Apparatus and method

Abstract

A method for simulating a RADAR system defined by a corresponding RADAR model comprising a transmission chain, a reception chain, and a data processing chain, which is implemented by a computer comprising a processor and memory. Transmitting radio frequency (RF) signals to a target and optionally to an environment including clutter via a transmission chain (2702), Calculating the reflected RF signals from the target and from optional clutter (2704), The receiving chain receives each reflected RF signal (2706), The data processing chain uses each received reflected RF signal to detect the target (2708), A method comprising (2710) determining a simulated output of a RADAR system using the results of detecting a target.

Description

[Technical Field]

[0001] This invention relates to a RADAR system. In particular, this invention relates to the simulation of a RADAR system and / or the verification of a simulated RADAR system. [Background technology]

[0002] A conventional RADAR system includes, for example, a transmitter (i.e., a transmitting chain) configured to emit radio frequency, RF signals (i.e., radio waves also known as radar signals) into an environment including a target (and optionally, clutter), and a receiver (i.e., a receiving chain) configured to receive the respective reflected RF signals from the target (and / or optionally from clutter). In this way, the target can be detected using the respective reflected RF signals received by, for example, a data processor (i.e., a data processing chain).

[0003] There is still a need to improve the development of RADAR systems. [Overview of the project]

[0004] A first aspect is a method for simulating a RADAR system defined by a corresponding RADAR model comprising a transmission chain, a reception chain, and a data processing chain, which is implemented by a computer comprising a processor and memory. The transmission chain transmits radio frequency (RF) signals to a target and optionally to an environment including clutter. Calculating the reflected RF signals from the target and from optional clutter, The receiving chain receives each reflected RF signal, The data processing chain uses each received reflected RF signal to detect the target, Using the results of target detection to determine the simulated output of the RADAR system and This provides a method for providing this.

[0005] A second aspect is a method for providing a RADAR system, which is at least partially implemented by a computer having a processor and memory, To simulate a RADAR model that includes a transmission chain, a reception chain, and a data processing chain, corresponding to a RADAR system. The results of simulating the RADAR model are used to give a RADAR system. This provides a method for providing this.

[0006] A third aspect provides a RADAR system provided by the method according to the second aspect.

[0007] The fourth aspect provides a computer comprising a processor and memory configured to implement the method according to either the first aspect and / or the second aspect, a computer program comprising instructions that, when executed by the computer comprising the processor and memory, cause the computer to perform the method according to either the first aspect and / or the second aspect, or a non-temporary computer-readable storage medium comprising instructions that, when executed by the computer comprising the processor and memory, cause the computer to perform the method according to either the first aspect and / or the second aspect.

[0008] A fifth aspect is a method for verifying a RADAR model comprising a transmission chain, a reception chain, and a data processing chain corresponding to a RADAR system, which is implemented by a computer comprising a processor and memory. To acquire output collected from the RADAR system regarding the target and, optionally, the environment including clutter, Creating a RADAR model compatible with the RADAR system, By simulating the RADAR model, the simulated output of the RADAR model for the environment, including the target and optionally clutter, is determined. The RADAR model is validated by comparing the simulated output with the collected output. This provides a method for providing this.

[0009] A sixth aspect is a method for providing a RADAR system, which is at least partially implemented by a computer having a processor and memory. To verify a RADAR model that includes a transmission chain, a reception chain, and a data processing chain compatible with a RADAR system, The results of validating the RADAR model are used to give the RADAR system This provides a method for providing this.

[0010] The seventh aspect provides a RADAR system provided by a method according to either the fifth aspect and / or the sixth aspect.

[0011] The eighth aspect provides a computer comprising a processor and memory configured to implement the method according to either the fifth aspect and / or the sixth aspect; a computer program comprising instructions that, when executed by the computer comprising the processor and memory, cause the computer to perform the method according to either the fifth aspect and / or the sixth aspect; or a non-temporary computer-readable storage medium comprising instructions that, when executed by the computer comprising the processor and memory, cause the computer to perform the method according to either the fifth aspect and / or the sixth aspect. Detailed description of the invention A first aspect is a method for simulating a RADAR system defined by a corresponding RADAR model comprising a transmission chain, a reception chain, and a data processing chain, which is implemented by a computer comprising a processor and memory. Transmitting, via a transmission chain, a radio frequency (RF) signal into an environment including a target and optionally clutter. Calculating, from the target and optionally from the clutter, each reflected RF signal of the transmitted RF signal. Receiving, via a reception chain, each reflected RF signal. Detecting a target using, via a data processing chain, each received reflected RF signal. Determining a simulated output of a RADAR system using a result of detecting the target. Providing a method comprising the above.

[0012] In one example, transmitting, via a transmission chain, an RF signal into an environment including a target and optionally clutter comprises generating a continuous RF signal having a given amplitude envelope and optionally converting the continuous RF signal into a pulsed, amplitude-modulated RF signal and optionally amplifying the pulsed, amplitude-modulated RF signal.

[0013] In one example, the RADAR model comprises a transmission phase-locked loop, PLL, and generating a continuous RF signal having a given amplitude envelope comprises generating the continuous RF signal using the PLL.

[0014] In one example, the RADAR model comprises an envelope generator and generating a continuous RF signal having a given amplitude envelope comprises applying the amplitude envelope using the envelope generator.

[0015] In one example, the RADAR model comprises a vector modulator and converting the continuous RF signal into a pulsed, amplitude-modulated RF signal comprises converting the continuous RF signal into a pulsed, amplitude-modulated RF signal using the vector modulator.

[0016] In one example, a RADAR model includes an amplifier to amplify a pulsed and amplitude-modulated RF signal, or an amplifier to amplify a pulsed and amplitude-modulated RF signal.

[0017] For example, a RADAR model includes a transmitter that transmits an RF signal to a target and optionally to an environment including clutter.

[0018] In one example, the target comprises and / or a moving target or a stationary target, and optionally, herein, the clutter comprises and / or a stationary clutter.

[0019] In one example, calculating the reflected RF signals from a transmitted RF signal and from an optional clutter comprises the following:

[0020] In one example, receiving each reflected RF signal by a receiving chain includes receiving each reflected RF signal of a transmitted RF signal from a target and an optional clutter, optionally adding thermal noise to each received reflected RF signal, optionally amplifying each received reflected RF signal, optionally downconverting each received reflected RF signal to, for example, a first intermediate frequency (IF1) signal, optionally filtering the IF1 signal, and optionally downconverting the IF1 signal to a second intermediate frequency (IF2) signal.

[0021] In one example, detecting a target using each received reflected RF signal by a detection chain involves processing each received reflected RF signal by a data processing chain, which includes one or more of the following: performing a Hilbert transform on each received reflected RF signal, matching and filtering on each received reflected RF signal, windowing on each received reflected RF signal, performing a Fast Fourier Transform (FFT) on each received reflected RF signal, and generating a range-Doppler plot.

[0022] In one example, this method involves adapting a RADAR model, for example, a transmission chain, a reception chain, and a data processing chain.

[0023] For example, updating a RADAR model involves updating the RADAR model using the simulated output of the RADAR system.

[0024] In one example, this method includes modifying target and / or optional clutter. Here, calculating the reflected RF signals of the transmitted RF signals from the target and from the optional clutter comprises calculating the reflected RF signals of the transmitted RF signals from the modified target and from the optional modified clutter.

[0025] In one example, the transmitting block includes a quantum oscillator.

[0026] A microwave generator unit (MGU) is a comprehensive device capable of generating signals in the microwave domain by phase-coherent splitting of high-fidelity optical signals using frequency comb technology. The MGU includes a frequency comb for downconverting the optical signal and an ultrastable optical system that acts as a reference for the frequency comb. The frequency comb can also be referenced by an external optical signal. The frequency comb is a pulsed mode-locked laser that generates uniformly spaced frequency outputs in the frequency domain. Mode locking is a technique for generating pulsed lasers on the order of picoseconds or femtoseconds. The uniformly spaced frequency outputs (known as "teeth") are the repetition rate frequency (f) of the pulsed laser. rep The teeth are separated by ) . The frequency of each tooth is equivalent to the number of teeth (number of modes) multiplied by the spacing between the teeth. The phase change between laser pulses is due to the carrier-envelope offset frequency (f ceo We introduce another parameter called ).

[0027] MGUs can generate frequencies in both the optical and RF domains. RF signals generated by MGUs have low phase noise characteristics due to the conversion of optical signals to RF. MGUs can also perform optical atomic clock locking and convert optical atomic clocks to ultra-high precision and stability in the RF domain. MGUs are used in applications including deep space navigation, precision measurement, telecommunications, and next-generation wireless communications. The low phase noise RF output makes MGUs a potential candidate for realizing quantum oscillators in radar. While an MGU-generated RF source locked to an optical atomic clock is the ultimate realization of a quantum oscillator for radar applications, an RF signal generated from an MGU locked to its cavity-stabilized internal laser is referred to in the paper as one manifestation of a quantum oscillator.

