A broadband low-noise amplifier end-to-end design method and apparatus

Abstract

This application provides an end-to-end design method and apparatus for a broadband low-noise amplifier. The method includes: matching the topology of the broadband low-noise amplifier based on a lookup table; optimizing the first parameters of active components in the broadband low-noise amplifier based on the topology; matching the range of second parameters of passive components in the broadband low-noise amplifier based on the first parameters of the active components; optimizing the physical parameters of the transformer using an adaptive weighted binary search algorithm, with the range of second parameters of the passive components as a pre-layout reference value; and assembling the active and passive components by interconnecting and wiring according to the first parameters, second parameters, and physical parameters and based on layout rules, to obtain the broadband low-noise amplifier layout. The end-to-end design method and apparatus for broadband low-noise amplifiers provided in this application address the problems of low efficiency, narrow applicability, and susceptibility to local optima in existing design methods, significantly improving design efficiency and optimization quality, and achieving globally optimal performance.

Description

Technical Field

[0001] This application relates to the field of integrated circuit design technology, and in particular to an end-to-end design method and apparatus for a broadband low-noise amplifier. Background Art

[0002] In the wave of 5G / 6G technology, high-throughput wireless communication and high-resolution sensing applications place stringent demands on the performance of broadband low-noise amplifiers (LNAs), urgently requiring wider bandwidth, lower noise figures, and superior S11 performance to ensure high-quality amplification and transmission of weak signals. However, traditional broadband LNA designs rely on experience-driven optimization strategies, requiring repeated trial-and-error parameter adjustments. As performance standards rise and bandwidth expands, design complexity increases exponentially, making it difficult to meet modern needs. Therefore, there is an urgent need for efficient end-to-end design solutions to accelerate the optimization process of active / passive circuits.

[0003] Currently, while commercial electronic design automation tools such as ADS and Cadence have the ability to optimize circuit parameters, they lack end-to-end automation processes and are prone to getting trapped in local optima in broadband low-noise amplifier design. Some studies have attempted to use artificial intelligence techniques, such as end-to-end passive component placement and synthesis methods based on neural networks, but these are only applicable to simple circuit topologies and the preparation and pre-training of sample datasets are extremely time-consuming. Other methods, such as hierarchical multi-objective optimization methods, involve a large number of parameters, resulting in a lengthy and inefficient optimization process.

[0004] Therefore, there is an urgent need for a method to solve the problems of low efficiency, narrow applicability, and susceptibility to local optima in existing design methods, so as to significantly improve design efficiency and optimization quality and achieve global optimal performance. Summary of the Invention

[0005] In view of this, this application provides an end-to-end design method and apparatus for broadband low-noise amplifiers to solve the problems of low efficiency, narrow applicability and susceptibility to local optima in existing design methods, thereby significantly improving design efficiency and optimization quality and achieving global optimal performance.

[0006] Specifically, this application is implemented through the following technical solution:

[0007] The first aspect of this application provides a broadband low-noise amplifier end-to-end design method, the method comprising:

[0008] The topology of a broadband low-noise amplifier is matched based on a lookup table.

[0009] Optimize the first parameter of the active element in the broadband low-noise amplifier based on the topology;

[0010] The first parameter of the active element is matched to the second parameter range of the passive element in the broadband low-noise amplifier.

[0011] Using the second parameter range of the passive component as a pre-layout reference value, the physical parameters of the transformer are optimized using an adaptive weighted binary search algorithm;

[0012] Based on the first parameter, the second parameter, and the physical parameters, the active and passive components are interconnected and assembled according to the layout rules to obtain a broadband low-noise amplifier layout.

[0013] A second aspect of this application provides an end-to-end design apparatus for a broadband low-noise amplifier, the apparatus comprising a matching module, an optimization module, and an assembly module;

[0014] The matching module is used to match the topology of the broadband low-noise amplifier based on a lookup table.

[0015] The optimization module is used to optimize the first parameters of the active components in the broadband low-noise amplifier based on the topology.

[0016] The matching module is also used to match the second parameter range of the passive components in the broadband low-noise amplifier based on the first parameter of the active component.

[0017] The optimization module is also used to optimize the physical parameters of the transformer using an adaptive weighted binary search algorithm, with the second parameter range of the passive component as the pre-layout reference value.

[0018] The assembly module is used to interconnect and assemble the active and passive components based on the first parameter, the second parameter, and the physical parameters, according to layout rules, to obtain a broadband low-noise amplifier layout.

[0019] The broadband low-noise amplifier end-to-end design method and apparatus provided in this application firstly utilize a lookup table to match the topology, enabling rapid selection of a suitable circuit topology. This avoids the uncertainty and redundant calculations associated with manually selecting topologies in complex circuit design processes, reducing time and computational burden in the initial design phase. Simultaneously, the lookup table method ensures that each selected topology is validated, minimizing the risk of design failure. Next, the first parameters of the active components are optimized based on the selected topology. This step is closely integrated with the topology, avoiding parameter mismatch issues that may occur in traditional designs. Based on the first parameters of the active components, the second parameter range of the passive components is matched, further improving the overall design consistency. In the design of broadband low-noise amplifiers, active and passive components must be well-matched to ensure circuit stability and performance at different frequencies. By optimizing the second parameter range, the electrical performance of the passive components is ensured to be consistent with the operating state of the active components. This method not only improves circuit reliability but also reduces performance degradation caused by parameter mismatch, ensuring the low-noise and high-gain characteristics of the broadband low-noise amplifier.

[0020] Furthermore, the physical parameters of the transformer are gradually optimized by adaptively adjusting the weights, and convergence to the optimal solution is achieved through continuous parameter updates. This algorithm allows for dynamic adaptation to the actual needs of the circuit during optimization and accelerates convergence, avoiding over-computation and inefficiency issues in traditional optimization methods. Finally, based on the first parameter, the second parameter, and the physical parameters, and combined with layout rules, the interconnection and routing of active and passive components are optimized, resulting in a well-designed circuit layout. By rationally arranging components according to design rules, not only can space utilization efficiency be improved, but also parasitic effects, signal interference, and performance losses caused by improper layout can be effectively reduced. Simultaneously, the optimized layout better ensures the thermal stability and electromagnetic compatibility of the circuit, providing a strong guarantee for the final broadband low-noise amplifier performance. In summary, the entire design process forms a mutually supportive closed-loop system. From topology selection to parameter optimization and physical layout optimization, each step lays the foundation for the next optimization step, ensuring close integration between each link and ultimately achieving a highly efficient and high-performance broadband low-noise amplifier design method. This end-to-end design process can significantly shorten the design cycle, improve the accuracy and reliability of the design, and ensure the stability and high efficiency of the broadband low-noise amplifier in practical applications. Attached Figure Description

[0021] Figure 1 A flowchart of the broadband low-noise amplifier end-to-end design method provided in Embodiment 1 of this application;

[0022] Figure 2A schematic diagram of the broadband low-noise amplifier provided in this application;

[0023] Figure 3 A schematic diagram of the layout of the broadband low-noise amplifier provided in this application;

[0024] Figure 4 This is a schematic diagram of the structure of the broadband low-noise amplifier end-to-end design device provided in Embodiment 2 of this application. Detailed Implementation

[0025] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application.

