate different types of active devices; in the active device layer, silicon-based CMOS devices, gallium arsenide-based HEMT devices and gallium nitride-based HEMT devices are selected and arranged according to preset microwave system performance indicators; in the passive device layer, a passive device network including inductors, capacitors and transformers is designed and arranged; in the interconnection layer, a multi-layer metal interconnection structure is designed to achieve low-loss connection between active devices and passive devices;

[0014] Based on the established multi-layer structure model, an electromagnetic simulation model of heterogeneous integrated circuits in microwave systems is constructed; various active devices in the active device layer are modeled as lumped parameter models, and the passive device network in the passive device layer is modeled as a distributed parameter model; the finite element method is used to perform full-wave electromagnetic simulation of the entire heterogeneous integrated circuit to obtain the electromagnetic coupling effect and parasitic effect between each device; according to the electromagnetic simulation results, the layout of active and passive devices is optimized, and the metal routing of the interconnect layer is adjusted to reduce signal interference and parasitic effects;

[0015] Based on the optimized electromagnetic simulation model, the performance of heterogeneous integrated circuits in microwave systems is evaluated and optimized; a multivariable optimization model including temperature changes, process deviations and load changes is established; large-scale random simulations are performed using the Monte Carlo method to evaluate the performance stability of microwave systems under different working conditions; based on the simulation results, an adaptive algorithm is used to dynamically adjust the key parameters of heterogeneous integrated circuits, including the bias voltage of active devices, the impedance matching network of passive devices and the transmission line structure of the interconnection layer; through iterative optimization, the performance indicators of heterogeneous integrated circuits in microwave systems meet the preset requirements, while ensuring the stability and reliability of the system under different working conditions.

[0016] In an optional embodiment,

[0017] A multi-layer structure model of heterogeneous integrated circuits of microwave systems is established, and the multi-layer structure model includes a substrate layer, an active device layer, a passive device layer and an interconnection layer; a plurality of different material regions are set on the substrate layer, and each material region is used to integrate different types of active devices; in the active device layer, silicon-based CMOS devices, gallium arsenide-based HEMT devices and gallium nitride-based HEMT devices are selected and arranged according to the preset microwave system performance indicators; in the passive device layer, a passive device network including inductors, capacitors and transformers is designed and arranged; in the interconnection layer, a multi-layer metal interconnection structure is designed to achieve low-loss connection between active devices and passive devices, including:

[0018] Selective epitaxial growth technology is used on a silicon substrate to grow high-quality gallium arsenide film and gallium nitride film to form multiple regions of different materials; lattice matching between the gallium arsenide film and the gallium nitride film and the silicon substrate is achieved by controlling epitaxial growth parameters; buffer layers are introduced at the interface between the gallium arsenide film and the silicon substrate and at the interface between the gallium nitride film and the silicon substrate respectively; the buffer layer is an Al GaN / GaN superlattice structure;

[0019] A silicon-based CMOS device, a gallium arsenide-based HEMT device, and a gallium nitride-based HEMT device are respectively prepared on the plurality of different material regions; wherein, for the gallium nitride-based HEMT device, a field plate structure and a passivation technology are used to improve the breakdown voltage and power density, and a T-type gate structure and a gate length of less than 100 nanometers are used to improve the cutoff frequency and the maximum oscillation frequency;

[0020] A passive device layer is arranged on the silicon-based CMOS device, the gallium arsenide-based HEMT device and the gallium nitride-based HEMT device; in the passive device layer, a high-Q value inductor is prepared by combining a suspended spiral structure with a magnetic material substrate, a high-density MIM capacitor with a multi-layer stacking structure is prepared by using a high dielectric constant material, the high dielectric constant material is selected from HfO2 and Al2O3, and a transformer is prepared by using a symmetrical structure and magnetic shielding technology;

[0021] A multi-layer metal interconnection layer is arranged on the passive device layer; in the multi-layer metal interconnection layer, copper is used as the metal wire material, and the characteristic impedance is controlled by optimizing the line width and spacing; a coplanar waveguide structure or a microstrip line structure is used on the key signal path; an electromagnetic shielding structure is introduced into the multi-layer metal interconnection layer, and the electromagnetic shielding structure includes a grounded shielding wall and metal filling;

[0022] The silicon-based CMOS device, the gallium arsenide-based HEMT device, the gallium nitride-based HEMT device and the device in the passive device layer are interconnected through the multi-layer metal interconnection layer to form a complete microwave system heterogeneous integrated circuit.

[0023] In an optional embodiment,

[0024] Based on the established multi-layer structure model, an electromagnetic simulation model of heterogeneous integrated circuits of microwave systems is constructed; various active devices in the active device layer are modeled as lumped parameter models, and the passive device network in the passive device layer is modeled as a distributed parameter model, including:

[0025] Establishing a lumped parameter model of the active device layer, including modeling of silicon-based CMOS devices, gallium arsenide-based HEMT devices, and gallium nitride-based HEMT devices; wherein, for gallium nitride-based HEMT devices, a small signal equivalent circuit model including gate-source capacitance, gate-drain capacitance, drain-source capacitance, transconductance, output conductance, and parasitic inductance and resistance is established; thermal resistance and thermal capacitance are introduced into the small signal equivalent circuit model to describe the self-heating effect, and an RC network is added to simulate the trap effect;

[0026] Establish a distributed parameter model of the passive device layer, including modeling of inductors, capacitors, and transformers. For high-Q spiral inductors, establish an accurate geometric model in 3D electromagnetic simulation software, including the cross-sectional shape of the conductor, the properties of the substrate material, and the surrounding metal structure. Use the vector fitting method to convert the full-wave electromagnetic simulation results into an equivalent circuit model.

[0027] In the substrate layer model, different dielectric constants and loss tangents are defined for different material regions; additional interface impedance layers are introduced at the material interface to simulate the effects of interface stress and defects on electromagnetic properties;

[0028] Integrate the lumped parameter model of active devices and the distributed parameter model of passive devices in circuit simulation software; insert the coupling matrix or parasitic parameter network obtained by electromagnetic simulation at key nodes to consider the electromagnetic coupling and parasitic effects in heterogeneous integrated circuits;

[0029] Design and fabricate test structures, including various active and passive devices and typical interconnect structures; evaluate model accuracy by comparing simulation results with measured data, including S-parameters, noise figure and power gain;

[0030] Based on the evaluation results, the model parameters are optimized and calibrated until the deviation between the simulation results and the measured data meets the preset accuracy requirements.