[0028] Atomic clocks can be further divided into two categories based on frequency: microwave and optical. Optical atomic clocks are oscillators with the highest level of frequency stability. There are two main categories of optical clocks: the first uses one or more trapped ions, and the second uses a number of neutral atoms in a lattice of light with a “magic” wavelength. Optical atomic clocks mainly consist of a local oscillator and an atomic reference. The local oscillator provides the optical frequency that will be governed by the atomic reference. Carefully generated samples of atoms or single ions generally have some “forbidden” electronic transitions that act as the atomic reference. Electronic transitions that are forbidden under the electric dipole approximation are known as forbidden transitions. Forbidden transitions can be enabled by applying an external magnetic field. These forbidden transitions have an ultranarrow linewidth that makes them a useful, well-defined frequency reference. By tuning the frequency of the local oscillator (laser) to the optical atomic resonant frequency, the optical clock references the atomic transitions. The highest frequency stability of optical atomic clocks is due to the narrow linewidth of the transitions. The RF output of an MGU locked to an optical atomic clock is another potential candidate for realizing a quantum oscillator in radar.

[0029] A second aspect is a method for providing a RADAR system, which is at least partially implemented by a computer having a processor and memory, To simulate a RADAR model that includes a transmission chain, a reception chain, and a data processing chain, corresponding to a RADAR system. The results of simulating the RADAR model are used to give a RADAR system. This provides a method for providing this.

[0030] The method according to the second embodiment may include any of the steps described with respect to the first embodiment.

[0031] In one example, this method includes updating the RADAR model based on the results of validating the RADAR model.

[0032] In one example, this method includes verifying an updated RADAR model.

[0033] A third aspect provides a RADAR system provided by the method according to the second aspect.

[0034] The fourth aspect provides a computer comprising a processor and memory configured to implement the method according to either the first aspect and / or the second aspect, a computer program comprising instructions that, when executed by the computer comprising the processor and memory, cause the computer to perform the method according to either the first aspect and / or the second aspect, or a non-temporary computer-readable storage medium comprising instructions that, when executed by the computer comprising the processor and memory, cause the computer to perform the method according to either the first aspect and / or the second aspect.

[0035] A fifth aspect is a method for verifying a RADAR model comprising a transmission chain, a reception chain, and a data processing chain corresponding to a RADAR system, which is implemented by a computer comprising a processor and memory. To acquire output collected from the RADAR system regarding the target and, optionally, the environment including clutter, Creating a RADAR model compatible with the RADAR system, By simulating the RADAR model, the simulated output of the RADAR model for the environment, including the target and optionally clutter, is determined. The RADAR model is validated by comparing the simulated output with the collected output. This provides a method for providing this.

[0036] The method according to the fifth embodiment may include any of the steps described with respect to the first embodiment and / or the second embodiment.

[0037] In one example, comparing the simulated output with the collected output involves comparing the signal power and / or signal-to-noise ratio, SNR, values ​​of the transmit chain, receive chain, and / or data processing chain to their expected values.

[0038] For example, comparing the signal power and / or signal-to-noise ratio (SNR) values ​​of the transmit chain, receive chain, and / or data processing chain to their respective expected values ​​comprises comparing the signal power and / or signal-to-noise ratio (SNR) values ​​of the outputs of the transmit chain, receive chain, and / or data processing chain to their respective expected values.

[0039] For example, comparing a simulated output with a collected output involves comparing the basic parameters of the simulated output with the basic parameters of the collected output.

[0040] For example, comparing the basic parameters of a simulated output with the basic parameters of a collected output involves comparing the basic parameters of a generated range-Doppler plot of the simulated output with the basic parameters of a range-Doppler plot of the collected output.

[0041] For example, comparing a simulated output with a collected output involves comparing the clutter-to-noise ratio, CNR, signal-to-noise ratio, SNR, and / or thermal noise floor between the simulated output and the collected output.

[0042] In one example, comparing the CNR, SNR, and / or thermal noise floor of a simulated output with that of a collected output involves comparing the CNR, SNR, and / or thermal noise floor of a generated range-Doppler plot of the simulated output with that of a range-Doppler plot of the collected output.

[0043] In one example, CNR takes thermal noise into account.

[0044] In one example, this method is Adapting a RADAR model, for example, a transmission chain, a reception chain, and a data processing chain, Optionally, modify the target and / or optional clutter. Equipped with, Here, calculating the reflected RF signals of the transmitted RF signals from the target and from the optional clutter comprises calculating the reflected RF signals of the transmitted RF signals from the modified target and from the optional modified clutter.

[0045] In one example, the transmitting block includes a quantum oscillator.

[0046] A sixth aspect is a method for providing a RADAR system, which is at least partially implemented by a computer having a processor and memory. To verify a RADAR model that includes a transmission chain, a reception chain, and a data processing chain, which are compatible with a RADAR system. The results of validating the RADAR model are used to give the RADAR system This provides a method for providing this.

[0047] The method according to the sixth embodiment may include any of the steps described with respect to the first embodiment, the second embodiment and / or the fifth embodiment.

[0048] In one example, this method involves adapting the RADAR model based on the results of validating the RADAR model.

[0049] In one example, this method includes verifying the adapted RADAR model.

[0050] The seventh aspect provides a RADAR system provided by a method according to either the fifth aspect and / or the sixth aspect.

[0051] The eighth aspect provides a computer comprising a processor and memory configured to implement the method according to either the fifth aspect and / or the sixth aspect; a computer program comprising instructions that, when executed by the computer comprising the processor and memory, cause the computer to perform the method according to either the fifth aspect and / or the sixth aspect; or a non-temporary computer-readable storage medium comprising instructions that, when executed by the computer comprising the processor and memory, cause the computer to perform the method according to either the fifth aspect and / or the sixth aspect.

[0052] definition Throughout this specification, the terms “comprising” or “comprises” mean including the specified components but not excluding the presence of other components. The terms “consisting essentially of” or “consists essentially of” mean including the specified components but excluding other components, except for materials present as impurities, unavoidable materials present as a result of processes used to give the components, and components added for purposes other than achieving the technical effects of the present invention, such as pigments.

[0053] The terms "consisting of" or "consists of" mean that a specified component is included but other components are excluded.

[0054] Where appropriate, and depending on the context, the use of the terms “comprises” or “comprising” may also be interpreted as including the meaning of “consists essentially of” or “consisting essentially of,” and also as including the meaning of “consists of” or “consisting of.”

[0055] The optional features presented herein may be used individually or in combination with each other, as appropriate, and in particular, in the combinations presented in the appended claims. The optional features for each aspect or exemplary embodiment of the Invention are also applicable, as appropriate, to all other aspects or exemplary embodiments of the Invention. In other words, those skilled in the art will, by reading this specification, consider the optional features for each aspect or exemplary embodiment of the Invention to be interchangeable and combinable between different aspects and exemplary embodiments.

[0056] For a better understanding of the present invention and to illustrate how exemplary embodiments of the present invention may be carried out, the accompanying figures will be referenced simply as examples. [Brief explanation of the drawing]