[0026] The terminology used in this application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. The singular forms “a,” “the,” and “the” used herein are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any and all possible combinations of one or more of the associated listed items.

[0027] It should be understood that although the terms first, second, third, etc., may be used in this application to describe various information, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, without departing from the scope of this application, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Depending on the context, the word "if" as used herein may be interpreted as "when," "when," or "in response to determination."

[0028] The following specific embodiments are given to illustrate the technical solution of this application in detail.

[0029] Figure 1 This is a flowchart illustrating the end-to-end design method for a broadband low-noise amplifier provided in Embodiment 1 of this application. Please refer to... Figure 1 The method provided in this embodiment may include:

[0030] S101. Topology of a broadband low-noise amplifier based on lookup table matching.

[0031] Specifically, the lookup table stores the correspondence between specific inputs and outputs. In this embodiment, the lookup table is used to quickly determine the topology of a broadband low-noise amplifier that matches given operating frequencies and other conditions. The lookup table typically contains input information such as different operating frequency ranges and performance requirements, along with corresponding suitable broadband low-noise amplifier topology information. For example, for a specific frequency range and desired performance parameters such as bandwidth, noise figure, and gain, the lookup table records which topology best meets these requirements.

[0032] Furthermore, a broadband low-noise amplifier (LNOA) is an electronic circuit device primarily used to amplify weak high-frequency signals while minimizing introduced noise. In 5G / 6G communication technologies and high-resolution sensing applications, LNOA can achieve low-noise signal amplification over a wide frequency range to meet the stringent signal quality requirements of high-throughput wireless communication and high-resolution sensing. Topology refers to the connection method and layout of the various components (such as transistors, resistors, capacitors, and inductors) in a broadband LNOA. Different topologies have different performance characteristics; for example, some topologies may excel in broadband performance, while others are more advantageous in low-noise performance. Common broadband LNOA topologies include cascode and inductive negative feedback structures.

[0033] In practice, the design requirements of the broadband low-noise amplifier are obtained, including center frequency, target bandwidth, expected gain, maximum noise figure, and power consumption limit. These requirements are input into a lookup table for matching calculations, retrieving the topology that best matches the input design requirements. If a perfectly matching topology exists, it is directly selected as the topology for the broadband low-noise amplifier. If no perfectly matching topology exists, a tolerance or scoring mechanism is introduced to select the optimal matching topology.

[0034] Optionally, before matching the topology of the broadband low-noise amplifier based on the lookup table, the method further includes: determining the key performance indicators of the broadband low-noise amplifier based on design target requirements; collecting the topologies of the broadband low-noise amplifier and recording the key performance indicators corresponding to each topology; constructing a database containing the topologies and key performance indicators, and establishing a lookup table based on the database.

[0035] In practical implementation, the design objectives and requirements are determined based on the application scenario of the broadband low-noise amplifier to be designed. Based on these requirements, key performance indicators (KPIs) of the broadband low-noise amplifier are extracted, including but not limited to gain, noise figure, power consumption, bandwidth, and input-output matching. Existing topologies of broadband low-noise amplifiers are obtained through relevant literature and public databases, and the KPIs corresponding to each topology under standard simulation conditions are recorded. The collected topologies and their corresponding KPIs are entered into a database, forming a structure-performance mapping relationship. Based on the established database, a lookup table for fast topology lookup is created, with each entry including a topology identifier and KPIs.

[0036] S102. Optimize the first parameters of the active elements in the broadband low-noise amplifier based on the topology.

[0037] Specifically, active components refer to electronic components in a broadband low-noise amplifier that amplify signals. These typically include transistors (such as BJTs or FETs) or the amplification section of integrated circuits. Active components increase the strength of the input signal through current or voltage amplification, and are the core component for achieving amplification. For broadband low-noise amplifiers, the performance of active components directly determines important performance characteristics such as gain, noise level, and linearity.

[0038] The first parameter refers to a key performance parameter that needs to be determined first when optimizing active components. It typically includes transconductance (gm) and drain current (ID). Transconductance (gm) is an important parameter reflecting the amplification capability of active components, representing the relationship between current change and input voltage change; drain current (ID) represents the amount of current passing through the transistor and has a direct impact on the gain and power consumption of broadband low-noise amplifiers.

[0039] It should be noted that the primary parameters of active components (such as transconductance and drain current) directly affect the gain, noise, and other performance characteristics of broadband low-noise amplifiers. Determining these primary parameters helps ensure optimal performance of the broadband low-noise amplifier within its operating frequency band. Furthermore, optimizing these primary parameters based on the design topology (e.g., the type of broadband low-noise amplifier architecture used) ensures that the selected active components are matched with other parts of the circuit design (such as passive components) to achieve optimal operation.

[0040] In specific implementation, optimizing the first parameters of the active components in the broadband low-noise amplifier based on the topology includes: determining the basic parameters of the active components based on the topology; analyzing the importance of different key performance indicators according to the application scenario of the broadband low-noise amplifier, and determining the weights of different key performance indicators based on the importance; constructing a comprehensive performance evaluation function based on the basic parameters and the weights of different key performance indicators using the transconductance-drain current ratio method; adjusting the transconductance-drain current ratio through iterative calculation, and calculating the comprehensive performance score under different combinations of the first parameters according to the comprehensive performance evaluation function in each iteration, until the number of convergences reaches a preset number, and selecting the first parameter combination with the optimal comprehensive performance score as the first parameter of the active component.

[0041] Specifically, firstly, based on the selected topology, the basic parameters of the active components are determined, including model, size, and threshold voltage. Then, based on the application scenario of the broadband low-noise amplifier (e.g., communication, radar), the relative importance of key performance indicators (e.g., gain, noise figure, bandwidth, power consumption) is determined. Each key performance indicator is assigned a weight based on its relative importance ratio, reflecting its importance in the design process. Further, a comprehensive performance evaluation function is constructed using the transconductance-drain current ratio method. Transconductance (gm) and drain current (ID) are selected as the main optimization parameters. A comprehensive performance evaluation function is constructed using the ratio of transconductance to drain current, considering the weights of the key performance indicators. The comprehensive performance evaluation function includes factors such as gain, noise figure, and power consumption, which are weighted and summed according to different weights. The comprehensive performance evaluation function can be expressed as:

[0042] ;

[0043] Among them, the This is a comprehensive performance evaluation function; the... For transconductance; the The drain current ratio; The gain is calculated based on the transconductance and drain current ratio; The noise figure is calculated based on the transconductance and drain current ratio; The power consumption is calculated based on the transconductance and drain current ratio; , , Weights for different key performance indicators.

[0044] Furthermore, initial values ​​for transconductance and drain current are initialized. In each iteration, based on the comprehensive performance evaluation function, the comprehensive performance score is calculated for different combinations of first parameters (different combinations of transconductance and drain current). Based on the evaluation results, the ratio of transconductance to drain current is adjusted to optimize the comprehensive performance. Iterative calculations continue until a preset number of convergences is reached, or iteration stops when the change in the comprehensive performance score falls below a preset threshold. In the final iteration results, the first parameter combination with the optimal comprehensive performance score is selected as the first parameter of the active component.

[0045] S103. Match the second parameter range of the passive element in the broadband low-noise amplifier based on the first parameter of the active element.