[0031] In an optional embodiment,

[0032] The finite element method is used to perform full-wave electromagnetic simulation on the entire heterogeneous integrated circuit to obtain the electromagnetic coupling effects and parasitic effects between various devices, including:

[0033] Establishing a three-dimensional geometric model of a heterogeneous integrated circuit, the three-dimensional geometric model includes a silicon substrate, a selectively epitaxially grown gallium arsenide region, a selectively epitaxially grown gallium nitride region, an active device layer, a passive device layer, and a multi-layer metal interconnect structure; wherein the active device layer includes simplified geometric models of CMOS devices, gallium arsenide HEMT devices, and gallium nitride HEMT devices, and the passive device layer includes accurate geometric models of high-Q inductors, high-density MIM capacitors, and transformers;

[0034] Specifying electromagnetic parameters for each material in the three-dimensional geometric model, including dielectric constant, conductivity and magnetic permeability; establishing a nonlinear characteristic model for nonlinear materials; setting radiation boundary conditions or absorption boundary conditions at the boundaries of the three-dimensional geometric model; and setting impedance boundary conditions for metal structures;

[0035] Meshing the three-dimensional geometric model, using adaptive meshing technology, using fine meshes at device interfaces and high-frequency current intensive areas, and using relatively coarse meshes in non-critical areas; performing mesh convergence analysis to ensure that the mesh quality meets the preset accuracy requirements;

[0036] Set the port excitation and select Gaussian pulse as the excitation source; set the solution frequency points to cover the entire frequency band of the device operation, and set more dense frequency points near the resonant frequency of the high Q value structure;

[0037] Use the finite element method to discretize continuous electromagnetic field problems into a system of linear equations; for time domain analysis, use explicit or implicit time integration methods to solve; for frequency domain analysis, directly solve the complex linear equations; use parallel computing technology and iterative solvers to improve computational efficiency;

[0038] Extract S parameters, impedance matrix and electromagnetic field distribution from full-wave electromagnetic simulation results; analyze the mutual inductance between adjacent spiral inductors and the parasitic capacitance between HEMT devices and nearby wires; identify hot spots and electromagnetic interference problems through current density distribution;

[0039] A parasitic effect network model is constructed based on the simulation results, and the extracted parasitic parameters are converted into an equivalent circuit model, including coupling capacitance, mutual inductance and parasitic resistance; for complex coupling structures, multi-port network parameter description is used.

[0040] In an optional embodiment,

[0041] According to the simulation results, an adaptive algorithm is used to dynamically adjust the key parameters of heterogeneous integrated circuits, including the bias voltage of active devices, the impedance matching network of passive devices, and the transmission line structure of the interconnect layer.

[0042] Establishing a full-wave electromagnetic model of a heterogeneous integrated circuit, wherein the full-wave electromagnetic model includes accurate geometric structures and material parameters of active devices, passive devices and interconnect layers;

[0043] defining an outer optimization objective function for evaluating active device performance, wherein the objective function is a weighted sum of output power, power added efficiency, and third-order intermodulation distortion;

[0044] The bias voltage of the active device is optimized by using the gradient descent algorithm, which calculates the partial derivatives of the objective function with respect to the gate-source voltage and the drain-source voltage and iteratively updates the gate-source voltage and the drain-source voltage in combination with the learning rate.

[0045] A middle-layer optimization fitness function is defined to evaluate the performance of the passive device impedance matching network, wherein the fitness function is a function of the scattering parameter;

[0046] The impedance matching network of passive devices is optimized by using genetic algorithm. The parameters of matching network components are encoded as chromosomes. A new generation of population is generated through selection, crossover and mutation operations. The performance of each individual is evaluated using fitness function.

[0047] An inner layer optimization objective function is defined to evaluate the performance of the interconnect layer transmission line structure, wherein the objective function includes a weighted square sum of a difference between a characteristic impedance and a target impedance, a difference between a transmission loss and a maximum allowable loss, and a difference between a phase velocity and a theoretical phase velocity;

[0048] The particle swarm optimization algorithm is used to optimize the transmission line structure of the interconnect layer. The transmission line geometric parameters are encoded as particle positions. The transmission line structure is iteratively optimized by calculating the particle speed and position update formula.

[0049] Construct a multi-layer nested optimization structure, where the outer loop controls the optimization of active devices, the middle loop is responsible for matching network optimization, and the inner loop handles the optimization of transmission line structure; each time the inner optimization is completed, the optimization result is fed back to the middle loop; each time the middle optimization is completed, the optimization result is fed back to the outer loop;

[0050] In each layer of optimization process, full-wave electromagnetic simulation is used to obtain performance indicators, and the optimization parameters are updated according to the simulation results; the iteration continues until the preset termination condition is met or the maximum number of iterations is reached;

[0051] The final optimization results are output, including the optimal bias voltage of active devices, the optimal impedance matching network parameters of passive devices, and the optimal transmission line structure parameters of the interconnect layer.

[0052] According to a second aspect of the embodiments of the present invention,

[0053] Provides a design system for heterogeneous integrated circuits for microwave systems, including:

[0054] The first unit is used to establish a multi-layer structure model of heterogeneous integrated circuits of microwave systems, which includes a substrate layer, an active device layer, a passive device layer and an interconnection layer; a plurality of different material regions are set on the substrate layer, and each material region is used to integrate different types of active devices; in the active device layer, silicon-based CMOS devices, gallium arsenide-based HEMT devices and gallium nitride-based HEMT devices are selected and arranged according to preset microwave system performance indicators; in the passive device layer, a passive device network including inductors, capacitors and transformers is designed and arranged; in the interconnection layer, a multi-layer metal interconnection structure is designed to achieve low-loss connection between active devices and passive devices;

[0055] The second unit is used to build an electromagnetic simulation model of a heterogeneous integrated circuit of a microwave system based on the established multi-layer structure model; model various active devices in the active device layer as lumped parameter models, and model the passive device network in the passive device layer as a distributed parameter model; use the finite element method to perform full-wave electromagnetic simulation on the entire heterogeneous integrated circuit to obtain the electromagnetic coupling effect and parasitic effect between each device; optimize the layout of active and passive devices based on the electromagnetic simulation results, and adjust the metal routing of the interconnect layer to reduce signal interference and parasitic effects;

[0056] The third unit is used to evaluate and optimize the performance of heterogeneous integrated circuits in microwave systems based on the optimized electromagnetic simulation model; establish a multivariable optimization model that includes temperature changes, process deviations and load changes; use the Monte Carlo method to perform large-scale random simulations to evaluate the performance stability of microwave systems under different working conditions; based on the simulation results, use adaptive algorithms to dynamically adjust the key parameters of heterogeneous integrated circuits, including adjusting the bias voltage of active devices, the impedance matching network of passive devices and the transmission line structure of the interconnection layer; through iterative optimization, the performance indicators of heterogeneous integrated circuits in microwave systems meet the preset requirements, while ensuring the stability and reliability of the system under different working conditions.

[0057] According to a third aspect of the embodiments of the present invention,

[0058] An electronic device is provided, comprising:

[0059] processor;

[0060] a memory for storing processor-executable instructions;

[0061] The processor is configured to call the instructions stored in the memory to execute the aforementioned method.

[0062] A fourth aspect of the embodiments of the present invention is:

[0063] A computer-readable storage medium is provided, on which computer program instructions are stored. When the computer program instructions are executed by a processor, the aforementioned method is implemented.