[0057] [Figure 1] A schematic diagram illustrating the front-end model of the entire radar model in a simulation, according to an exemplary embodiment. The front-end comprises four main sections. The transmit chain transmits an amplified and pulsed sinusoidal signal to the environment through the transmit antenna. Clutter and targets in the environment reflect the signal to the radar receive antenna. In the receive chain, the received signal undergoes amplification and down-conversion stages to generate an IF signal. The IF signal is converted to a baseband signal in the data processing chain. The baseband signal undergoes further data processing stages to extend the signal-to-noise ratio of the target. [Figure 2] A schematic diagram illustrating a typical PLL including a reference oscillator, a phase detector, a filter, a VCO, and a divider, according to an exemplary embodiment. The phase detector compares the output of the reference oscillator with the output of the VCO. The output of the phase detector is used to optimize the phase of the VCO until the VCO is locked with the reference oscillator. The divider is used to divide the VCO output frequency, and the filter is used to filter out noise. [Figure 3] A schematic diagram of a charge pump integer N PLL for generating a phase-locked 1 GHz signal using a 10 MHz reference signal. The VCO output is divided into 100 segments and compared to the reference oscillator in a PFD. The up (UP) and down (DOWN) signals at the PFD output drive the charge pump. The charge pump output is forwarded through a loop filter to the VCO control input. The PLL is finally locked when both the reference signal and the divided signals are synchronized with each other. [Figure 4] (a) A figure showing the time-domain representation of a 10 MHz pulse signal at the input of the PLL and (b) a 1 GHz phase-locked VCO signal at the output of the PLL. [Figure 5] This figure shows the power spectrum of a 1 GHz phase-locked VCO signal at the output of a PLL. [Figure 6] A diagram showing the time-domain representation of a signal at the output of different blocks in the transmit chain. The representation shows only a partial section (10s) from the complete PRI. (a) Output of the envelope generator showing the amplitude envelope for approximately 1s, (b) Output of the vector modulator showing the continuous sinusoidal signal within the amplitude envelope. The transmit frequency in the L band makes it difficult to resolve the sinusoidal signal in the representation. The particular shape of the vector modulator output is an artifact of the envelope generator's sampling rate. [Figure 7]This figure shows the power spectra of the signals at the outputs of the amplifier and transmitting antenna. The power spectrum at the amplifier output peaks at 7.9 dB, equivalent to a signal power of 33 dB. The power spectrum at the transmitting antenna output peaks at 20.4 dB, equivalent to a signal power of 45.5 dB. The transmitting antenna gain is equivalent to 12.5 dB. The side lobes in the power spectrum are artifacts of the envelope generator's sampling rate. [Figure 8] A diagram showing the time-domain representation of a signal at antenna output, exhibiting a 4s time delay equivalent to a distance of 600m. The representation shows only a partial section (10s) from the complete PRI. (a) Signal at the output of the transmitting antenna, (b) Signal at the output of the receiving antenna. [Figure 9] This figure shows a histogram of thermal noise at the output of the thermal noise block during the simulation. The histogram shows a Gaussian noise distribution peaking at -107 dB for bandwidths of 5 GHz and 290 K, and temperature, respectively. [Figure 10] This figure shows the power spectra of the signals at the outputs of the receiving antenna and the thermal noise block. The power spectrum at the output of the receiving antenna peaks at -99.6 dB, which is equivalent to a received signal power of -74.5 dB. The power spectrum at the output of the thermal noise block shows a uniform thermal noise floor at -137 dB. Since the spectrum has an RBW of 5 MHz, -137 dB is equivalent to an expected thermal noise power of -107 dB. [Figure 11] This figure shows the power spectra at the outputs of the thermal noise block and the LNA. The power spectrum at the output of the LNA is amplified to -35.1 dB, equivalent to a peak power of -10 dB. The noise floor is amplified to -68 dB, equivalent to a noise power of -38 dB. The SNR at the output of the LNA is 4.5 dB lower than the SNR at the output of the thermal noise block (equivalent to the LNA NF). [Figure 12]Figure showing histograms of the output of a thermal noise block and an LNA, representing peak signal and noise power values. (a) Histogram of the output of the thermal noise block, showing signal power and noise power at -74.5 dB and -107 dB, respectively. (b) Histogram of the output of the LNA, showing signal power and noise power at -10 dB and -38 dB, respectively. [Figure 13] A schematic diagram of a typical mixer including RF and LO inputs and outputs. The output includes two frequencies: the sum of the two input frequencies and the difference between the two input frequencies. [Figure 14] This figure shows the power spectra of the signals at the outputs of the LNA and mixer. The signal peak at the mixer output is reduced by 6 dB compared to the peak at the LNA output. The noise power at the mixer output is reduced by 3 dB compared to the noise power at the LNA output. The reduction in signal and noise power values ​​is as expected. [Figure 15] Figures showing histograms at the outputs of a mixer and a band-pass filter (BPF) indicating peak signal and noise power values. (a) Histogram at the mixer output showing signal and noise power at -10dB and -41dB, respectively. (b) Histogram at the output of the BPF showing signal and noise power at -16dB and -61dB, respectively. [Figure 16] This diagram schematically shows a typical ADC including a sample block, a stop block, a quantization block, and an encoder block. The input to the ADC is sampled at the required frequency in the sample block and stored in the stop block until the next sample. The signal is quantized using the quantization block. The digital signal at the output of the quantization block is converted to binary format in the binary block. [Figure 17] This figure shows histograms of the ADC output, representing the signal power and noise power at -16dB and -61dB, respectively. [Figure 18]A schematic diagram showing the data matrix and data cube in the data processing chain of a radar system. (a) A data matrix containing a single PRI on the fast time axis and a stack of PRIs on the slow time axis. (b) A data cube containing data matrices from each receiving channel stacked on top of each other. [Figure 19] (a) A diagram showing the time-domain representation of the signal at the input of a matched filter and (b) the signal at the output of a matched filter. [Figure 20] The figure shows histograms at the output of a matched filter, representing the signal power and noise power at 79 dB and 24.5 dB, respectively. [Figure 21] Figure showing simulated range-Doppler plots of a fixed test target at 600m, including 15 distance bins and a central Doppler bin equivalent to 1400Hz. (a) Range-Doppler plot without Blackman-Harris windowing, showing target peak power of 145dB and average noise power of 56.5dB, equivalent to an SNR of 88.5dB. (b) Simulated range-Doppler plot of a fixed test target at 600m, including 15 distance bins and a central Doppler bin equivalent to 1400Hz. [Figure 22] A figure showing a simulated range-Doppler plot with distance bins on the vertical axis and Doppler frequency on the horizontal axis. The spectrum is shown for -700 to +700 Hz. A single fixed clutter and a single Doppler target are highlighted in the green and red boxes, respectively. The range-Doppler plot is normalized to the clutter. [Figure 23]Figure showing a comparison between simulated and real-world range-Doppler plots. Both range-Doppler plots are normalized to the strongest clutter. Targets are highlighted in red boxes. Targets are present across both plots within the same distance and Doppler bins. (a) Simulated range-Doppler plot without oscillator phase noise. The plot includes a fixed clutter and a single simulated target within a uniform thermal noise floor. (b) Real-world range-Doppler plot from a steering radar trial at an airfield. A phase noise floor is seen emerging from the thermal noise floor for the distance bin with higher clutter. The plot also includes other unwanted targets seen by the radar. [Figure 24] The figure shows a comparison graph illustrating clutter power and thermal noise floor for each distance bin. The comparison graph includes data from both a real-world range-Doppler plot and a simulated range-Doppler plot. The clutter power for both the simulation and real data overlaps only within a good range. The thermal noise floor in the simulation completely overlaps with the real data and is therefore not differentiable. [Figure 25(a)] A figure showing a simulated range-Doppler plot of a classical oscillator with phase noise. [Figure 25(b)] The figure shows a real-world range-Doppler plot from a steering radar trial with phase noise from a classical oscillator. The range-Doppler plot also consists of other unwanted targets observed by the radar. [Figure 25(c)] This figure shows a comparison between a simulated range-Doppler plot with phase noise for a classical oscillator and a real-world range-Doppler plot. A phase noise floor can be seen emerging from the thermal noise floor for distance bins with higher clutter power. The target of interest is highlighted in the red box. [Figure 26]The figure shows a comparison graph illustrating the peak clutter power per distance bin, the overall noise floor, and the thermal noise floor for both simulated range-Doppler plots with phase noise for a classical oscillator and real-world range-Doppler plots. The overall noise floors for both the simulation and real data overlap by only a good range. The overall noise floor for distance bins close to the radar increases by approximately 25 dB above the thermal noise floor. The thermal noise floor in the simulation completely overlaps with the real data and is therefore not differentiable. [Figure 27] A diagram illustrating a method in a schematic manner according to an exemplary embodiment. [Figure 28] A diagram illustrating a method in a schematic manner according to an exemplary embodiment. [Modes for carrying out the invention]

[0058] There are many radar architectures in the world today, and the radar architecture depends on the application of the radar. Some common radar architectures include frequency-modulated continuous-wave radar, pulsed-Doppler radar, and pulse-compressed radar. Any common radar includes a transmit chain, a receive chain, and a data processing chain. A good radar model is one that can simulate the building blocks in the transmit chain, receive chain, and data processing chain, along with environmental target and clutter models. The radar model developed and described herein is based on the L-band steering pulsed-Doppler radar at the University of Birmingham with a reasonable number of simplifications and focuses on modeling the behavior of the hardware building blocks.

[0059] There are several platforms on which radar can be modeled, including MATLAB® and LabVIEW. The radar model described herein was developed from scratch using a bottom-up approach in MATLAB and MATLAB's graphical interface known as Simulink. Most of the building blocks in the radar were modeled in Simulink to develop a power-efficient, user-friendly radar model with all parameters optimized based on requirements. The front end of the entire simulated radar model is given in Figure 1. The radar model can be divided into four main sections as follows:

[0060] • Transmission chain ·environment • Receiving chain • Data processing chain The components within the transmit chain, environment, and receive chain were modeled in Simulink as separate system-level building blocks. Each radar building block contained further subblocks to appropriately represent the functionality of the radar components in the model. The Simulink blocks representing the radar components were connected to perform simulations against the complete radar model. Test probes were placed to assess signals at the inputs and outputs of every building block during the simulation, in both the time and frequency domains. Data at the output of the last block in the receive chain was imported into the MATLAB workspace. The data processing chain was implemented in MATLAB using separate MATLAB code.