[0046] Specifically, passive components refer to electronic components that do not rely on an external power source for amplification or energy conversion. They typically do not provide gain or power amplification functions and are used for signal processing, filtering, and regulation in circuits. Common passive components include resistors, capacitors, inductors, transformers, filters, couplers, and splitters.

[0047] The second parameter refers to a key parameter related to circuit design and performance tuning. It is typically associated with the type of passive component. For example, for a capacitor, the second parameter might be capacitance. For a resistor, it might be resistance. For an inductor, it might be inductance. For a filter, it might be cutoff frequency, bandwidth, or Q-value. In the design of broadband low-noise amplifiers, the second parameter usually refers to the characteristics of passive components that determine circuit performance. These parameters are often related to filtering, matching network design, etc., and affect the bandwidth, gain, and noise characteristics of the broadband low-noise amplifier.

[0048] It is important to note that passive components directly affect the frequency response, bandwidth, gain, and noise characteristics of a circuit. By adjusting the signal path, achieving impedance matching, filtering, and suppressing unwanted frequency components, passive components ensure effective signal transmission and processing during amplification. Therefore, the proper selection and optimization of the second parameter of passive components can effectively improve the overall performance of the amplifier, especially in terms of bandwidth, gain, and noise suppression. Furthermore, the determination of the second parameter is closely related to the first parameter (i.e., the parameter of the active component). The first parameter is usually a key factor determining the operating state of the active component, such as transconductance and drain current, which directly affect the amplifier's gain and power consumption. Since active and passive components cooperate in the circuit, optimizing the first parameter affects the circuit's operating point, thus influencing the operating state of the passive component. Therefore, the determination of the second parameter must be coordinated with the first parameter. By adjusting the second parameter of the passive component, the overall circuit performance can be optimized. This interaction needs to be continuously adjusted through simulation and optimization during the design process to ensure that the gain, noise, bandwidth, and other performance characteristics of the entire circuit meet the design requirements.

[0049] In specific implementation, matching the second parameter range of the passive components in the broadband low-noise amplifier based on the first parameter of the active component includes: determining the input port and output port based on the first parameter of the active component; extracting the real and imaginary parts from the input and output ports and modeling them as equivalent resistance-capacitance networks; calculating the initial value of the second parameter based on the equivalent resistance-capacitance network; optimizing the second parameter related to the magnetically coupled resonator at each stage using a particle swarm optimization algorithm; iteratively calculating based on the particle dynamic change formula during the optimization process; matching the overall gain difference of the frequency corresponding to the gain peak of the magnetically coupled resonator at each stage; calculating the second parameter corresponding to the frequency; using the particle swarm optimization algorithm, based on the equivalent resistance-capacitance network and the circuit state after optimization of the magnetically coupled resonator at each stage, iteratively calculating based on the particle dynamic change formula to obtain the remaining second parameter; combining capacitors greater than a first preset value into a capacitor array, and resistors less than a second preset value into a resistor array; organizing all the second parameters obtained after optimization; and determining the second parameter range of the passive components in the broadband low-noise amplifier.

[0050] Specifically, based on the determined first parameters of the active components, a corresponding topology model is built in the circuit simulation software. AC small-signal analysis is used to extract the impedances of the input and output ports, obtaining their real parts (equivalent resistance) and imaginary parts (equivalent reactance). The resistance and reactance are then converted to capacitance and inductance using conversion formulas, completing the modeling of the equivalent resistance-capacitance network. Based on the parameter values ​​extracted from the equivalent resistance-capacitance network, the required initial values ​​of inductance and capacitance are estimated using impedance matching formulas for matching networks (such as conjugate matching), yielding the initial values ​​of the second parameters. Further, for each stage of the magnetically coupled resonator, the passive parameters to be optimized (such as resonant inductance, capacitance, coupling inductance, coupling coefficient, etc.) are determined. A particle swarm optimization model is built in the simulation software, where each particle represents a parameter combination. The objective function is designed as: the gain reaches its peak at multiple frequency points. The target value for each particle is obtained using a circuit simulator or a custom gain calculation function. The particle swarm is iteratively updated according to the standard PSO formula until convergence, outputting the optimal second parameter for each stage of the magnetically coupled resonator. Based on the known second parameters and the optimized magnetically coupled resonator state, the complete circuit structure is rebuilt. An initial search range is set for other passive components, and the same PSO optimization strategy is used. An objective function (e.g., maximizing power gain, minimizing noise figure) is selected, and iterative optimization is performed to obtain the remaining parameters. Finally, it is determined whether the capacitance in the optimization result is greater than the first preset threshold (e.g., >70pF). If so, multiple smaller capacitors are connected in parallel to form the target capacitor, creating a capacitor array. Similarly, it is determined whether the resistance is less than the second preset threshold (e.g., <3kΩ). If so, multiple large resistors are connected in series to form the target resistor, creating a resistor array. All the final optimized second parameters are then compiled and used as reference values ​​for optimizing the third parameter.

[0051] Optionally, the step of using a particle swarm optimization algorithm to optimize the second parameters related to the magnetically coupled resonator at each stage includes: determining the second parameters related to the first magnetically coupled resonator; determining the first gain peak value of the first magnetically coupled resonator; determining the first frequency based on the first gain peak value; starting from an initial value, optimizing the second parameters of the first magnetically coupled resonator using a particle swarm optimization algorithm; iteratively calculating the second parameters that reach the first gain peak value at the first frequency based on the particle swarm dynamic change formula during the optimization process; and determining the second parameters that reach the first gain peak value at the first frequency. Determine the second parameters related to the second magnetically coupled resonator, determine the second peak gain of the second magnetically coupled resonator, and determine the second and third frequencies based on the second peak gain. Starting from the initial value, and combining the optimized circuit state of the first magnetically coupled resonator, calculate the second parameters that reach the second peak gain at the second and third frequencies. Determine the second parameters related to the third magnetically coupled resonator, determine the second parameters related to the fourth magnetically coupled resonator, and calculate the overall gain difference between the third and fourth magnetically coupled resonators between the second and third frequencies. Starting from the initial value, and combining the optimized circuit state of the first and second magnetically coupled resonators, calculate the second parameters when the overall gain difference reaches its minimum value.

[0052] Specifically, in the first stage of optimizing the second parameters of the first magnetically coupled resonator, the initial values ​​of the second parameters extracted from the equivalent resistance-capacitance network are used as input. Initial inductance, capacitance, and other parameters related to the first magnetically coupled resonator are extracted and set as the second parameters to be optimized. Through circuit simulation or response analysis, the first gain peak of the first magnetically coupled resonator is determined, and the corresponding target frequency (denoted as the first frequency) is further extracted. Particle swarm optimization is then performed, setting the optimization objective to achieve the gain peak of the first magnetically coupled resonator at the first frequency. The particle swarm is initialized, with the particle positions representing different combinations of the second parameters. Based on the particle dynamic change formula (including velocity and position updates), iteration is performed, calculating the gain response under the current particle combination in each iteration, and recording the optimal particle. When the iteration reaches convergence or the maximum number of iterations is reached, the optimal second parameters of the first magnetically coupled resonator at the first frequency are obtained. Second stage: Parameter optimization of the second magnetically coupled resonator.