[0064] The establishment of the above multi-layer structure model laid the foundation for subsequent electromagnetic simulation and performance optimization. The model not only takes into account the characteristics and layout of various devices, but also fully considers key issues such as material compatibility, electromagnetic compatibility and thermal management, providing a feasible technical path for realizing high-performance microwave system heterogeneous integrated circuits. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 A schematic diagram of a flow chart of a method for designing a heterogeneous integrated circuit for a microwave system according to an embodiment of the present invention;

[0066] Figure 2 It is a schematic diagram of the structure of a design system for heterogeneous integrated circuits of a microwave system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0067] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0068] The technical solution of the present invention is described in detail with specific embodiments below. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments.

[0069] Figure 1 FIG. 1 is a flow chart of a method for designing a heterogeneous integrated circuit for a microwave system according to an embodiment of the present invention. Figure 1 As shown, the method includes:

[0070] S101. Establish a multi-layer structure model of a microwave system heterogeneous integrated circuit, the multi-layer structure model includes a substrate layer, an active device layer, a passive device layer and an interconnection layer; set a plurality of different material regions on the substrate layer, each material region is used to integrate different types of active devices; in the active device layer, select and arrange silicon-based CMOS devices, gallium arsenide-based HEMT devices and gallium nitride-based HEMT devices according to preset microwave system performance indicators; in the passive device layer, design and arrange a passive device network including inductors, capacitors and transformers; in the interconnection layer, design a multi-layer metal interconnection structure to achieve low-loss connection between active devices and passive devices;

[0071] S102. Based on the established multi-layer structure model, an electromagnetic simulation model of the heterogeneous integrated circuit of the microwave system is constructed; various active devices in the active device layer are modeled as lumped parameter models, and the passive device network in the passive device layer is modeled as a distributed parameter model; the finite element method is used to perform full-wave electromagnetic simulation on the entire heterogeneous integrated circuit to obtain electromagnetic coupling effects and parasitic effects between various devices; according to the electromagnetic simulation results, the layout of active devices and passive devices is optimized, and the metal routing of the interconnection layer is adjusted to reduce signal interference and parasitic effects;

[0072] S103. Based on the optimized electromagnetic simulation model, the performance of heterogeneous integrated circuits of microwave systems is evaluated and optimized; a multivariable optimization model including temperature changes, process deviations and load changes is established; large-scale random simulation is performed using the Monte Carlo method to evaluate the performance stability of microwave systems under different working conditions; based on the simulation results, an adaptive algorithm is used to dynamically adjust the key parameters of heterogeneous integrated circuits, including the bias voltage of active devices, the impedance matching network of passive devices and the transmission line structure of the interconnection layer; through iterative optimization, the performance indicators of heterogeneous integrated circuits of microwave systems meet the preset requirements, while ensuring the stability and reliability of the system under different working conditions.

[0073] In an optional embodiment,

[0074] A multi-layer structure model of heterogeneous integrated circuits of microwave systems is established, and the multi-layer structure model includes a substrate layer, an active device layer, a passive device layer and an interconnection layer; a plurality of different material regions are set on the substrate layer, and each material region is used to integrate different types of active devices; in the active device layer, silicon-based CMOS devices, gallium arsenide-based HEMT devices and gallium nitride-based HEMT devices are selected and arranged according to the preset microwave system performance indicators; in the passive device layer, a passive device network including inductors, capacitors and transformers is designed and arranged; in the interconnection layer, a multi-layer metal interconnection structure is designed to achieve low-loss connection between active devices and passive devices, including:

[0075] Selective epitaxial growth technology is used on a silicon substrate to grow high-quality gallium arsenide film and gallium nitride film to form multiple regions of different materials; lattice matching between the gallium arsenide film and the gallium nitride film and the silicon substrate is achieved by controlling epitaxial growth parameters; buffer layers are introduced at the interface between the gallium arsenide film and the silicon substrate and at the interface between the gallium nitride film and the silicon substrate respectively; the buffer layer is an Al GaN / GaN superlattice structure;

[0076] A silicon-based CMOS device, a gallium arsenide-based HEMT device, and a gallium nitride-based HEMT device are respectively prepared on the plurality of different material regions; wherein, for the gallium nitride-based HEMT device, a field plate structure and a passivation technology are used to improve the breakdown voltage and power density, and a T-type gate structure and a gate length of less than 100 nanometers are used to improve the cutoff frequency and the maximum oscillation frequency;

[0077] A passive device layer is arranged on the silicon-based CMOS device, the gallium arsenide-based HEMT device and the gallium nitride-based HEMT device; in the passive device layer, a high-Q value inductor is prepared by combining a suspended spiral structure with a magnetic material substrate, a high-density MIM capacitor with a multi-layer stacking structure is prepared by using a high dielectric constant material, the high dielectric constant material is selected from HfO2 and Al2O3, and a transformer is prepared by using a symmetrical structure and magnetic shielding technology;

[0078] A multi-layer metal interconnection layer is arranged on the passive device layer; in the multi-layer metal interconnection layer, copper is used as the metal wire material, and the characteristic impedance is controlled by optimizing the line width and spacing; a coplanar waveguide structure or a microstrip line structure is used on the key signal path; an electromagnetic shielding structure is introduced into the multi-layer metal interconnection layer, and the electromagnetic shielding structure includes a grounded shielding wall and metal filling;

[0079] The silicon-based CMOS device, the gallium arsenide-based HEMT device, the gallium nitride-based HEMT device and the device in the passive device layer are interconnected through the multi-layer metal interconnection layer to form a complete microwave system heterogeneous integrated circuit.

[0080] For example, the multi-layer structure model of heterogeneous integrated circuits in microwave systems is the key to achieving high-performance, multi-functional integration. The model consists of substrate layer, active device layer, passive device layer and interconnect layer, each of which has specific functions and design requirements.

[0081] The substrate layer is the foundation of the entire heterogeneous integrated circuit, and it is necessary to achieve the coexistence of multiple materials on a single substrate. Selective epitaxial growth technology is used to grow high-quality gallium arsenide and gallium nitride films on silicon substrates to form multiple regions of different materials. By precisely controlling epitaxial growth parameters such as temperature, pressure, and gas flow, good lattice matching between gallium arsenide and gallium nitride films and silicon substrates can be achieved, reducing interface defects and stress. At the same time, a buffer layer is introduced at the material junction, such as an Al GaN / GaN superlattice structure, to further improve the quality of the heterogeneous interface.

[0082] In the active device layer, different types of active devices are selected and arranged according to the preset microwave system performance indicators. Silicon-based CMOS devices are mainly used for digital signal processing and low-frequency analog circuits, gallium arsenide-based HEMT devices are suitable for intermediate frequency amplification and mixing circuits, and gallium nitride-based HEMT devices are used for high-power, high-frequency RF amplification. In order to give full play to the advantages of various devices, it is necessary to optimize the device structure and process parameters. For example, for gallium nitride-based HEMT, the breakdown voltage and power density of the device can be significantly improved by introducing a field plate structure and adopting passivation technology. At the same time, the use of a T-type gate structure and a sub-100nm gate length can achieve a higher cutoff frequency and maximum oscillation frequency.