[0061] The front-end transmit chain of the radar model included a transmit phase-locked loop (PLL) for generating the transmit RF signal and an envelope generator for providing the amplitude envelope. The continuous RF signal generated at the output of the transmit PLL was converted into a pulsed and amplitude-modulated RF signal with the help of a vector modulator. The RF signal was then amplified using an amplifier and transmitted to the environment using a transmit antenna.

[0062] The environment included a target and clutter model to reflect the RF signal transmitted to the receiving antenna. In the receiving chain, the signal reflected from the environment was collected by the receiving antenna. A thermal noise block after the receiving antenna added Gaussian thermal noise to the simulation. An LNA amplified the received signal at the output of the thermal noise block. A combination of a receiving PLL block and a mixer block down-converted the received RF signal to a first intermediate frequency (IF1) signal. The IF1 signal was filtered using a BPF to remove all unwanted frequencies. An ADC placed at the output of the BPF down-converted the IF1 signal to a second intermediate frequency (IF2). The ADC was used only for frequency down-conversion, as the signal was already in the digital domain.

[0063] Several data processing techniques were performed on the signal at the ADC output during the data processing chain. The data processing chain included Hilbert transform and matched filtering, windowing, FFT, and generation of range-Doppler plots. Matched filtering was used to improve the SNR in the presence of added noise. In the radar domain, matched filtering is often referred to as pulse compression. The FFT was used to generate range-Doppler plots from the time-domain signal and to perform pulse integration to improve the SNR. Spectral leakage due to the Fourier transform was reduced through windowing.

[0064] 2 Transmission Chain The transmit chain of the simulated full radar model included components that perform signal generation, signal modulation, signal amplification, and signal transmission. One significant aspect of the radar transmit chain is the generation of the RF transmit frequency. As described herein, the transmit frequency is either directly generated or synthesized using frequency upconversion processes. The earliest method for directly synthesizing RF signals used a magnetron. More recently, direct digital synthesis (DDS) has been used to generate signals directly at the required RF frequency. A typical method for synthesizing RF signals using frequency upconversion includes a mixer. Another method for frequency upconversion uses a PLL. The RF signal was generated using a transmit PLL in the full radar model. The Simulink simulations described herein also include a general standalone PLL model to verify the working principle of the PLL. In the full radar model, due to modeling limitations when integrating the standalone PLL model, a simplified version of the transmit PLL for generating signals directly at RF frequencies was modeled.

[0065] Most radar signals undergo some form of waveform modulation. Waveform modulation can be of different types, including amplitude modulation, frequency modulation, and phase modulation, as described herein. The entire radar model included amplitude modulation to generate pulsed radar signals. The amplitude envelope for amplitude modulation in the entire radar model was generated by an envelope generator block, as shown in Figure 1. Amplitude modulation is generally added to the RF signal with the help of a mixer. Mixers are also used in radar for frequency up-conversion and frequency down-conversion. In the radar model, a vector modulator performs the function of a mixer. The vector modulator integrated amplitude modulation into the RF signal. In most radars, the generated RF signal needs to be amplified to high power before transmission to the environment. The power required for transmission and amplification depends on the radar application. The entire radar model simulated a simple amplifier design with constant amplifier gain. The transmitting antenna design can follow different architectures depending on the application. The transmitting antenna was modeled as a single transmitting antenna with constant antenna gain in the entire radar model.

[0066] 2.1 Transmit PLL The transmit PLL generates sinusoidal RF signals in the transmit chain of the radar model implemented during the simulation. This subsection describes the fundamentals of PLLs, how PLLs operate in simulations, and the design of transmit PLLs in simulations.

[0067] 2.1.1 Fundamentals of PLL A PLL is an electronic circuit that uses a voltage or voltage-driven oscillator to continuously adjust its frequency to match an input signal. PLLs generate, stabilize, modulate, demodulate, filter, or restore signals in "noisy" communication channels. PLLs play a crucial role in the most modern electronic devices, particularly RF devices. In radar, PLLs are used to generate a stable RF signal at the radar transmission frequency using a highly stable reference oscillator.

[0068] A typical PLL block diagram is given in Figure 2. The main components in a PLL include a reference oscillator, a voltage-controlled oscillator (VCO), a phase detector, a filter, and, optionally, a divider. The PLL operates on the principle of tuning the phase and frequency of the VCO with the help of the reference oscillator. A control input in the VCO tunes the VCO frequency, and the phase detector compares the phase of the VCO to that of the reference oscillator. When the phase of the VCO is synchronized with that of the reference oscillator, the state is known as phase-locked and is therefore known as a phase-locked loop. The VCO can also be locked to the reference oscillator at a different reference frequency with the help of a divider placed before the phase detector, as shown in Figure 2. The divider can also be placed after the reference oscillator when a fractional relationship is required between the reference oscillator and the VCO frequency.

[0069] There are various types of phase detectors. A mixer phase detector works on the principle of mixing two input signals. The mixer performs multiplication of the two inputs, and one of the outputs of the mixer phase detector will be the sine of the phase difference between the two inputs. A mixer phase detector is only effective for the same frequency or frequencies that are very close to each other, as it only detects the phase difference between the two inputs.

[0070] Phase-frequency detectors (PFDs) are the most common type of phase detector. PFDs can operate even when the frequencies of the two input signals to the phase detector are different. PFDs operate on the principle of zero transitions of the input signals to track the difference between phase and frequency. The output of a PFD includes up (UP) and down (DOWN) signals. The up (UP) and down (DOWN) signals indicate the direction of the frequency change required to lock the PLL. Since the voltage signal controls the VCO, the up (UP) and down (DOWN) signals at the PFD output need to be converted to voltage signals. One standard method of converting up (UP) and down (DOWN) signals is using a charge pump. The current source in the charge pump can be controlled by the up (UP) and down (DOWN) outputs from the PFD. The charge pump current is drawn into the output when the up (UP) signal is active, and drawn out when the down (DOWN) signal is active. A capacitor at the output of the charge pump transfers the current fluctuations to the voltage fluctuations. Voltage fluctuations are supplied to the VCO's control input.

[0071] Filters in a PLL are primarily used to control the loop dynamics and are therefore also known as loop filters. Filters are also used to low-pass filter noise and unwanted products at the output of the phase detector. The filter is placed before the VCO, and its output is fed to the VCO's control input.

[0072] 2.1.2 PLL operation in simulation A typical example of the operation of a PLL with integer N charge pumps to generate an RF signal at 1 GHz using a 10 MHz stable reference oscillator is given in Figure 3. In the PLL simulation, the 1 GHz output of the VCO was downconverted to 10 MHz with the help of a divider with N equal to 100. The reference signal and the divider signal were compared in a PFD, which generated an error signal depending on the phase difference between the two input signals. The error signal drove the charge pump, and the output of the charge pump was applied to the control input of the VCO. When both the reference signal and the divided signal are phase-locked, the 1 GHz PLL output signal is finally locked to the 10 MHz reference signal. The pulsed 10 MHz reference signal at the PLL input and the 1 GHz phase-locked sinusoidal signal at the PLL output are given in Figure 4. The spectrum of the phase-locked VCO signal at the PLL output given in Figure 5 shows a peak at 1 GHz, confirming the operation of the PLL on the simulation platform.

[0073] 2.1.3 Transmit PLL in Simulation An ideal case scenario would be to integrate a fully functional PLL model, as described in Section 2.1.2, within the entire radar model simulation. Design and timing considerations made it impossible to integrate a fully functional PLL into the radar model. As an alternative, the transmit PLL simulated in the model was modeled in Simulink as a direct digital synthesizer for directly generating signals at RF frequencies, based on the RF block set. The transmit PLL also has the option of adding phase noise to the generated signal. Phase noise is critical to radar characterization and performance, and the effects of phase noise in radar systems are described in detail herein. The transmit PLL can be assigned amplitude and frequency to arbitrary user-specified values. The block also has the option of assigning amplitude to both the in-phase and quadrature components of the signal. In addition to generating a signal with a specific amplitude, the signal can also be specified in terms of current and power. The output frequency of the transmit PLL was set to the transmit frequency of an L-band steering radar. The amplitude of the generated signal was set to 1V. A simplified equation for the mathematical model of the signal at the output of the transmit PLL is given by:

[0074]

number

[0075] Here, S PLL (t) represents the sinusoidal signal at the output of the transmitting PLL, and A PLL and θ represent the amplitude and phase of a sinusoidal signal, respectively, and f T This represents the transmission frequency.