[0053] After the first stage of optimization, the second stage of optimization continues based on the optimized second parameters of the first magnetically coupled resonator, maintaining the same circuit state. Through circuit response analysis, two target frequencies for the second magnetically coupled resonator are determined: the second frequency and the third frequency (representing the bandwidth they cover). Further particle swarm optimization is performed, with the optimization objective being to achieve the peak response of the second magnetically coupled resonator at both the second and third frequencies. First, the particle swarm is initialized using the output from the first stage. During iteration, the optimization objective is set to ensure that the gain at both frequencies simultaneously reaches its peak. A composite objective function is used to evaluate the response at multiple frequencies, ultimately obtaining the optimal second parameters that satisfy the dual-frequency gain condition.

[0054] After the second stage of optimization is completed, based on the circuit states optimized in the first and second stages, the initial second parameters of the third and fourth magnetically coupled resonators are extracted. Further joint particle swarm optimization is then performed: each particle in the particle swarm represents a parameter combination of the third and fourth magnetically coupled resonators. In each iteration, the gain curves of the third and fourth magnetically coupled resonators between the second and third frequencies are simulated; the overall gain difference between the gain curves is calculated, and the optimization objective is to minimize this overall gain difference; the particle swarm parameters are dynamically updated until convergence is achieved, yielding the joint optimization result and determining the second parameters.

[0055] For example, Figure 2 This is a schematic diagram of the broadband low-noise amplifier provided in this application. Please refer to... Figure 2 ,exist Figure 2In this process, there are 32 second parameters to be optimized. First, for the first magnetically coupled resonator, the relevant second parameters are determined as follows: k4, LL4, LR4, CL4, CR4, RL4, RR4. The first frequency is then determined as fc, and the optimization objective is to reach the peak gain at fc. Next, for the second magnetically coupled resonator, the relevant second parameters are determined as follows: k1, LL1, LR1, CL1, CR1, RL1, RR1. The second frequency is then determined as fL, and the third frequency is fH. The optimization objective is to reach the peak gain at both fL and fH. Finally, for the third and fourth magnetically coupled resonators, the relevant second parameters are determined as follows: k2, LL2, LR2, CL2, CR2, RL2, RR2, k3, LL3, LR3, CL3, CR3, RL3, RR3. The optimization objective is to minimize the difference between the maximum and minimum gain points of the overall gain curve within the operating frequency band between fL and fH. Where k is the coupling coefficient, characterizing the degree of coupling between magnetically coupled resonators; LL is the inductance associated with the primary coil in the magnetically coupled resonator; LR is the inductance associated with the secondary coil in the magnetically coupled resonator; CL is the capacitance associated with the primary part of the magnetically coupled resonator; CR is the capacitance associated with the secondary part of the magnetically coupled resonator; RL is the equivalent resistance of the primary coil of the magnetically coupled resonator; and RR is the equivalent resistance of the secondary coil of the magnetically coupled resonator. Further, after optimizing the 28 second parameters, Lg, Lpar, Cff, and Ls are then optimized to achieve input matching. These four second parameters are determined by minimizing NF while ensuring that the overall S11 remains below -8dB. Where Lg is the gate inductance, Lpar is the gate parallel inductance, Cff is the feedback capacitance, and Ls is the source inductance. S11 is one of the scattering parameters used to describe the reflection coefficient at port 1 when a signal is input from port 1 in a multi-port network (such as a broadband low-noise amplifier). S11 reflects how much of the input signal is reflected back to the input terminal, and its value is expressed in decibels (dB). The smaller the S11 value, the less reflection and the better the input matching. NF is an abbreviation for Noise Figure, used to measure the amount of noise introduced by a broadband low-noise amplifier while amplifying the signal. The noise figure is a dimensionless parameter, usually expressed in decibels (dB), and is defined as the ratio of the input signal-to-noise ratio (SNR) to the output SNR. The smaller the NF value, the less noise introduced by the broadband low-noise amplifier itself, and the higher the amplification quality of the signal.

[0056] The method provided in this embodiment achieves efficient solution of the second parameter of passive components in a broadband low-noise amplifier by introducing a step-by-step parameter optimization strategy based on particle swarm optimization algorithm. This optimization process transforms the originally highly complex and dimensional parameter optimization problem into a step-by-step calculation process with physical meaning and phased objectives, significantly reducing computational complexity and convergence difficulty while ensuring global performance indicators are met. In specific implementation, this application first optimizes the key parameters of the first magnetically coupled resonator with gain control at the center frequency as the objective; then, it optimizes the parameters of the second magnetically coupled resonator around the two edge frequencies of low and high frequencies to expand the overall bandwidth; further, to flatten the gain response, it jointly optimizes all parameters of the third and fourth magnetically coupled resonators to minimize the difference between the maximum and minimum gain points within the operating bandwidth, thereby effectively controlling gain fluctuations. The optimization process iteratively solves each step based on the previous stage circuit state using a particle dynamic change formula, ensuring the consistency and feasibility of the optimization results at each stage within the overall circuit framework. Firstly, this application significantly reduces the dimensionality of optimization variables and improves global convergence and optimization efficiency through phased modeling and iteration. Secondly, the optimization objectives at each stage have clear physical meaning, making the correspondence between circuit parameters and frequency response clearer and facilitating optimization path control and result debugging. Thirdly, multi-stage optimization fully considers the mutual influence of parameters at each stage during circuit cascading, making the optimization results closer to actual circuit performance. Fourthly, the optimized second parameter facilitates the subsequent synthesis of device arrays and the physical realization of transformer layout, meeting the requirements of end-to-end automated design. In summary, guided by engineering experience, this application combines particle swarm optimization algorithm with the gain control requirements of magnetically coupled resonators, ensuring circuit performance while balancing design efficiency, algorithm stability, and physical feasibility, providing an efficient and practical solution for the automated design of broadband low-noise amplifiers.

[0057] Optionally, the iterative calculation based on the particle dynamic change formula includes: determining the local optimal solution of the target second parameter at the target iteration number, and calculating a first difference between the local optimal solution and the target second parameter; determining the global optimal solution of the target second parameter at the target iteration number, and calculating a second difference between the global optimal solution and the target second parameter; calculating a first product of a first acceleration coefficient and the first difference, and calculating a second product of a second acceleration coefficient and the second difference; determining the parameter adjustment step size of the target second parameter at the target iteration number, and calculating a third product of the inertia weight and the parameter adjustment step size; determining the sum of the first product, the second product, and the third product as the parameter adjustment step size of the target second parameter at the next target iteration number; and determining the sum of the parameter adjustment step size and the target second parameter at the target iteration number as the target second parameter for the next target iteration number.

[0058] Specifically, the above process can be represented by the following formula:

[0059] ;

[0060] ;

[0061] Among them, the The target number of iterations; The step size for adjusting the second parameter of the target at the next target iteration number; For inertial weights; the The step size for adjusting the second parameter of the target at the target number of iterations; , The first acceleration coefficient and the second acceleration coefficient; , It is a constant; the stated The second parameter of the objective is the local optimal solution for the objective at the objective iteration number; The second parameter is the target number of iterations; The global optimal solution for the second parameter of the objective at the target iteration number; The second parameter of the target is the number of iterations for the next target.

[0062] S104. Using the second parameter range of the passive component as a pre-layout reference value, the physical parameters of the transformer are optimized using an adaptive weighted binary search algorithm.