[0083] The design of the passive device layer directly affects the performance and integration of the system. High-Q inductors are key components for realizing high-performance RF circuits. They can be made by combining a suspended spiral structure with a magnetic material substrate to reduce the size of the inductor while ensuring a high Q value. For high-density MIM capacitors, high dielectric constant materials (such as HfO2 and Al2O3) can be used as dielectrics, and the capacitance per unit area can be increased through a multi-layer stacking structure. The design of the transformer needs to consider the coupling coefficient and insertion loss, and a symmetrical structure and magnetic shielding technology can be used to optimize performance.

[0084] The interconnect layer is the key to connecting active and passive devices, and its design directly affects signal integrity and system performance. The use of multi-layer metal interconnect structures can achieve high-density, low-loss signal transmission. In order to reduce the transmission loss of high-frequency signals, low-loss materials (such as copper) can be used as metal wires, and the characteristic impedance can be controlled by optimizing the line width and spacing. On the key signal path, coplanar waveguide or microstrip line structures can be used to achieve impedance matching and low-loss transmission by precisely controlling the transmission line geometry parameters. At the same time, in order to suppress parasitic effects and electromagnetic interference, electromagnetic shielding structures such as grounded shielding walls and metal fillings are introduced in the interconnect layer.

[0085] The establishment of the above multi-layer structure model laid the foundation for subsequent electromagnetic simulation and performance optimization. The model not only takes into account the characteristics and layout of various devices, but also fully considers key issues such as material compatibility, electromagnetic compatibility and thermal management, providing a feasible technical path for realizing high-performance microwave system heterogeneous integrated circuits.

[0086] In an optional embodiment,

[0087] Based on the established multi-layer structure model, an electromagnetic simulation model of heterogeneous integrated circuits of microwave systems is constructed; various active devices in the active device layer are modeled as lumped parameter models, and the passive device network in the passive device layer is modeled as a distributed parameter model, including:

[0088] Establishing a lumped parameter model of the active device layer, including modeling of silicon-based CMOS devices, gallium arsenide-based HEMT devices, and gallium nitride-based HEMT devices; wherein, for gallium nitride-based HEMT devices, a small signal equivalent circuit model including gate-source capacitance, gate-drain capacitance, drain-source capacitance, transconductance, output conductance, and parasitic inductance and resistance is established; thermal resistance and thermal capacitance are introduced into the small signal equivalent circuit model to describe the self-heating effect, and an RC network is added to simulate the trap effect;

[0089] Establish a distributed parameter model of the passive device layer, including modeling of inductors, capacitors, and transformers. For high-Q spiral inductors, establish an accurate geometric model in 3D electromagnetic simulation software, including the cross-sectional shape of the conductor, the properties of the substrate material, and the surrounding metal structure. Use the vector fitting method to convert the full-wave electromagnetic simulation results into an equivalent circuit model.

[0090] In the substrate layer model, different dielectric constants and loss tangents are defined for different material regions; additional interface impedance layers are introduced at the material interface to simulate the effects of interface stress and defects on electromagnetic properties;

[0091] Integrate the lumped parameter model of active devices and the distributed parameter model of passive devices in circuit simulation software; insert the coupling matrix or parasitic parameter network obtained by electromagnetic simulation at key nodes to consider the electromagnetic coupling and parasitic effects in heterogeneous integrated circuits;

[0092] Design and fabricate test structures, including various active and passive devices and typical interconnect structures; evaluate model accuracy by comparing simulation results with measured data, including S-parameters, noise figure and power gain;

[0093] Based on the evaluation results, the model parameters are optimized and calibrated until the deviation between the simulation results and the measured data meets the preset accuracy requirements.

[0094] For example, building an accurate electromagnetic simulation model is a key step in realizing the design of heterogeneous integrated circuits for high-performance microwave systems. Based on the multi-layer structure model established above, a hybrid modeling strategy is adopted to model various active devices in the active device layer as lumped parameter models, and the passive device network in the passive device layer as distributed parameter models. This hybrid modeling method can greatly improve the simulation efficiency while ensuring the simulation accuracy.

[0095] First, for silicon-based CMOS devices, gallium arsenide-based HEMT devices, and gallium nitride-based HEMT devices in the active device layer, lumped parameter models are used for modeling. Taking gallium nitride-based HEMT as an example, its small signal equivalent circuit model includes parameters such as gate-source capacitance Cgs, gate-drain capacitance Cgd, drain-source capacitance Cds, transconductance gm, output conductance gds, and parasitic inductance and resistance. These parameters can be extracted through the S parameter measurement results of the actual device combined with the optimization algorithm. In order to improve the accuracy of the model, nonlinear effects such as self-heating effect and trap effect also need to be considered. The self-heating effect can be described by introducing thermal resistance Rth and thermal capacitance Cth in the equivalent circuit, while the trap effect can be simulated by adding an RC network.

[0096] For devices such as inductors, capacitors and transformers in the passive device layer, distributed parameter models are used for modeling. Taking high-Q spiral inductors as an example, its electromagnetic simulation model needs to consider the resistance and inductance of metal wires as well as the mutual inductance and parasitic capacitance between coils. By establishing an accurate geometric model in a three-dimensional electromagnetic simulation software (such as HFSS or CST), including the cross-sectional shape of the wire, the properties of the substrate material and the surrounding metal structure, a high-precision S parameter model can be obtained. In order to improve the simulation efficiency, model order reduction techniques such as vector fitting can be used to convert the full-wave electromagnetic simulation results into an equivalent circuit model.

[0097] When building a complete electromagnetic simulation model, special attention should be paid to the effects of heterogeneous interfaces and multilayer structures. For different material regions in the substrate layer, their electromagnetic properties can be described by defining different dielectric constants and loss tangents. At the material interface, the effects of interface stress and defects on electromagnetic properties need to be considered, and these effects can be simulated by introducing additional interface impedance layers.

[0098] For the modeling of interconnect layers, it is more appropriate to use a distributed parameter model. For coplanar waveguide or microstrip line structures on key signal paths, a quasi-TEM transmission line model can be used to describe them. The characteristic impedance and effective dielectric constant of the transmission line model can be calculated using analytical formulas or numerical methods.

[0099] After completing the modeling of each part, it is necessary to integrate the lumped parameter model of the active device and the distributed parameter model of the passive device. This can be achieved through circuit-electromagnetic joint simulation technology. Specifically, the S parameter model or equivalent circuit model of the passive device can be imported into the circuit simulation software (such as ADS) and connected with the lumped parameter model of the active device. At the same time, considering the electromagnetic coupling and parasitic effects in heterogeneous integrated circuits, it is necessary to insert the coupling matrix or parasitic parameter network obtained by electromagnetic simulation at the key nodes.

[0100] Finally, in order to verify the accuracy of the constructed electromagnetic simulation model, experimental verification is required. Test structures can be designed and produced, including various active devices, passive devices, and typical interconnect structures. The accuracy of the model is evaluated by comparing the simulation results with the measured data, such as S parameters, noise figure, power gain and other performance indicators. If there is a significant deviation, the model parameters need to be further optimized and calibrated.