[0076] 2.2 Envelope Generators and Vector Modulators Signal modulation is one of the important aspects of a radar transmission chain. Amplitude modulation and frequency modulation are the two most common types of signal modulation used in radar. Amplitude modulation gives an amplitude envelope over the generated RF signal, and the envelope can be of different shapes depending on the radar requirements. The most common type of amplitude modulation is the rectangular envelope. In linear frequency modulation, the transmission signal frequency is modulated in a linear manner. Linear frequency modulation is used to achieve both high distance resolution and a large detection range simultaneously using the increased bandwidth of the transmission signal. An envelope generator is one way of generating the modulation required for a radar transmission signal. A vector modulator is essentially a mixer for adding baseband modulation to an RF signal. The architecture of a vector modulator is similar to that of a general mixer.

[0077] 2.2.1 Envelope Generators in Simulations In all radar models, an envelope generator was used to provide the required amplitude envelope for the signal generated by the transmitting PLL. The envelope generator can generate any user-defined amplitude envelope. The envelope generator may also include further subblocks to enable phase modulation and frequency modulation. In the simulation, the envelope generator produced a shaped envelope with a pulse width of approximately 1 s and a PRI of approximately 136 s. The envelope generator was implemented as a block that could provide amplitude values ​​at a user-specified sampling rate. The amplitude values ​​were loaded into MATLAB as a .mat file and read by Simulink at the specified sampling rate. The amplitude envelope at the output of the envelope generator is given in Figure 6(a).

[0078] 2.2.2 Vector Modulators in Simulation In all radar models, the vector modulator mixes a continuous sinusoidal signal at the output of a transmit PLL, which has an amplitude envelope at the output of an envelope generator. The vector modulator was simulated using an RF blockset mixer. The mathematical model of the signal at the output of the vector modulator is given by the following equation:

[0079]

number

[0080] Here, S V (t) represents the signal at the output of the vector modulator, and A V ,f T θ, and θ represent the amplitude, frequency, and phase of the signal, respectively. Even if the signal at the output of the vector modulator were a pulsed signal, for simplicity, the time dependence of the pulse is not explicitly mentioned in Equation 2. Thus, Equation 2 is a simplified mathematical signal model. The pulsed RF signal at the output of the vector modulator is shown in Figure 6(b). Figure 6(b) clearly shows the amplitude envelope over the sinusoidal RF signal. The particular shape of the vector modulator output is an artifact of the sampling rate of the envelope generator.

[0081] 2.3 Amplifier An RF amplifier is an electronic amplifier used in most communication systems to amplify an input signal to the required output signal. In radar, an RF amplifier is used to convert a low-power RF signal generated by a transmitter into a high-power RF transmit signal. Generally, RF amplifiers are placed before the radar transmit antenna and drive the transmit antenna. Design parameters for an RF amplifier include gain, power output, bandwidth, input and output impedance matching, and linearity.

[0082] The gain of an RF amplifier is defined as the ratio of the output power to the input signal power. Bandwidth defines the operating signal bandwidth of the amplifier, and ideally, different amplifiers would be required depending on the signal frequency. Impedance is defined as the opposite measure of the flow of electricity. Impedance is a complex-valued quantity with resistance as its real part and reactance as its imaginary part. Impedance matching equalizes the input impedance and output impedance to maximize signal power transfer by minimizing signal reflection. Linearity is described as the behavior of a circuit, where fluctuations in output signal strength are directly proportional to fluctuations in input signal strength. The ratio of output signal power to input remains the same for a linear amplifier, regardless of the input signal power.

[0083] 2.3.1 Amplifiers in Simulation In all radar models, a simplified signal amplifier was used to amplify the input signal to an arbitrary user-defined value. The amplifier in the simulation has the option to perform both linear and nonlinear amplification. The amplifier in the simulation was set to linear amplification. The amplification factor in dB can also be given to the amplifier. From the radar's power budget, the expected value of the signal power at the amplifier output is 33 dB. The power spectrum outputs the RMS value of the signal. For an ideal rectangular pulse with a width of 1 s and a PRI of 136 s, the RMS peak would be reduced by 21.3 dB, resulting in a power spectrum that peaks at 11.7 dB (instead of 33 dB). Since the amplitude envelope in the simulation is not a perfect rectangle, the peak in the spectrum is reduced by a further 3.8 dB, resulting in a total reduction of 25.1 dB. The power spectrum of the signal at the amplifier output is given in Figure 7. In Figure 7, it can be seen that the peak power of the signal at the amplifier output is 7.9 dB, which is equivalent to a transmit power of 33 dB.

[0084] 2.4 Transmitting Antenna Antennas are widely used in all wireless systems. Antenna behavior depends on the components within the antenna, and antenna design depends on the application. In radar, transmitting antennas are used to broadcast RF transmission signals to the environment. Key parameters of an antenna are radiation pattern, antenna gain, and directivity.

[0085] The radiation pattern of an antenna is a description of the angular dependence of its emission, which is a result of its three-dimensional shape. In a three-dimensional coordinate system, the horizontal plane (XY) is defined as the azimuthal plane, and the vertical plane (YZ) is the upright plane. The azimuthal angle is defined as the horizontal axis angle from the antenna boresight direction. The elevation angle is defined as the angle above the horizon in the vertical direction. Based on the radiation pattern, antennas can be divided into isotropic antennas and directional antennas. An isotropic antenna is a theoretical antenna pattern in which the radiated power is equal in all directions of azimuthal and elevation. In a directional antenna, the energy is dominated in one or more directions. Antenna gain is defined as the ratio of the power transmitted in the main direction of a directional antenna to the power in an isotropic antenna. The main direction of a directional antenna is called the radar boresight. The directivity of an antenna is defined as the maximum signal power radiated in a given direction, divided by the average power in all directions.

[0086] 2.4.1 Transmitting Antenna in Simulation The transmitting antenna was designed as a simplified antenna with a constant antenna gain across all radar models. Simulink has an antenna design feature that allows for the inclusion of parameters such as directivity. While simulating the simplified antenna model, the constraint of total simulation time for the entire radar simulation was taken into consideration. Further modifications may be incorporated into the antenna design depending on the application. The antenna gain in the simulation was achieved using an amplifier as described in Section 2.3.1. The simplified mathematical model of the transmitted signal is given by the following equation:

[0087]

Number

[0088] Here, S T (t) represents the transmitted signal, and A T represents the amplitude of the transmitted signal. The power spectrum of the signal at the output of the transmitting antenna is given in FIG. 7. The transmitting antenna is expected to amplify the signal by 12.5 dB in transmission. The value of the antenna gain was selected based on the system parameters of the L-band steering radar. From FIG. 7, it can be seen that the signal at the output of the transmitting antenna reaches a peak at 20.4 dB and is equivalent to a signal power of 45.5 at the output of the transmitting antenna. The 12.5 dB transmitting antenna gain is visible in FIG. 7. In the simulation, the output of the transmitting antenna was connected to a block in the environment.

[0089] 3 Environment The radar environment includes both targets and clutter. A target is an object of interest in an environment that may be stationary or moving. All moving targets have a Doppler frequency associated with the target, and the Doppler frequency is used to measure the speed of the target. Radar targets include drones, aircraft, ships, vehicles, and birds. Clutter is generally fixed objects and can include those of buildings, trees, land, ocean, and other geographical features in the field of view. Weather is also classified as clutter. The electromagnetic signal interacts with the atmosphere, clutter, and targets, and various types of interactions are described herein. The interaction of the electromagnetic signal with the environment is adapted by the radar range equation as described herein.

[0090] 3.1 Targets and Clutter in Simulation In the entire radar model, a simplified representation of the amplitude responses of both targets and clutter was simulated. Each of the subblocks in the environment shown in Figure 1 represents the amplitude response of a different object in the environment, which can be either a target or clutter. Targets and clutter are realized as point scatterers in the entire radar model. The output of the transmitting antenna block was connected to the input of the amplitude response block. Each amplitude response block was designed based on the radar distance equation. The distance, velocity, and RCS of each object can be defined within the amplitude response block. The environment can ideally represent any number of objects with varying distances, velocities, and RCS values ​​as point scatterers. Since antenna directivity is not included in the simulation, the position of an object corresponds to its distance at the radar boresight. The simulation may include additional algorithms for positioning objects at specific angles in azimuth and elevation units. The output of the amplitude response block was connected to the receiving chain through an interface. The interface combined the reflected signals from all different objects in the environment and directed them to the receiving chain. Atmospheric loss is currently kept as 1 in the simulation. Both the target and the clutter are modeled as simple point scatterers, with the target modeled within a specific distance bin and the clutter dispersed within each distance bin.

[0091] 4 Receiving Chain The receiving chain of a radar model includes components for performing signal reception, signal amplification, signal down-conversion, and analog-to-digital conversion. The receiving antenna receives signals reflected from the environment. The receiving antenna also amplifies the signal. Generally, radar receivers contain thermal noise due to random thermal fluctuations of electrons, as described herein. In all radar models, a separate block was used to generate thermal noise in the receiver chain of the radar system. The thermal noise block was located after the receiving antenna, as shown in Figure 1.