[0063] Specifically, transformers are commonly used passive circuit components in broadband low-noise amplifiers, primarily for impedance transformation, signal coupling, and bandpass response control. They typically employ an integrated magnetically coupled resonant structure, consisting of two mutually coupled helical inductors. The physical parameters of a transformer refer to its adjustable geometric and structural characteristics during actual layout design and manufacturing, including the linewidth, spacing, number of turns, inner diameter of the helical inductors, and the coupling distance between the two coils. These parameters directly determine the transformer's electromagnetic behavior and equivalent circuit characteristics.

[0064] It should be noted that there is a mapping relationship between the physical parameters and the second parameter. The second parameter reflects the design target's requirements for electrical performance, while the physical parameters are the geometric means to achieve these performances. By optimizing the physical parameters, the transformer can meet the predetermined electrical performance requirements while remaining within manufacturing feasibility.

[0065] In specific implementation, using the second parameter range of the passive component as a pre-layout reference value, the physical parameters of the transformer are optimized using an adaptive weighted binary search algorithm, including: determining a first initial iteration interval for the physical parameters of the transformer to be optimized based on the second parameter range; at the beginning of each iteration, calculating the updated weights based on the initial weights and the weights of the previous iteration; determining the endpoints of the first interval and the second interval based on the first initial iteration interval, calculating the first function value and the second function value of the first interval endpoint and the second interval endpoint respectively, matching the corresponding segmentation point calculation method based on the magnitude of the first function value and the second function value, and calculating the first segmentation point based on the segmentation point calculation method; dividing the first initial iteration interval according to the first segmentation point to obtain two sub-intervals, determining the target sub-interval based on the magnitude of the first function value and the second function value, and based on the target... The process of iteratively calculating the sub-intervals and the updated weights is repeated until the maximum number of iterations is reached, thereby determining the physical parameters of the first transformer. Based on the physical parameters of the first transformer and the range of the second parameter, a second initial iteration interval is determined. Based on the second initial iteration interval, the process of calculating the physical parameters of the first transformer is repeated to determine the physical parameters of the second transformer. Based on the physical parameters of the second transformer and the range of the second parameter, a third initial iteration interval is determined. Based on the third initial iteration interval, the process of calculating the physical parameters of the first transformer is repeated to determine the physical parameters of the third transformer. Based on the physical parameters of the third transformer and the range of the second parameter, a fourth initial iteration interval is determined. Based on the fourth initial iteration interval, the process of calculating the physical parameters of the first transformer is repeated to determine the physical parameters of the fourth transformer.

[0066] Specifically, firstly, the second parameter range obtained from the optimization of passive components in circuit modeling is used as the target performance constraint. Based on the second parameter range, the first initial iteration interval of the transformer's physical parameters is determined, i.e., the selectable range of parameters such as line width, number of turns, spacing, and inner diameter. This interval is set as the starting point for the physical parameter search. At the beginning of each iteration, the current updated weight is calculated based on an initial weight and the updated weight obtained from the previous iteration. The updated weight is used to adjust the search direction and accuracy of the next step. Further, using the two endpoints of the current iteration interval (the first interval endpoint and the second interval endpoint), the corresponding first function value and second function value are calculated respectively. The function value usually represents the error between the target performance index and the target performance index. Based on the magnitude of the two function values, different division point calculation methods are selected to calculate a new candidate solution point, i.e., the first division point. Using the first division point, the first initial iteration interval is divided into two sub-intervals, and the merits of the function values ​​at both ends are compared to determine the target sub-interval. The target sub-interval is then used as the new iteration interval, and the above steps are repeated in conjunction with weight updates until the set maximum number of iterations is reached. The physical parameters obtained at this point are the optimal physical parameters of the first transformer.

[0067] Furthermore, based on the optimization results of the physical parameters of the first transformer, and combined with the second parameter range, a second initial iteration interval is set, and the same optimization process is performed on the second transformer to obtain the physical parameters of the second transformer; similarly, the optimization process for the third and fourth transformers is continued, and finally four sets of optimized transformer physical parameters are obtained in sequence to meet the performance coverage requirements of the entire broadband LNA.

[0068] The updated weights can be calculated based on the following formula:

[0069] ;

[0070] Among them, the The updated weights; The initial weights; The weights from the previous iteration; The value is a function of the weights from the previous iteration; It is a constant; the stated This is the target value.

[0071] For example, in one embodiment, the transformer is composed of two octagonal helical coils with a turns ratio of m:n, where m:n is determined empirically. The physical parameters of the transformer include the inner radii rL and rR of the coils and the center-to-center distance d, which are the target variables to be optimized. Before optimization, based on the second parameter range obtained in circuit modeling, i.e., the numerical range of the target inductance L and coupling coefficient k obtained in the passive component circuit simulation stage, an initial search range for rL, rR, and d is set. This search range corresponds to the "first initial iteration range". At the beginning of each iteration, according to the set initial weights... and the weights in the previous iteration result The current iterative weights are updated using an exponential decay function. This is used to dynamically adjust the position of the bisection point in the next round. This weight update corresponds to the "calculation of updated weights" step. Next, the two endpoints of the first initial iteration interval are selected, namely the endpoints of the first interval (e.g., the smaller combination of rL, rR, d) and the endpoints of the second interval (the larger combination), and their first and second function values ​​are calculated by substituting them into the performance error function. Based on the relationship between the function values ​​of the two endpoints, the current first bisection point is calculated using the bisection point calculation method. The original iteration interval is divided into two sub-intervals. Then, based on the quality of the function values, the target sub-interval that is retained, i.e., the side with smaller error, is determined as the search range for the next iteration. This process is repeated until the preset maximum number of iterations is reached. At this point, rL, rR, and d are the optimal physical parameters of the first transformer.

[0072] Based on this, using the current transformer optimization results as a reference and combining the second parameter range corresponding to the second transformer in the circuit, a new second initial iteration interval is set, and the above optimization process is repeated to obtain the optimal physical parameters of the second transformer; then the optimization of the third and fourth transformers is performed in the same way. Furthermore, to address the gain peak shift problem caused by self-resonance during the layout process, this scheme further introduces fine-tuning of the radius ratio rratio (while maintaining the scaling rL, rR), and fine-tunes d to ensure that the gain peak position in the simulation results after layout is consistent with that before layout. The entire optimization process is executed in two rounds to ensure the layout consistency and performance accuracy of the entire broadband LNA system.