[0101] The electromagnetic simulation model constructed through the above steps not only takes into account the material and structural complexity brought by heterogeneous integration, but also takes into account simulation accuracy and efficiency. This provides a reliable foundation for subsequent circuit design and performance optimization, and helps to achieve the design goal of heterogeneous integrated circuits for high-performance microwave systems.

[0102] In an optional embodiment,

[0103] The finite element method is used to perform full-wave electromagnetic simulation on the entire heterogeneous integrated circuit to obtain the electromagnetic coupling effects and parasitic effects between various devices, including:

[0104] Establishing a three-dimensional geometric model of a heterogeneous integrated circuit, the three-dimensional geometric model includes a silicon substrate, a selectively epitaxially grown gallium arsenide region, a selectively epitaxially grown gallium nitride region, an active device layer, a passive device layer, and a multi-layer metal interconnect structure; wherein the active device layer includes simplified geometric models of CMOS devices, gallium arsenide HEMT devices, and gallium nitride HEMT devices, and the passive device layer includes accurate geometric models of high-Q inductors, high-density MIM capacitors, and transformers;

[0105] Specifying electromagnetic parameters for each material in the three-dimensional geometric model, including dielectric constant, conductivity and magnetic permeability; establishing a nonlinear characteristic model for nonlinear materials; setting radiation boundary conditions or absorption boundary conditions at the boundaries of the three-dimensional geometric model; and setting impedance boundary conditions for metal structures;

[0106] Meshing the three-dimensional geometric model, using adaptive meshing technology, using fine meshes at device interfaces and high-frequency current intensive areas, and using relatively coarse meshes in non-critical areas; performing mesh convergence analysis to ensure that the mesh quality meets the preset accuracy requirements;

[0107] Set the port excitation and select Gaussian pulse as the excitation source; set the solution frequency points to cover the entire frequency band of the device operation, and set more dense frequency points near the resonant frequency of the high Q value structure;

[0108] Use the finite element method to discretize continuous electromagnetic field problems into a system of linear equations; for time domain analysis, use explicit or implicit time integration methods to solve; for frequency domain analysis, directly solve the complex linear equations; use parallel computing technology and iterative solvers to improve computational efficiency;

[0109] Extract S parameters, impedance matrix and electromagnetic field distribution from full-wave electromagnetic simulation results; analyze the mutual inductance between adjacent spiral inductors and the parasitic capacitance between HEMT devices and nearby wires; identify hot spots and electromagnetic interference problems through current density distribution;

[0110] A parasitic effect network model is constructed based on the simulation results, and the extracted parasitic parameters are converted into an equivalent circuit model, including coupling capacitance, mutual inductance and parasitic resistance; for complex coupling structures, multi-port network parameter description is used.

[0111] For example, in heterogeneous integrated circuits, due to the close integration of different materials and devices, the electromagnetic coupling effects and parasitic effects between devices become particularly complex and important. In order to accurately capture these effects, using the finite element method for full-wave electromagnetic simulation is an effective method. The specific steps are as follows:

[0112] First, a 3D geometric model of the entire heterogeneous integrated circuit is established. This includes the silicon substrate, the selectively epitaxially grown gallium arsenide and gallium nitride regions, the active device layer (CMOS, GaAs HEMT, GaN HEMT), the passive device layer (high Q value inductor, high density MIM capacitor, transformer) and the multi-layer metal interconnect structure. During the modeling process, the geometric dimensions and material properties of each structure need to be accurately described. For active devices, a simplified geometric model can be used, but key physical structures such as gate, source and drain need to be retained.

[0113] Next, define the material properties and boundary conditions. Specify the appropriate electromagnetic parameters for each material, including the dielectric constant, conductivity, and permeability. For nonlinear materials, such as ferrites, their nonlinear properties need to be considered. Set appropriate boundary conditions at the model boundaries, such as radiation boundary conditions or absorption boundary conditions to simulate open space. For metal structures, impedance boundary conditions can be used to reduce the complexity of meshing.

[0114] Then, meshing is performed. Due to the presence of multi-scale structures in heterogeneous integrated circuits, from nanometer-scale transistor channels to millimeter-scale spiral inductors, adaptive meshing technology is required. In critical areas, such as device interfaces and high-frequency current-intensive areas, finer meshes are used. At the same time, in order to improve computational efficiency, relatively coarse meshes can be used in non-critical areas. The quality of the mesh directly affects the accuracy of the simulation results, so mesh convergence analysis is required to ensure the reliability of the results.

[0115] Set the excitation source and solution parameters. Set appropriate port excitation according to the frequency range to be analyzed. For broadband analysis, you can use Gaussian pulses as the excitation source. Set the solution frequency points, usually covering the entire frequency band of the device. For high Q value structures, such as spiral inductors, it is necessary to set more dense frequency points near the resonant frequency.

[0116] Solve Maxwell equations. The continuous electromagnetic field problem is discretized into a linear system of equations using the finite element method. For time domain analysis, explicit or implicit time integration methods are used; for frequency domain analysis, the complex linear system of equations is solved directly. Given the scale of the problem, parallel computing techniques and iterative solvers are usually required to improve computational efficiency.

[0117] Post-processing and data extraction. Extract various electromagnetic parameters from full-wave electromagnetic simulation results, including S parameters, impedance matrix, electromagnetic field distribution, etc. Pay special attention to the mutual coupling effect between devices, such as the mutual inductance between adjacent spiral inductors, the parasitic capacitance between HEMT devices and nearby wires, etc. By analyzing the current density distribution, potential hot spots and EMI problems can be identified.

[0118] Based on the simulation results, a parasitic effect network model is constructed. The extracted parasitic parameters are converted into equivalent circuit models, such as coupling capacitance, mutual inductance, and parasitic resistance. These models can be directly integrated into circuit-level simulation to improve the accuracy of the overall design. For complex coupling structures, multi-port network parameters can be used to describe them, such as Y parameters or Z parameters.

[0119] Finally, perform experimental verification and model calibration. Design a dedicated test structure to verify the accuracy of the simulation results. Evaluate the accuracy of the model by comparing the simulated and measured S parameters, group delay, and other parameters. If there is a significant deviation, it is necessary to repeatedly adjust the model parameters and simulation settings until the desired accuracy is achieved.

[0120] Through the above steps, the electromagnetic coupling effects and parasitic effects in heterogeneous integrated circuits can be fully captured. These effects have a significant impact on the high-frequency performance, power consumption and reliability of the circuit. Accurate full-wave electromagnetic simulation results provide an important basis for subsequent circuit optimization and layout design, which helps to achieve the design goals of high-performance microwave systems.

[0121] In an optional embodiment,

[0122] According to the simulation results, an adaptive algorithm is used to dynamically adjust the key parameters of heterogeneous integrated circuits, including the bias voltage of active devices, the impedance matching network of passive devices, and the transmission line structure of the interconnect layer.