[0092] One of the key blocks in the receiving chain of all radar models was the receiving PLL. The frequency of the signal at the output of the receiving PLL was different from the frequency of the signal at the output of the transmitting PLL. The frequency offset between the transmitting and receiving PLLs was equivalent to the IF. A mixer and bandpass filter (BPF) combination performed the down-conversion of the received signal to the IF. The output of the receiving PLL was mixed with the received signal. In all radar models, since both signals were digital signals, a signal multiplier was used to mix the received signals. The ADC is another key block in the receiving chain of the radar system. The ADC performs the basic operation of converting an analog signal to a digital signal. Whenever the sampling rate of the ADC does not conform to the Nyquist criterion, aliasing occurs, and the input signal will be represented as a down-converted signal at the output of the ADC. To reliably reproduce the signal according to the Nyquist criterion, the sampling frequency must be at least twice the highest frequency of the signal. The Nyquist frequency is defined as twice the highest frequency of interest. When frequencies above half the Nyquist frequency are sampled, the frequency is incorrectly detected as a lower frequency; this process is known as aliasing. In all radar models, the signal is already in the digital domain, and the ADC performs a second round of downconversion through aliasing.

[0093] 4.1 Receiving antenna and thermal noise Receiving antennas are present in most wireless communication devices. In radar, the receiving antenna is used to receive signals from the environment. Depending on the application, there are various types of radar antenna configurations, including a single receiving antenna or an array of receiving antenna elements. These designs apply to both transmitting and receiving antennas. An antenna array contains a collection of identical, uniformly spaced antenna elements that operate as a single unit. The signals received in each array element are combined with the phase relationships required to amplify the signal from the desired direction. Phased array antenna systems include phase shifters to manipulate the individual array elements in the desired direction.

[0094] Thermal noise is an unavoidable factor in any electronic device. Thermal noise defines a lower limit for target detection, and the range-Doppler plot of a radar is always limited by the thermal noise floor. As described herein, thermal noise depends on the bandwidth and temperature of the radar's receiving chain. Thermal noise can be reduced by reducing the bandwidth and temperature of the radar's receiving chain.

[0095] 4.1.1 Receiving antenna in simulation In the entire radar model, the receiving antenna was designed as a simplified antenna with a constant antenna gain. The L-band steering radar used for verification includes a two-dimensional array of 64 receiving antenna elements. Given the time constraints for simulating the entire radar model, the simulation currently has a single receiving antenna element. Adding more antenna elements would add more receiving chain components, and the model could be adapted to more antenna elements at the expense of simulation time. The output of the amplitude response block in the environment was connected to the input of the receiving antenna through an interface. The receiving antenna gain in the simulation was achieved using an amplifier similar to that described in Section 2.3.1.

[0096] Consider a radar that transmits a series of M pulses separated by PRI of T, where 0 ≤ m ≤ M-1. o For a target with the first distance and velocity v, the distance to the target for the mth transmitted pulse is R o -νmT. For the mth transmitted pulse, the signal arriving at the receiver antenna is:

[0097]

number

[0098] This will result in an equivalent time delay. The mathematical model for the delayed received signal is given by the following equation.

[0099]

number

[0100] Here,

[0101]

number

[0102] This is the Doppler frequency of the target moving towards the radar.

[0103]

number

[0104] This is the phase of the received signal. R This represents the amplitude of the received signal and follows the radar distance equation.

[0105] Consider a radar model with a single fixed test target at a distance of 600m and a specific RCS. The target at a distance of 600m and a specific RCS is used. For a target at 600m, the received signal will be delayed by 4s compared to the transmitted signal. The time-domain representations of the signal at the output of the transmitting antenna and the signal at the output of the receiving antenna are given in Figure 8. Figure 8 clearly shows a 4s delay between the transmitted and received signals, as expected. The radar distance equation gives the expected value of the received signal power.

[0106] For a test target at a given 600m and a specific RCS value, the signal power at the receiving antenna output is expected to be -74.4dB. The power spectrum of the signal at the receiving antenna output is given in Figure 10. Figure 10 shows the receiving antenna output, which peaks at -99.6dB. Since the peak in the power spectrum is reduced by 25.1dB by the RMS value, the received signal power is -74.5dB, which is very close to the expected value of -74.4dB.

[0107] 4.1.2 Thermal noise in simulations In the simulation, thermal noise was added using a separate block. Thermal noise depends on both bandwidth and temperature. The thermal noise block in the simulation can be given any value for noise temperature. The noise bandwidth in the simulation was equivalent to the simulation's sampling rate. The simulation's sampling rate was set to 5 GHz, which is greater than twice any frequency value used in the simulation and satisfies the Nyquist criterion. For a noise bandwidth of 5 GHz and a noise temperature equivalent to room temperature (290 K), the expected thermal noise power would be -107 dBW. Figure 9 shows a histogram of the thermal noise power values ​​generated by the thermal noise block. A Gaussian distribution of thermal noise power values, peaking at -107 dB, can be seen in the histogram of Figure 9.

[0108] SNR is an important parameter that defines radar performance. The SNR of a received signal is given by equation (5). The SNR of a received signal with an atmospheric loss coefficient of 1 is given by the following equation.

[0109]

number

[0110] In Figure 10, the power spectrum of the signal at the output of the thermal noise block is compared with the power spectrum of the signal at the output of the receiving antenna. In Figure 10, the thermal noise at all frequencies can be seen. The resolution bandwidth (RBW) of the power spectrum is 5 MHz, and the noise bandwidth is 5 GHz. In Figure 10, a thermal noise floor of -137 dB was produced, which is 30 dB lower than the expected thermal noise power of -107 dB.

[0111] 4.2 LNA LNAs (Low-Range Amplifiers) are essential components in most communication systems, particularly in radar receiving chains. An LNA is an electronic amplifier capable of amplifying weak signals without significant degradation of signal-to-noise ratio (SNR). In radar, LNAs are generally placed after the receiving antenna to amplify weak received signals and provide a power level suitable for analog-to-digital conversion or further processing in the analog domain. LNA performance can be measured using different parameters, including noise figure, gain, and dynamic range. LNA gain is defined as the amplification factor of the LNA. LNA gain is specified in dB. LNA noise figure (NF) is defined as a measure of the degradation of the SNR of the received signal. For example, if an LNA has a 3 dB SNR degradation at its output compared to its input, the LNA's NF will be 3 dB. Radar targets are designed to produce LNAs with the lowest possible NF to minimize SNR degradation by the LNA.

[0112] 4.2.1 LNA in Simulation In all radar models, the LNA was implemented using an amplifier with an option to specify the NF. The LNA design is based on the amplifier as described in Section 2.3.1. The LNA can provide amplification of any user-defined value. Along with the LNA gain, the LNA block can also specify an NF of any value. LNA(S LNA The mathematical model for the received signal at the output of ) is given by the following equation:

[0113]

number

[0114] Here, A LNA This represents the amplitude of the received signal after the LNA. The SNR (SNR) at the output of the LNA. LNA ) is given by the following equation,

[0115]

number

[0116] Here, NF is the noise figure of the LNA. In Figure 11, the power spectrum of the signal at the output of the LNA is compared to the signal at the output of the thermal noise block. It can be seen that the signal power and noise floor have been amplified by an amount equivalent to the LNA amplification factor. Figure 11 also shows the SNR at the outputs of the thermal noise block and the LNA, which are 37.4 dB and 32.9 dB, respectively. The SNR at the output of the thermal noise block is reduced by only 4.5 dB compared to the SNR at the output of the thermal noise block, and is equivalent to the LNA NF.

[0117] In Figure 12, the signal and noise power at the outputs of the thermal noise block and the LNA are also compared by plotting histograms. In Figure 12, broad peaks correspond to noise power, and smaller peaks towards the right edge of the histogram correspond to peak signal power. Figure 12(a) shows the signal and noise power after the thermal noise block, peaking at -74.5 dB and -107 dB, respectively. Figure 12(b) shows the signal and noise power after the LNA, peaking at -10 dB and -38 dB, respectively. Figure 12 also shows that the SNR at the output of the LNA is reduced by 4.5 dB compared to the output of the thermal noise block.

[0118] 4.3 Receiver PLL and Mixer One of the critical steps in any of the radar system's receiving chains is down-converting the received RF signal. Down-conversion to baseband can be either a single step or a multi-step process, as described herein. In most cases, the received RF signal is down-converted through multiple stages, each having a corresponding IF. Down-conversion to IF is generally performed by mixing the received signal with either the LO signal in the transmitting chain or a signal derived from the LO signal, thus maintaining coherence.