[0073] The method provided in this embodiment, firstly, utilizes an adaptive weighting mechanism that iteratively optimizes parameters. This mechanism accelerates convergence when errors are large and maintains stability when approaching the optimal solution, effectively balancing the exploration and refinement stages. This characteristic makes the adaptive weighted bisection algorithm more advantageous than traditional bisection or fixed-step-size methods when facing complex nonlinear objective functions. In practical applications, this weighting mechanism can quickly narrow the search range in the early stages and maintain stability in the later stages, avoiding the jitter or missing of the global optimum that may occur with traditional methods. Therefore, the adaptive weighted bisection algorithm can significantly improve the convergence speed and efficiency of optimization, and is particularly suitable for design tasks with high precision requirements, such as the optimization of transformer physical parameters for broadband low-noise amplifiers. Furthermore, the adaptive mechanism balances exploration and stability when adjusting weights. At the beginning of the algorithm, larger weights cause the algorithm to focus more on the side with smaller errors, accelerating convergence. As the number of iterations increases, the weights gradually return to 0.5, ensuring that the search direction remains sufficiently flexible when approaching the optimal solution. This mechanism effectively avoids the algorithm getting trapped in local optima and maintains global search capability. This characteristic is particularly important in complex circuit optimization tasks, especially when multiple design constraints need to be considered simultaneously. By gradually adjusting the weights, the adaptive weighted binary search algorithm can dynamically balance accuracy and efficiency, thereby achieving a better global optimum. Secondly, by using the range of the second parameter as the starting point for physical parameter optimization, the adaptive weighted binary search algorithm closely links the optimization of the target inductance and coupling coefficient with the physical structural parameters of the transformer. The optimization process is not merely a blind adjustment of the transformer's physical parameters, but rather a reverse calculation based on the target electrical parameters, ensuring the practical feasibility of the optimization results in terms of circuit performance. This design approach starting from electrical parameters avoids unrealistic or unsuitable combinations of physical parameters, greatly improving the effectiveness and accuracy of optimization, and is particularly suitable for system designs such as RF circuits that are highly sensitive to inductance and coupling coefficients. Thirdly, after optimization by the adaptive weighted binary search algorithm, due to parasitic effects in the layout (such as self-resonance), even if the inductance and coupling coefficients match the simulated values, the position of the gain peak may shift. Therefore, the adaptive weighted binary search algorithm can further adjust the transformer's geometric parameters by optimizing the radius ratio and center spacing to ensure that the gain peak after layout is precisely aligned with the gain peak in the simulation. This subsequent fine-tuning process ensures a high degree of consistency between layout and simulation performance, improving the accuracy and reliability of the final design. This two-stage optimization process maximizes design flexibility and accuracy while ensuring performance consistency, meeting more stringent design requirements.

[0074] Optionally, the method for calculating the segmentation point based on matching the magnitudes of the first and second function values ​​includes the following steps: when the first function value is greater than the second function value, calculating the fourth product of the weight of the previous iteration and the endpoint of the first interval; calculating the third difference between 1 and the weight of the previous iteration; calculating the fifth product of the third difference and the endpoint of the second interval; and determining the sum of the fourth and fifth products as the first segmentation point; when the first function value is not greater than the second function value, calculating the sixth product of the weight of the previous iteration and the endpoint of the second interval; calculating the seventh product of the third difference and the endpoint of the first interval; and determining the sum of the sixth and seventh products as the first segmentation point.

[0075] Specifically, the process of finding the first dividing point can be expressed by the following formula:

[0076] ;

[0077] Among them, the The first function value; The second function value; The target value; The weights from the previous iteration; The endpoints of the first interval; The endpoint of the second interval; This is the first dividing point.

[0078] The method provided in this embodiment employs dynamic weight adjustment and selects different calculation strategies based on the magnitude of the error, enabling the algorithm to adaptively adjust the search direction according to the current optimization state. During the iteration process, different strategies are adopted to determine the calculation method of the split point based on the relationship between the first function value and the second function value. This allows for dynamic adjustment based on the characteristics of the error when facing different optimization stages, ensuring that the optimization process can be effectively carried out under different conditions. Secondly, through multiple weighted and product calculations, the weight changes and error adjustments during the search process can be meticulously balanced. Each iteration not only depends on the position of the interval endpoints but also considers the impact of weight changes on the split point, making the search process more accurate. This method not only ensures that the selection of each split point is more in line with the target requirements but also avoids making undesirable split decisions when the error is too large, thereby improving accuracy and stability. Furthermore, by calculating the error difference and combining it with the product of the interval endpoints, the asymmetry of the error is considered during the optimization process. In practical optimization, errors may not be symmetrically distributed. This discrepancy can cause traditional split-point calculation methods to fail or accumulate errors. This method effectively addresses this asymmetric error by adjusting the split points, ensuring its robustness even in complex optimization tasks. Furthermore, this calculation method improves the efficiency of the optimization process. In traditional bisection methods, the selection of split points typically relies on simple interval partitioning. However, by combining weights and error differences, the algorithm can precisely adjust the search direction at each step, thereby accelerating the convergence of the search process. This means that the algorithm can approach the optimal solution faster before reaching the set maximum number of iterations, reducing computational resource consumption and improving the efficiency of the entire optimization process.

[0079] S105. Based on the first parameter, the second parameter, and the physical parameter, the active and passive components are interconnected and assembled according to the layout rules to obtain a broadband low-noise amplifier layout.

[0080] In specific implementation, the process of interconnecting and assembling the active and passive components based on the first parameter, the second parameter, and the physical parameters, according to layout rules, to obtain a broadband low-noise amplifier layout includes: determining the layout requirements of the resistor and capacitor arrays through simulation analysis based on the first parameter, the second parameter, and the physical parameters; determining routing instructions based on the extracted layout information and the layout requirements; determining the placement positions of the active and passive components according to the routing instructions and layout rules for component alignment, non-overlap, and minimizing the overall occupied area; using pin information to guide the A* algorithm, and using Chebyshev distance as a heuristic function, interconnecting and routing according to the routing instructions; during the routing process, identifying traces that violate design rule check constraints as obstacles, and adjusting the routing based on the check results; and assembling the active and passive components with completed layout and routing to obtain the broadband low-noise amplifier layout.

[0081] Specifically, firstly, based on the first parameters of active components, the second parameters of passive components, and the physical parameters of each device obtained from previous optimization, simulation analysis is performed to extract the layout requirements of the corresponding resistor and capacitor arrays, including the required area, arrangement, pin orientation, and electrical connection requirements. Then, based on the extracted layout information, combined with the inter-array connection relationships and coupling requirements, routing instructions are generated as the target input for subsequent routing. Further, according to the routing instructions, active and passive components are arranged in positions while satisfying the following layout rules: component alignment rule: uniform orientation for improved routing feasibility; non-overlapping rule: all component physical outlines must not overlap; minimum area rule: the overall layout area is minimized to optimize chip area utilization. Through these layout rule constraints, candidate placement areas for each component are calculated, generating the optimal placement scheme. After the component locations are determined, the routing process is guided by the A* (A-Star) algorithm: pin information is used as the start and end points of the path; Chebyshev distance is used as the heuristic function, taking into account diagonal connections; path planning is performed sequentially according to the routing instructions; during the routing process, design rule checks (DRC) are dynamically performed, such as minimum trace width, minimum spacing, and via restrictions; if the routing path violates the rules, the corresponding path is marked as an obstacle area, and the local routing search is re-executed to achieve automatic adjustment and repair. When all routing tasks are completed and the electrical connections between all active and passive components meet the design constraints, the entire assembly is assembled into the final broadband low-noise amplifier layout. Figure 3 This is a schematic diagram of the layout of the broadband low-noise amplifier provided in this application.