[0123] Establishing a full-wave electromagnetic model of a heterogeneous integrated circuit, wherein the full-wave electromagnetic model includes accurate geometric structures and material parameters of active devices, passive devices and interconnect layers;

[0124] defining an outer optimization objective function for evaluating active device performance, wherein the objective function is a weighted sum of output power, power added efficiency, and third-order intermodulation distortion;

[0125] The bias voltage of the active device is optimized by using the gradient descent algorithm, which calculates the partial derivatives of the objective function with respect to the gate-source voltage and the drain-source voltage and iteratively updates the gate-source voltage and the drain-source voltage in combination with the learning rate.

[0126] A middle-layer optimization fitness function is defined to evaluate the performance of the passive device impedance matching network, wherein the fitness function is a function of the scattering parameter;

[0127] The impedance matching network of passive devices is optimized by using genetic algorithm. The parameters of matching network components are encoded as chromosomes. A new generation of population is generated through selection, crossover and mutation operations. The performance of each individual is evaluated using fitness function.

[0128] An inner layer optimization objective function is defined to evaluate the performance of the interconnect layer transmission line structure, wherein the objective function includes a weighted square sum of a difference between a characteristic impedance and a target impedance, a difference between a transmission loss and a maximum allowable loss, and a difference between a phase velocity and a theoretical phase velocity;

[0129] The particle swarm optimization algorithm is used to optimize the transmission line structure of the interconnect layer. The transmission line geometric parameters are encoded as particle positions. The transmission line structure is iteratively optimized by calculating the particle speed and position update formula.

[0130] Construct a multi-layer nested optimization structure, where the outer loop controls the optimization of active devices, the middle loop is responsible for matching network optimization, and the inner loop handles the optimization of transmission line structure; each time the inner optimization is completed, the optimization result is fed back to the middle loop; each time the middle optimization is completed, the optimization result is fed back to the outer loop;

[0131] In each layer of optimization process, full-wave electromagnetic simulation is used to obtain performance indicators, and the optimization parameters are updated according to the simulation results; the iteration continues until the preset termination condition is met or the maximum number of iterations is reached;

[0132] The final optimization results are output, including the optimal bias voltage of active devices, the optimal impedance matching network parameters of passive devices, and the optimal transmission line structure parameters of the interconnect layer.

[0133] For example, the performance optimization of heterogeneous integrated circuits is a complex multivariable problem involving the coordinated design of active devices, passive devices, and interconnect structures. Based on the above full-wave electromagnetic simulation results, an adaptive algorithm can be used to dynamically adjust key parameters to achieve global optimization of circuit performance.

[0134] First, an adaptive algorithm based on gradient descent is used to adjust the bias voltage of active devices. Taking the GaN HEMT power amplifier as an example, its key performance indicators include output power, efficiency and linearity. After each update, the circuit simulation is re-performed to calculate the new objective function value. The iterative process continues until the objective function converges or the preset number of iterations is reached.

[0135] For the optimization of impedance matching networks of passive devices, genetic algorithms combined with electromagnetic simulation are used. Taking the input matching network as an example, the chromosome encoding is first defined, and each gene represents the parameters of a matching network component (such as inductance or capacitance). The initial population is randomly generated, and each individual corresponds to a set of matching network parameters. Selection, crossover and mutation operations are performed to generate a new generation of populations. Electromagnetic simulation is performed on each new individual to calculate the fitness. This process is repeated until the termination condition is met, such as reaching the maximum number of generations or fitness convergence. The optimal individual finally obtained is the optimized matching network parameters.

[0136] The particle swarm optimization algorithm is used to optimize the transmission line structure of the interconnect layer. Taking the coplanar waveguide structure as an example, the optimization variables include the signal line width W, the distance S between the signal line and the ground plane, and the dielectric layer thickness H.

[0137] Initialize the particle swarm, each particle represents a set of (W, S, H) parameters. Perform electromagnetic simulation on each particle and calculate the objective function value. Update the individual optimal position and global optimal position of each particle.

[0138] The above three optimization processes are interrelated. Adjustment of the bias voltage of the active device will affect its input and output impedance, which in turn requires re-optimization of the matching network. Changes in the matching network may cause adjustments to the transmission line structure. Therefore, it is necessary to establish a collaborative optimization framework at a higher level. A nested loop structure is adopted, in which the outer loop controls the optimization of active devices, the middle loop is responsible for matching network optimization, and the inner loop handles the optimization of transmission line structure. After each inner layer optimization is completed, the result is fed back to the middle layer; after each middle layer optimization is completed, the result is fed back to the outer layer. This multi-layer nested optimization strategy can fully consider the mutual influence between each part, and ultimately achieve global performance optimization of heterogeneous integrated circuits.

[0139] Through this adaptive algorithm-driven dynamic parameter adjustment process, the advantages of heterogeneous heterogeneous integration technology can be fully utilized on the basis of considering electromagnetic coupling and parasitic effects, and the performance of microwave systems can be significantly improved. At the same time, this method also provides a systematic and automated optimization path for the design of more complex heterogeneous heterogeneous integrated circuits in the future.

[0140] Figure 2 FIG. 1 is a schematic diagram of a design system for heterogeneous integrated circuits of a microwave system according to an embodiment of the present invention. Figure 2 As shown, the system comprises:

[0141] The first unit is used to establish a multi-layer structure model of heterogeneous integrated circuits of microwave systems, which includes a substrate layer, an active device layer, a passive device layer and an interconnection layer; a plurality of different material regions are set on the substrate layer, and each material region is used to integrate different types of active devices; in the active device layer, silicon-based CMOS devices, gallium arsenide-based HEMT devices and gallium nitride-based HEMT devices are selected and arranged according to preset microwave system performance indicators; in the passive device layer, a passive device network including inductors, capacitors and transformers is designed and arranged; in the interconnection layer, a multi-layer metal interconnection structure is designed to achieve low-loss connection between active devices and passive devices;

[0142] The second unit is used to build an electromagnetic simulation model of a heterogeneous integrated circuit of a microwave system based on the established multi-layer structure model; model various active devices in the active device layer as lumped parameter models, and model the passive device network in the passive device layer as a distributed parameter model; use the finite element method to perform full-wave electromagnetic simulation on the entire heterogeneous integrated circuit to obtain the electromagnetic coupling effect and parasitic effect between each device; optimize the layout of active and passive devices based on the electromagnetic simulation results, and adjust the metal routing of the interconnect layer to reduce signal interference and parasitic effects;

[0143] The third unit is used to evaluate and optimize the performance of heterogeneous integrated circuits in microwave systems based on the optimized electromagnetic simulation model; establish a multivariable optimization model that includes temperature changes, process deviations and load changes; use the Monte Carlo method to perform large-scale random simulations to evaluate the performance stability of microwave systems under different working conditions; based on the simulation results, use adaptive algorithms to dynamically adjust the key parameters of heterogeneous integrated circuits, including adjusting the bias voltage of active devices, the impedance matching network of passive devices and the transmission line structure of the interconnection layer; through iterative optimization, the performance indicators of heterogeneous integrated circuits in microwave systems meet the preset requirements, while ensuring the stability and reliability of the system under different working conditions.

[0144] According to a third aspect of the embodiments of the present invention,

[0145] An electronic device is provided, comprising:

[0146] processor;

[0147] a memory for storing processor-executable instructions;

[0148] The processor is configured to call the instructions stored in the memory to execute the aforementioned method.