[0119] Mixers are generally used in a vast number of RF applications. An RF mixer is a three-port device used to upconvert or downconvert signal frequencies. Mixers can be either passive or active. An ideal mixer modifies the frequency of an input signal while maintaining the phase of the input signal. Mixers are used in radar receiver chains to downconvert RF signals to frequencies convenient for performing further processing steps. A mixer has two inputs: an RF input and an LO input. The output of a mixer contains two frequencies: the sum of the two input frequencies and the difference between the two input frequencies. A typical mixer is shown in Figure 13.

[0120] Two important parameters that affect the SNR of a mixer output are the mixer's conversion loss and noise figure (NF). The mixer's conversion loss is defined as the ratio between the input RF power and the output IF power. The mixer's NF is defined as the ratio between the input SNR and the output SNR. In an ideal mixer with an LO input amplitude of 1 and an RF input amplitude of A, each output amplitude will be equivalent to A / 2. Since power is proportional to the square of the amplitude, the power of each output signal of the mixer will be 1 / 4 of the input power. Therefore, the conversion loss of an ideal mixer will be 6 dB. MXRThis relates to the image frequency. Along with the desired frequency band, there is also an undesirable frequency band known as the image frequency for mixers that use LO to downconvert RF to IF. If noise is present in both the desired frequency band and the image frequency band, the downconversion folds the noise in the image band over the noise in the desired band, resulting in a 3dB NF.

[0121] 4.3.1 Receiver PLL in Simulation In all radar models, the receive PLL was designed similarly to the transmit PLL. The receive PLL was designed as a direct digital synthesizer to output any user-defined frequency. The frequency at the mixer output was equivalent to the sum of the transmit frequency and the IF. Coherence between the receive and transmit PLLs was achieved within the Simulink simulation. Both the transmit and receive PLLs were run simultaneously in the simulation, always on, and the phase relationship between the transmit and receive PLLs was made coherent.

[0122] 4.3.2 Mixers in Simulations In all radar models, the mixer's RF input was connected to the LNA output, and the mixer's LO input was connected to the receiver PLL output. The mixer output contained two frequencies. The difference frequency at the mixer output was equivalent to IF1. In the simulation, the mixer was implemented using a multiplication block to mix both input signals.

[0123] The amplitude at the output of the receiving PLL was kept at 1, and therefore the mixer conversion loss was obtained at 6 dB, following the example of an ideal mixer. The conversion loss affected both the signal and noise equally, and the signal power and noise power at the mixer output were 6 dB lower than at the LNA output. The effective SNR gain due to the conversion loss in the mixer was 0. In the simulation, the real-valued received signal was accompanied by complex-valued thermal noise. Therefore, the down-conversion folded the thermal noise present in the image band on top of the thermal noise present in the desired band, making the noise figure of the mixer equivalent to 3 dB. MXR It is given by the following equation:

[0124]

number

[0125] Here, NF MXR is the mixer's NF. In Figure 14, the power spectrum of the signal at the mixer output is compared to the signal at the LNA output. In Figure 14, two peaks for the power spectrum at the mixer output can be seen. The two peaks correspond to the sum frequency and the difference frequency (IF1 signal). As expected, the mixer output spectrum peaks in Figure 14 are reduced by 6 dB compared to the LNA output spectrum. Since the mixer's conversion loss is 6 dB and the mixer's NF is 3 dB, the effective noise power at the mixer output is reduced by 3 dB compared to the LNA output, as shown in Figure 14.

[0126] 4.4 BPF Filters are present in most communication systems. Filters are electronic components used to allow or block specific signal frequencies. Filters are crucial for improving the performance of communication systems by filtering out noise and reducing interference from external signals. Generally, filters can be divided into four categories: bandpass filters, bandreject filters, lowpass filters, and highpass filters. A BPF transmits a specific frequency band and blocks all other frequencies. A bandreject filter blocks a specific band of frequencies and transmits all other frequencies. A lowpass filter transmits all frequencies below a certain value, and a highpass filter transmits all frequencies above a certain value. In radar, a BPF filters out all out-of-band noise, thereby reducing noise present in the radar system's receiving chain. The BPF is located outside the mixer to transmit the difference frequency at the mixer output and block the sum frequency.

[0127] 4.4.1 BPF in Simulation In all radar models, the BPF is located outside the mixer to transmit the difference frequency centered on IF1. The signal at the output of the BPF (S BPF The mathematical model of ) is given by the following equation:

[0128]

number

[0129] Here,

[0130]

number

[0131] This is the amplitude of the signal after the BPF. The signal power at the output of the BPF will be equivalent to the signal power at the difference frequency. The noise power at the output of the BPF is B / B BPFIt is reduced by only this, where B is the noise bandwidth before the BPF, and B BPF This is the bandwidth of the BPF. The SNR at the output of the BPF is (SNR BPF ) is given by the following equation.

[0132]

number

[0133] In all radar models, the narrowband bandpass filter (BPF) was maintained at the mixer output. Histograms of the mixer and BPF outputs are given in Figure 15. In Figure 15(a), the signal power and noise power at the mixer output, peaking at -41dB and -10dB, respectively, can be seen. Even though the signal power at the mixer output is reduced by 6dB, since both frequency peaks exist before filtering, the signal peak at the mixer output in Figure 15(a) is the same as the signal peak at the LNA output, as shown in Figure 12(b). The noise power at the mixer output is reduced by 3dB compared to the LNA output, and is therefore expected to be -41dB. In Figure 15(b), the signal power and noise power at the BPF output, peaking at -61dB and -16dB, respectively, can be seen.

[0134] Since the BPF allows the difference frequency only at the IF, the signal power at the output of the BPF is expected to be -16 dB. The noise power at the output of the BPF will be reduced by a factor equivalent to the bandwidth ratio, as given by Equation 10. For the BPF bandwidth given in the simulation, the expected reduction in noise power after the BPF is 23 dB, which was also confirmed by the power budget calculation. Thermal noise includes both the desired frequency band and the image frequency band. The BPF operates over both the desired frequency band and the image frequency band, resulting in an additional 3 dB noise contribution from the image frequency band. Thus, the effective reduction in noise power after the BPF is 20 dB. In Figure 15, the noise power at the outputs of the mixer and the BPF are -41 dB and -61 dB, respectively, showing a 20 dB reduction in noise power after the BPF, as expected.

[0135] 4.5 ADC Analog-to-digital converters (ADCs) are crucial in digital data processing systems. An ADC is an electronic integrated circuit that converts signals from the analog to the digital domain. While signals in the analog domain are continuous signals with continuous values, signals in the digital domain are represented by a sequence of discrete values. In radar, the ADC is positioned before the data processing block. The ADC in radar facilitates the conversion of analog signals to digital signals, enabling digital data processing to generate target information. Key parameters in an ADC include the sampling rate, resolution, and quantization error.

[0136] 4.5.1 Fundamentals of ADCs A typical ADC block diagram is given in Figure 16. The sample block is used to sample the ADC input signal at a specific sampling frequency. The ADC input signal is continuous in both time and amplitude. The output of the sample block is a signal with discrete time and continuous amplitude. The stop block is used to stop the output of the sample block until the next batch of the sample block's output. The signal at the output of the stop block is discrete in time, continuous in amplitude, and remains unchanged until the next sample set. The quantization block is used for quantizing the signal. The continuous analog amplitude at the input is converted to a discrete amplitude at the output of the quantization block. Since the signal is discrete in both time and amplitude, the signal at the output of the quantization block is digital. The last block, the encoder block, converts the digital signal to binary format.

[0137] The sampling rate in an ADC is defined as the number of samples taken per second. As the sampling rate increases, the ADC can handle higher frequencies. The reconstruction of an analog signal in the digital domain depends on the sampling rate; the higher the sampling rate, the better the signal is reconstructed. If the ADC's sampling rate is lower than twice the frequency of the analog signal, aliasing occurs, and the reconstructed digital signal characteristics differ from those of the analog signal.

[0138] Digital signals have a discrete, finite number of values ​​and are represented using bits. Each bit corresponds to a specific number of steps in the digital signal. The resolution of an ADC is equivalent to the number of bits that represent the amplitude of the digital signal. The number of steps increases exponentially with the number of bits. As the bit resolution increases, the digital signal becomes closer to the original analog signal.