[0082] The method provided in this embodiment, firstly, by constructing a lookup table, can quickly find a suitable broadband low-noise amplifier topology based on design requirements. Compared to the traditional method of manually selecting topologies, this avoids a large amount of screening and trial-and-error processes, greatly shortening the time consumed in the initial design phase and reducing the computational burden. Furthermore, in the optimization of the first parameter of active components, basic parameters are determined based on the topology, key performance index weights are allocated according to the application scenario, a comprehensive performance evaluation function is constructed, and the optimal parameter combination is found through iterative calculation. The optimization of the second parameter of passive components adopts the particle swarm optimization algorithm, decomposing the complex parameter optimization problem into a step-by-step calculation process, reducing computational complexity and convergence difficulty, and improving the efficiency of parameter optimization. Secondly, the optimization of the first parameter of active components is closely related to the topology, ensuring good matching with other parts of the circuit (especially passive components). Simultaneously, the determination of the range of the second parameter of passive components is related to the first parameter of active components, ensuring the coordinated operation of active and passive components in the circuit, reducing performance degradation caused by parameter mismatch, and improving circuit reliability. By employing an adaptive weighted binary search algorithm, the physical parameters of the transformer are optimized with the second parameter range of passive components as a reference. This closely links the optimization of the target inductance and coupling coefficient with the physical structural parameters of the transformer, avoiding combinations of physical parameters that do not meet actual manufacturing and circuit performance requirements, thus improving the effectiveness and accuracy of the optimization. Thirdly, this embodiment optimizes the parameters of active and passive components, as well as the physical parameters of the transformer, in stages, comprehensively considering key performance indicators such as circuit gain, noise figure, and bandwidth. When optimizing passive component parameters, specific optimization objectives are set for different magnetically coupled resonators, such as maximizing gain at a specific frequency and minimizing gain difference, effectively controlling gain fluctuations and achieving globally optimal performance. Fourthly, in the interconnect wiring assembly stage, this embodiment strictly adheres to layout rules of component alignment, non-overlap, and minimizing overall area. Component alignment improves wiring feasibility, the non-overlap rule avoids physical conflicts between components, and the area minimization rule optimizes chip area utilization, reducing parasitic effects, signal interference, and performance losses caused by improper layout. The A* algorithm is used for interconnect routing, guided by pin information and using Chebyshev distance as a heuristic function, combined with Design Rule Check (DRC). During routing, once a violation of the rules is detected, it is immediately marked as an obstacle area and the routing path is searched again, achieving automatic adjustment and repair, thus ensuring the electrical performance and electromagnetic compatibility of the circuit.

[0083] Corresponding to the aforementioned embodiment of a broadband low-noise amplifier end-to-end design method, this application also provides an embodiment of a broadband low-noise amplifier end-to-end design apparatus.

[0084] Figure 4 This is a schematic diagram of the end-to-end design device for a broadband low-noise amplifier provided in Embodiment 2 of this application. Please refer to... Figure 4 The device provided in this embodiment includes a matching module 210, an optimization module 220, and an assembly module 230;

[0085] The matching module 210 is used to match the topology of the broadband low-noise amplifier based on a lookup table.

[0086] The optimization module 220 is used to optimize the first parameters of the active components in the broadband low-noise amplifier based on the topology.

[0087] The matching module 210 is further configured to match the second parameter range of the passive components in the broadband low-noise amplifier based on the first parameter of the active component.

[0088] The optimization module 220 is also used to optimize the physical parameters of the transformer using an adaptive weighted binary search algorithm, with the second parameter range of the passive component as the pre-layout reference value.

[0089] The assembly module 230 is used to interconnect and assemble the active and passive components based on the first parameter, the second parameter, and the physical parameters, according to the layout rules, to obtain a broadband low-noise amplifier layout.

[0090] The apparatus of this embodiment can be used to perform... Figure 1 The steps of the method embodiment shown are similar in principle and process, and will not be repeated here.

[0091] The specific implementation process of the functions and roles of each unit in the above device can be found in the implementation process of the corresponding steps in the above method, and will not be repeated here.

[0092] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to in the description of the method embodiments. The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this application according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0093] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A broadband low-noise amplifier end-to-end design method, characterized in that, The method includes: The topology of a broadband low-noise amplifier is matched based on a lookup table. Optimize the first parameter of the active element in the broadband low-noise amplifier based on the topology; The first parameter of the active element is matched to the second parameter range of the passive element in the broadband low-noise amplifier. The process involves determining the input and output ports based on the first parameters of the active components, extracting the real and imaginary parts from the input and output ports, and modeling them as equivalent resistance-capacitance networks. Based on these equivalent resistance-capacitance networks, initial values ​​for the second parameters are calculated. A particle swarm optimization algorithm is used to optimize the second parameters related to the magnetically coupled resonators at each stage. During optimization, iterative calculations are performed based on the particle dynamics formula to match the frequency corresponding to the peak gain of the magnetically coupled resonators at each stage and the overall gain difference, calculating the second parameters corresponding to those frequencies. Using the particle swarm optimization algorithm, based on the equivalent resistance-capacitance network and the optimized circuit state of the magnetically coupled resonators at each stage, iterative calculations are performed based on the particle dynamics formula to obtain the remaining second parameters. Capacitors greater than a first preset value are combined into a capacitor array, and resistors less than a second preset value are combined into a resistor array. All the optimized second parameters are then organized to determine the range of the second parameters for the passive components in the broadband low-noise amplifier. The iterative calculation based on the particle dynamic change formula includes: determining the local optimal solution of the target second parameter at the target iteration number, and calculating the first difference between the local optimal solution and the target second parameter; determining the global optimal solution of the target second parameter at the target iteration number, and calculating the second difference between the global optimal solution and the target second parameter; calculating the first product of the first acceleration coefficient and the first difference, and calculating the second product of the second acceleration coefficient and the second difference; determining the parameter adjustment step size of the target second parameter at the target iteration number, and calculating the third product of the inertia weight and the parameter adjustment step size, and determining the sum of the first product, the second product, and the third product as the parameter adjustment step size of the target second parameter at the next target iteration number; and determining the sum of the parameter adjustment step size of the target second parameter at the next target iteration number and the target second parameter at the target iteration number as the target second parameter for the next target iteration number. Using the second parameter range of the passive component as a pre-layout reference value, the physical parameters of the transformer are optimized using an adaptive weighted binary search algorithm; Based on the first parameter, the second parameter, and the physical parameters, the active and passive components are interconnected and assembled according to the layout rules to obtain a broadband low-noise amplifier layout.

2. The method according to claim 1, characterized in that, The optimization of the first parameter of the active element in the broadband low-noise amplifier based on the topology includes: The basic parameters of the active components are determined based on the aforementioned topology. Based on the application scenarios of broadband low-noise amplifiers, the importance of different key performance indicators is analyzed, and the weights of different key performance indicators are determined based on the importance. Using the transconductance-drain current ratio method, and based on the aforementioned fundamental parameters, a comprehensive performance evaluation function is constructed according to the weights of different key performance indicators. The ratio of transconductance to drain current is adjusted by iterative calculation. In each iteration, the comprehensive performance score under different combinations of first parameters is calculated according to the comprehensive performance evaluation function until the number of convergences reaches a preset number. Then, the first parameter combination with the best comprehensive performance score is selected as the first parameter of the active element.