[0149] A fourth aspect of the embodiments of the present invention is:

[0150] A computer-readable storage medium is provided, on which computer program instructions are stored. When the computer program instructions are executed by a processor, the aforementioned method is implemented.

[0151] The present invention may be a method, an apparatus, a system and / or a computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for executing various aspects of the present invention.

[0152] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A design method for heterogeneous integrated circuits for microwave systems, characterized in that: include: A multi-layer structure model of heterogeneous integrated circuits of microwave systems is established, and the multi-layer structure model includes a substrate layer, an active device layer, a passive device layer and an interconnection layer; a plurality of different material regions are set on the substrate layer, and each material region is used to integrate different types of active devices; in the active device layer, silicon-based CMOS devices, gallium arsenide-based HEMT devices and gallium nitride-based HEMT devices are selected and arranged according to preset microwave system performance indicators; in the passive device layer, a passive device network including inductors, capacitors and transformers is designed and arranged; in the interconnection layer, a multi-layer metal interconnection structure is designed to achieve low-loss connection between active devices and passive devices; Based on the established multi-layer structure model, an electromagnetic simulation model of heterogeneous integrated circuits in microwave systems is constructed; various active devices in the active device layer are modeled as lumped parameter models, and the passive device network in the passive device layer is modeled as a distributed parameter model; the finite element method is used to perform full-wave electromagnetic simulation of the entire heterogeneous integrated circuit to obtain the electromagnetic coupling effect and parasitic effect between each device; according to the electromagnetic simulation results, the layout of active and passive devices is optimized, and the metal routing of the interconnect layer is adjusted to reduce signal interference and parasitic effects; Based on the optimized electromagnetic simulation model, the performance of heterogeneous integrated circuits in microwave systems is evaluated and optimized; a multivariable optimization model including temperature changes, process deviations and load changes is established; large-scale random simulations are performed using the Monte Carlo method to evaluate the performance stability of microwave systems under different working conditions; based on the simulation results, an adaptive algorithm is used to dynamically adjust the key parameters of heterogeneous integrated circuits, including the bias voltage of active devices, the impedance matching network of passive devices and the transmission line structure of the interconnection layer; through iterative optimization, the performance indicators of heterogeneous integrated circuits in microwave systems meet the preset requirements, while ensuring the stability and reliability of the system under different working conditions.

2. The method according to claim 1, characterized in that A multi-layer structure model of heterogeneous integrated circuits of microwave systems is established, and the multi-layer structure model includes a substrate layer, an active device layer, a passive device layer and an interconnection layer; a plurality of different material regions are set on the substrate layer, and each material region is used to integrate different types of active devices; in the active device layer, silicon-based CMOS devices, gallium arsenide-based HEMT devices and gallium nitride-based HEMT devices are selected and arranged according to the preset microwave system performance indicators; in the passive device layer, a passive device network including inductors, capacitors and transformers is designed and arranged; in the interconnection layer, a multi-layer metal interconnection structure is designed to achieve low-loss connection between active devices and passive devices, including: Selective epitaxial growth technology is used on a silicon substrate to grow high-quality gallium arsenide film and gallium nitride film to form multiple regions of different materials; lattice matching between the gallium arsenide film and the gallium nitride film and the silicon substrate is achieved by controlling epitaxial growth parameters; buffer layers are introduced at the interface between the gallium arsenide film and the silicon substrate and at the interface between the gallium nitride film and the silicon substrate respectively; the buffer layer is an AlGaN / GaN superlattice structure; A silicon-based CMOS device, a gallium arsenide-based HEMT device, and a gallium nitride-based HEMT device are respectively prepared on the plurality of different material regions; wherein, for the gallium nitride-based HEMT device, a field plate structure and a passivation technology are used to improve the breakdown voltage and power density, and a T-type gate structure and a gate length of less than 100 nanometers are used to improve the cutoff frequency and the maximum oscillation frequency; A passive device layer is arranged on the silicon-based CMOS device, the gallium arsenide-based HEMT device and the gallium nitride-based HEMT device; in the passive device layer, a high-Q value inductor is prepared by combining a suspended spiral structure with a magnetic material substrate, a high-density MIM capacitor with a multi-layer stacking structure is prepared by using a high dielectric constant material, the high dielectric constant material is selected from HfO2 and Al 2O3, and a transformer is prepared by using a symmetrical structure and magnetic shielding technology; A multi-layer metal interconnection layer is arranged on the passive device layer; in the multi-layer metal interconnection layer, copper is used as the metal wire material, and the characteristic impedance is controlled by optimizing the line width and spacing; a coplanar waveguide structure or a microstrip line structure is used on the key signal path; an electromagnetic shielding structure is introduced into the multi-layer metal interconnection layer, and the electromagnetic shielding structure includes a grounded shielding wall and metal filling; The silicon-based CMOS device, the gallium arsenide-based HEMT device, the gallium nitride-based HEMT device and the device in the passive device layer are interconnected through the multi-layer metal interconnection layer to form a complete microwave system heterogeneous integrated circuit.

3. The method according to claim 1, characterized in that Based on the established multi-layer structure model, an electromagnetic simulation model of heterogeneous integrated circuits of microwave systems is constructed; various active devices in the active device layer are modeled as lumped parameter models, and the passive device network in the passive device layer is modeled as a distributed parameter model, including: Establishing a lumped parameter model of the active device layer, including modeling of silicon-based CMOS devices, gallium arsenide-based HEMT devices, and gallium nitride-based HEMT devices; wherein, for gallium nitride-based HEMT devices, a small signal equivalent circuit model including gate-source capacitance, gate-drain capacitance, drain-source capacitance, transconductance, output conductance, and parasitic inductance and resistance is established; thermal resistance and thermal capacitance are introduced into the small signal equivalent circuit model to describe the self-heating effect, and an RC network is added to simulate the trap effect; Establish a distributed parameter model of the passive device layer, including modeling of inductors, capacitors, and transformers. For high-Q spiral inductors, establish an accurate geometric model in 3D electromagnetic simulation software, including the cross-sectional shape of the conductor, the properties of the substrate material, and the surrounding metal structure. Use the vector fitting method to convert the full-wave electromagnetic simulation results into an equivalent circuit model. In the substrate layer model, different dielectric constants and loss tangents are defined for different material regions; additional interface impedance layers are introduced at the material interface to simulate the effects of interface stress and defects on electromagnetic properties; Integrate the lumped parameter model of active devices and the distributed parameter model of passive devices in circuit simulation software; insert the coupling matrix or parasitic parameter network obtained by electromagnetic simulation at key nodes to consider the electromagnetic coupling and parasitic effects in heterogeneous integrated circuits; Design and fabricate test structures, including various active and passive devices and typical interconnect structures; evaluate model accuracy by comparing simulation results with measured data, including S-parameters, noise figure and power gain; Based on the evaluation results, the model parameters are optimized and calibrated until the deviation between the simulation results and the measured data meets the preset accuracy requirements.