3. The method according to claim 1, characterized in that, The particle swarm optimization algorithm is used to optimize the second parameters related to the magnetically coupled resonator at each stage. During the optimization process, iterative calculations are performed based on the particle dynamics formula to match the frequency corresponding to the peak gain of the magnetically coupled resonator at each stage and the overall gain difference. The second parameter corresponding to the frequency is then calculated, including: Determine the second parameters related to the first magnetically coupled resonator, determine the first gain peak value of the first magnetically coupled resonator, and determine the first frequency based on the first gain peak value; Starting from the initial value, the second parameter of the first magnetically coupled resonator is optimized using the particle swarm optimization algorithm. During the optimization process, iterative calculations are performed based on the dynamic change formula of the particle swarm to calculate the second parameter that reaches the first gain peak at the first frequency. Determine the second parameters related to the second magnetically coupled resonator, determine the second gain peak value of the second magnetically coupled resonator, and determine the second frequency and the third frequency based on the second gain peak value; Starting from the initial value and combining the optimized circuit state of the first magnetically coupled resonator, calculate the second parameter that reaches the second gain peak at the second and third frequencies. Determine the second parameters related to the third magnetically coupled resonator, determine the second parameters related to the fourth magnetically coupled resonator, and calculate the overall gain difference between the third and fourth magnetically coupled resonators between the second and third frequencies. Starting from the initial value, and combining the optimized circuit states of the first and second magnetically coupled resonators, the second parameter is calculated when the overall gain difference reaches its minimum value.

4. The method according to claim 1, characterized in that, Using the second parameter range of the passive component as a pre-layout reference value, the physical parameters of the transformer are optimized using an adaptive weighted binary search algorithm, including: The first initial iteration interval for the physical parameters of the transformer to be optimized is determined based on the range of the second parameter. At the beginning of each iteration, the updated weights are calculated based on the initial weights and the weights from the previous iteration. Based on the first initial iteration interval, determine the endpoints of the first interval and the second interval, calculate the first function value and the second function value of the first interval endpoint and the second interval endpoint respectively, match the corresponding segmentation point calculation method based on the magnitude of the first function value and the second function value, and calculate the first segmentation point based on the segmentation point calculation method. The first initial iteration interval is divided according to the first dividing point to obtain two sub-intervals. The target sub-interval is determined based on the magnitude of the first function value and the second function value. The iterative calculation process is repeated based on the target sub-interval and the updated weight until the number of iterations reaches the maximum number of iterations, and then the physical parameters of the first transformer are determined. Based on the physical parameters of the first transformer and the range of the second parameter, a second initial iteration interval is determined, and the process of calculating the physical parameters of the first transformer is repeated based on the second initial iteration interval to determine the physical parameters of the second transformer. Based on the physical parameters of the second transformer and the range of the second parameters, a third initial iteration interval is determined. Based on the third initial iteration interval, the process of calculating the physical parameters of the first transformer is repeated to determine the physical parameters of the third transformer. Based on the physical parameters of the third transformer and the range of the second parameter, a fourth initial iteration interval is determined. Based on the fourth initial iteration interval, the process of calculating the physical parameters of the first transformer is repeated to determine the physical parameters of the fourth transformer.

5. The method according to claim 4, characterized in that, The method for calculating the segmentation point based on matching the magnitudes of the first and second function values, wherein the first segmentation point is calculated based on the method for calculating the segmentation point, includes: When the first function value is greater than the second function value, calculate the fourth product of the weight of the previous iteration and the endpoint of the first interval; Calculate the third difference between the weight of 1 and the weight of the previous iteration, calculate the fifth product of the third difference and the endpoint of the second interval, and determine the sum of the fourth product and the fifth product as the first dividing point; When the first function value is not greater than the second function value, calculate the sixth product of the weight of the previous iteration and the endpoint of the second interval; Calculate the seventh product of the third difference and the endpoint of the first interval, and determine the sum of the sixth product and the seventh product as the first dividing point.

6. The method according to claim 1, characterized in that, The step of interconnecting and assembling the active and passive components based on the first parameter, the second parameter, and the physical parameters, according to layout rules, to obtain a broadband low-noise amplifier layout includes: Based on the first parameter, the second parameter, and the physical parameters, the layout requirements of the resistor and capacitor array are determined through simulation analysis. Based on the extracted layout information and the layout requirements, the wiring instructions are determined. Based on the wiring instructions and the layout rules of component alignment, non-overlap, and minimizing the overall footprint, determine the placement positions of active and passive components. The A-Star algorithm is guided by pin information and Chebyshev distance is used as a heuristic function. Interconnection routing is performed according to the routing instructions. During the routing process, traces that violate the design rule check constraints are identified as obstacles, and routing adjustments are made based on the check results. The active and passive components with completed layout and wiring are assembled to obtain the layout of a broadband low-noise amplifier.

7. The method according to claim 1, characterized in that, Prior to the topology matching of the broadband low-noise amplifier based on the lookup table, the method further includes: Determine the key performance indicators of the broadband low-noise amplifier based on the design objectives and requirements. Collect the topologies of broadband low-noise amplifiers and record the key performance indicators corresponding to each topology; Construct a database containing the aforementioned topology and key performance indicators, and create a lookup table based on the database.

8. A broadband low-noise amplifier end-to-end design device, characterized in that, The device includes a matching module, an optimization module, and an assembly module; The matching module is used to match the topology of the broadband low-noise amplifier based on a lookup table. The optimization module is used to optimize the first parameters of the active components in the broadband low-noise amplifier based on the topology. The matching module is also used to match the second parameter range of the passive components in the broadband low-noise amplifier based on the first parameter of the active component. The process involves determining the input and output ports based on the first parameters of the active components, extracting the real and imaginary parts from the input and output ports, and modeling them as equivalent resistance-capacitance networks. Based on these equivalent resistance-capacitance networks, initial values ​​for the second parameters are calculated. A particle swarm optimization algorithm is used to optimize the second parameters related to the magnetically coupled resonators at each stage. During optimization, iterative calculations are performed based on the particle dynamics formula to match the frequency corresponding to the peak gain of the magnetically coupled resonators at each stage and the overall gain difference, calculating the second parameters corresponding to those frequencies. Using the particle swarm optimization algorithm, based on the equivalent resistance-capacitance network and the optimized circuit state of the magnetically coupled resonators at each stage, iterative calculations are performed based on the particle dynamics formula to obtain the remaining second parameters. Capacitors greater than a first preset value are combined into a capacitor array, and resistors less than a second preset value are combined into a resistor array. All the optimized second parameters are then organized to determine the range of the second parameters for the passive components in the broadband low-noise amplifier. The iterative calculation based on the particle dynamic change formula includes: determining the local optimal solution of the target second parameter at the target iteration number, and calculating the first difference between the local optimal solution and the target second parameter; determining the global optimal solution of the target second parameter at the target iteration number, and calculating the second difference between the global optimal solution and the target second parameter; calculating the first product of the first acceleration coefficient and the first difference, and calculating the second product of the second acceleration coefficient and the second difference; determining the parameter adjustment step size of the target second parameter at the target iteration number, and calculating the third product of the inertia weight and the parameter adjustment step size, and determining the sum of the first product, the second product, and the third product as the parameter adjustment step size of the target second parameter at the next target iteration number; and determining the sum of the parameter adjustment step size of the target second parameter at the next target iteration number and the target second parameter at the target iteration number as the target second parameter for the next target iteration number. The optimization module is also used to optimize the physical parameters of the transformer using an adaptive weighted binary search algorithm, with the second parameter range of the passive component as the pre-layout reference value. The assembly module is used to interconnect and assemble the active and passive components based on the first parameter, the second parameter, and the physical parameters, according to layout rules, to obtain a broadband low-noise amplifier layout.

Images