4. The method according to claim 3, characterized in that The finite element method is used to perform full-wave electromagnetic simulation on the entire heterogeneous integrated circuit to obtain the electromagnetic coupling effects and parasitic effects between various devices, including: Establishing a three-dimensional geometric model of a heterogeneous integrated circuit, the three-dimensional geometric model includes a silicon substrate, a selectively epitaxially grown gallium arsenide region, a selectively epitaxially grown gallium nitride region, an active device layer, a passive device layer, and a multi-layer metal interconnect structure; wherein the active device layer includes simplified geometric models of CMOS devices, gallium arsenide HEMT devices, and gallium nitride HEMT devices, and the passive device layer includes accurate geometric models of high-Q inductors, high-density MIM capacitors, and transformers; Specifying electromagnetic parameters for each material in the three-dimensional geometric model, including dielectric constant, conductivity and magnetic permeability; establishing a nonlinear characteristic model for nonlinear materials; setting radiation boundary conditions or absorption boundary conditions at the boundaries of the three-dimensional geometric model; and setting impedance boundary conditions for metal structures; Meshing the three-dimensional geometric model, using adaptive meshing technology, using fine meshes at device interfaces and high-frequency current intensive areas, and using relatively coarse meshes in non-critical areas; performing mesh convergence analysis to ensure that the mesh quality meets the preset accuracy requirements; Set the port excitation and select Gaussian pulse as the excitation source; set the solution frequency points to cover the entire frequency band of the device operation, and set more dense frequency points near the resonant frequency of the high Q value structure; Use the finite element method to discretize continuous electromagnetic field problems into a system of linear equations; for time domain analysis, use explicit or implicit time integration methods to solve; for frequency domain analysis, directly solve the complex linear equations; use parallel computing technology and iterative solvers to improve computational efficiency; Extract S parameters, impedance matrix and electromagnetic field distribution from full-wave electromagnetic simulation results; analyze the mutual inductance between adjacent spiral inductors and the parasitic capacitance between HEMT devices and nearby wires; identify hot spots and electromagnetic interference problems through current density distribution; A parasitic effect network model is constructed based on the simulation results, and the extracted parasitic parameters are converted into an equivalent circuit model, including coupling capacitance, mutual inductance and parasitic resistance; for complex coupling structures, multi-port network parameter description is used.

5. The method according to claim 1, characterized in that According to the simulation results, an adaptive algorithm is used to dynamically adjust the key parameters of heterogeneous integrated circuits, including the bias voltage of active devices, the impedance matching network of passive devices, and the transmission line structure of the interconnect layer. Establishing a full-wave electromagnetic model of a heterogeneous integrated circuit, wherein the full-wave electromagnetic model includes accurate geometric structures and material parameters of active devices, passive devices and interconnect layers; defining an outer optimization objective function for evaluating active device performance, wherein the objective function is a weighted sum of output power, power added efficiency, and third-order intermodulation distortion; The bias voltage of the active device is optimized by using the gradient descent algorithm, which calculates the partial derivatives of the objective function with respect to the gate-source voltage and the drain-source voltage and iteratively updates the gate-source voltage and the drain-source voltage in combination with the learning rate. A middle-layer optimization fitness function is defined to evaluate the performance of the passive device impedance matching network, wherein the fitness function is a function of the scattering parameter; The impedance matching network of passive devices is optimized by using genetic algorithm. The parameters of matching network components are encoded as chromosomes. A new generation of population is generated through selection, crossover and mutation operations. The performance of each individual is evaluated using fitness function. An inner layer optimization objective function is defined to evaluate the performance of the interconnect layer transmission line structure, wherein the objective function includes a weighted square sum of a difference between a characteristic impedance and a target impedance, a difference between a transmission loss and a maximum allowable loss, and a difference between a phase velocity and a theoretical phase velocity; The particle swarm optimization algorithm is used to optimize the transmission line structure of the interconnect layer. The transmission line geometric parameters are encoded as particle positions. The transmission line structure is iteratively optimized by calculating the particle speed and position update formula. Construct a multi-layer nested optimization structure, where the outer loop controls the optimization of active devices, the middle loop is responsible for matching network optimization, and the inner loop handles the optimization of transmission line structure; each time the inner optimization is completed, the optimization result is fed back to the middle loop; each time the middle optimization is completed, the optimization result is fed back to the outer loop; In each layer of optimization process, full-wave electromagnetic simulation is used to obtain performance indicators, and the optimization parameters are updated according to the simulation results; the iteration continues until the preset termination condition is met or the maximum number of iterations is reached; The final optimization results are output, including the optimal bias voltage of active devices, the optimal impedance matching network parameters of passive devices, and the optimal transmission line structure parameters of the interconnect layer.

6. A design system for heterogeneous integrated circuits of a microwave system, used to implement the method according to any one of claims 1 to 5, characterized in that: include: The first unit is used to establish a multi-layer structure model of heterogeneous integrated circuits of microwave systems, which includes a substrate layer, an active device layer, a passive device layer and an interconnection layer; a plurality of different material regions are set on the substrate layer, and each material region is used to integrate different types of active devices; in the active device layer, silicon-based CMOS devices, gallium arsenide-based HEMT devices and gallium nitride-based HEMT devices are selected and arranged according to preset microwave system performance indicators; in the passive device layer, a passive device network including inductors, capacitors and transformers is designed and arranged; in the interconnection layer, a multi-layer metal interconnection structure is designed to achieve low-loss connection between active devices and passive devices; The second unit is used to build an electromagnetic simulation model of a heterogeneous integrated circuit of a microwave system based on the established multi-layer structure model; model various active devices in the active device layer as lumped parameter models, and model the passive device network in the passive device layer as a distributed parameter model; use the finite element method to perform full-wave electromagnetic simulation on the entire heterogeneous integrated circuit to obtain the electromagnetic coupling effect and parasitic effect between each device; optimize the layout of active and passive devices based on the electromagnetic simulation results, and adjust the metal routing of the interconnect layer to reduce signal interference and parasitic effects; The third unit is used to evaluate and optimize the performance of heterogeneous integrated circuits in microwave systems based on the optimized electromagnetic simulation model; establish a multivariable optimization model that includes temperature changes, process deviations and load changes; use the Monte Carlo method to perform large-scale random simulations to evaluate the performance stability of microwave systems under different working conditions; based on the simulation results, use adaptive algorithms to dynamically adjust the key parameters of heterogeneous integrated circuits, including adjusting the bias voltage of active devices, the impedance matching network of passive devices and the transmission line structure of the interconnection layer; through iterative optimization, the performance indicators of heterogeneous integrated circuits in microwave systems meet the preset requirements, while ensuring the stability and reliability of the system under different working conditions.

7. An electronic device, characterized in that: include: processor; a memory for storing processor-executable instructions; The processor is configured to call the instructions stored in the memory to execute the method described in any one of claims 1 to 5.

8. A computer-readable storage medium having computer program instructions stored thereon, characterized in that: When the computer program instructions are executed by a processor, the method according to any one of claims 1 to 5 is implemented.

Images