[0139] The digitized signal at the output of an ADC will always be different from its analog counterpart. Quantization error is the difference between the analog signal and the nearest accessible digital signal at each sampling moment. Quantization error introduces noise into the system known as quantization noise. Quantization noise is one of the significant noise components in the receiving chain of a radar system. As bit resolution increases, the difference between the digital and analog signals decreases, and both quantization error and quantization noise decrease. In an ideal ADC, quantization noise and bit resolution are defined by the signal-to-quantization noise ratio. The signal-to-quantization noise ratio for a number of bits equal to n is given by the following equation:

[0140]

number

[0141] 4.5.2 ADC in Simulation The ideal scenario was to add a comprehensive ADC simulation to the entire radar model. Since the signals in Simulink are already in the digital domain, the ADC in the entire radar model was used to perform frequency down-conversion through aliasing. The ADC was sampled at a specific ADC clock frequency, and the ADC clock frequency was below IF1 at the ADC input. Through frequency aliasing, the ADC output was down-converted to IF2. The ADC was implemented in the entire radar model using a signal-to-workspace block. The ADC clock frequency was achieved by sampling the signal-to-workspace block at the required sampling frequency. The ADC in the entire radar model may have additional sub-blocks to implement quantization noise. The mathematical model of the signal at the ADC output is given by the following equation:

[0142]

number

[0143] Here, A ADC This is the amplitude of the signal after the ADC and is equivalent to the amplitude of the signal before the ADC. The ADC in the simulation does not change the SNR, and therefore the SNR at the output of the ADC is given by the following equation.

[0144]

number

[0145] The histogram at the ADC output is given in Figure 17. Comparing Figure 17 with Figure 15(b), it can be seen that there is no relative difference in signal power or noise power due to the ADC. The signal-to-workspace block transported the Simulink data to the MATLAB workspace. All radar data processing stages were implemented in MATLAB.

[0146] 5. Data Processing Chain The data processing chain in a radar system is as important as other subsections within the radar. The radar's data processing chain performs both functions: improving the signal-to-noise ratio (SNR) and calculating target characteristics. Target characteristics primarily include the target distance and the target's Doppler velocity. The processing chain can distinguish between targets and clutter with respect to distance and velocity. One important aspect of the data processing chain is the data cube. The data cube is a three-dimensional matrix with three axes: number of received channels, slow time, and fast time. The fast time axis contains a single PRI. In the final stage of data processing, the fast time axis is further subdivided into different distance bins with specific distance resolutions. The slow time axis is updated for each PRI. Slow time corresponds to a single PRI, and adjacent PRI data are stacked along the slow time axis. Together, fast time and slow time represent the data matrix shown in Figure 18(a).

[0147] A data cube is an extension of a data matrix using multiple receiving channels. The fast time and slow time from each receiving channel are stacked to form a data matrix and then a data cube, as shown in Figure 18(b).

[0148] In the data processing chain, one way to improve the signal-to-noise ratio (SNR) is by performing beamforming. Beamforming is performed when multiple receiving antenna elements are present in a radar system. In radar, beamforming is a technique where the receiving antenna array can be focused across arbitrary angles of elevation and azimuth to improve the SNR. The overall radar model currently includes only one receiving antenna element, and therefore beamforming is not performed in the data processing chain of the radar simulation. The components in the data processing chain of a radar system depend on the radar's functionality. In a pulsed Doppler radar system, the data processing chain generally includes Hilbert transform, matched filtering, windowing, FFT, and range-Doppler plot generation. The Hilbert transform is the process of generating a complex-valued signal from a real-valued signal. Matched filtering involves the convolution of an unknown signal with a conjugate time-inverted reference signal to improve the SNR of the signal in the presence of random noise. In the overall radar model, the Hilbert transform and matched filtering blocks are used to perform both operations on the signal at the output of the ADC.

[0149] In radar, the Fast Fourier Transform (FFT) is performed to extract frequency-domain information from a time-domain signal. The Fourier transform is also used to improve the signal-to-noise ratio (SNR) through pulse integration. Pulse integration can follow coherent and non-coherent forms depending on the signal characteristics. The FFT is applied across the slow-time axis. In the entire radar model, windowing is performed before the FFT to reduce the leakage effect of frequency components on the Fourier transform. A range-Doppler plot is generated by performing the FFT across the slow-time axis of the data matrix. In the entire radar model, after windowing and the FFT, the generation of a range-Doppler plot is performed to give distance and Doppler information for both the target and the clutter.

[0150] 5.1 Hilbert Transform and Matched Filtering In data processing, the Hilbert transform is defined as a linear operator that, when applied to a real-valued function x(t), generates a function H[x(t)]. Mathematically, the Hilbert transform of x(t) is given by the following equation:

[0151]

number

[0152] Here, * indicates a convolution operation. The Hilbert transform helps generate an analytic signal. The analytic signal is a complex-valued function with no negative frequency components. The real and imaginary parts of the analytic signal are real-valued functions related to each other by the Hilbert transform. The analytic signal at the output of the Hilbert transform can be defined as follows:

[0153]

number

[0154] The real and imaginary parts of Equation 15 are known as the common-mode and quadrature values. In the frequency domain, the Hilbert transform imparts a 90° phase shift to the signal. In radar, the Hilbert transform is one of the methods used to downconvert a signal to the baseband. The Hilbert transform is generally performed through a four-phase detection, where the IF signal is downconverted to the baseband along with the generation of common-mode and four-phase baseband signals. The Hilbert transform and generation of the analyzed common-mode and quadrature signals are important in data processing and Doppler velocity calculations.

[0155] In communication systems, matched filters are used as linear filters to improve the signal-to-noise ratio (SNR) of a signal in the presence of additive white Gaussian noise. A matched filter is a time-domain linear filter based on the principle of correlation between a received signal and a reference signal to improve the SNR of a received signal. In an ideal matched filter, the maximum SNR is obtained when the reference signal is a time-delayed mirror image of the received signal.

[0156] 5.1.1 Hilbert Transform and Matched Filtering in Simulations In all radar models, the Hilbert transform and matched filtering blocks were used to bring IF2 down to the baseband and perform matched filtering. A simplified mathematical model of the baseband complex signal at the output of the Hilbert transform is given by the following equation:

[0157]

number

[0158] Here, A HTis the amplitude of the signal after the Hilbert transform. The complex baseband radar signal is used to measure the phase and amplitude independently. The matched filter for the entire radar model was derived from the transmitted pulse. MATLAB code for the matched filtering process was used to generate the matched filter output. The mathematical model of the signal at the output of the matched filter is given by the following equation:

[0159]

number

[0160] Here, A MF G is the amplitude of the signal at the output of the matched filter. An important parameter in matched filtering is the gain of the matched filter. The gain of the matched filter depends on the matched filter coefficients. MF In a matched filter that gives an SNR gain of , the SNR at the output of the matched filter is (SNR MF ) is given by the following equation.

[0161]

number

[0162] Figure 19 shows the time-domain representation of the signals at the input and output of the matched filter. A rectangular pulse is subjected to matched filtering to produce a triangular waveform. Since matched filtering performs correlation between the input signal and the reference signal, the matched-filtered output of the rectangular pulse will peak at the endpoint of the input signal. For any finite input signal, the duration of the matched-filter output is twice the duration of the input signal. Since the input to the matched filter is approximately a rectangular pulse, and the reference derived from the transmitted signal is also approximately a rectangular pulse, the output of the matched filter in Figure 19 is a triangular pulse as expected. The duration of the matched-filter output in Figure 19 is twice the duration of the input signal, and likewise, it peaks at the endpoint of the input signal. A histogram of the output of the matched filter is given in Figure 20. The signal and noise power at the output of the matched filter are 79 dB and 23.5 dB, respectively. Comparing Figure 20 with Figure 17, it can be seen that the SNR at the output of the matched filter has improved by 10.5 dB. The improvement achieved through matched filtering is very close to the expected value confirmed using the steering radar's power budget.

[0163] 5.2 Windowing, FFT, and Range-Doppler Plots Fourier analysis transforms a signal in its time or space domain into a representation in the frequency domain, and vice versa. The Fast Fourier Transform (FFT) is an algorithm that performs Fourier analysis, encompassing both the Discrete Fourier Transform (DFT) and the Inverse Discrete Fourier Transform (IDFT). The DFT of a signal is obtained by decomposing the signal into its frequency components. FFT is used in a wide range of applications in communication systems. In radar, FFT transforms time-domain signals into frequency-domain signals. FFT is one of the key data processing techniques used in radar to acquire Doppler information of a target.

[0164] The FFT is performed along the slow-time axis of the data matrix or data cube. Performing an FFT on the data matrix or data cube results in a range-Doppler plot where the slow-time axis is converted to the Doppler axis and the fast-time axis is converted to the distance axis. The Doppler axis contains Doppler bins, and the distance axis contains distance bins. Generally, in pulsed Doppler radar, range-Doppler plots are used to represent processed data from the radar. Range-Doppler plots contain distance and Doppler information between the target and the clutter. Range-Doppler plots can also be used to analyze the relative signal intensity from different scatterers and to evaluate the effect of phase noise in radar target detection.

[0165] In radar, the FFT also performs pulse integration to improve the signal-to-noise ratio (SNR). Since the FFT is performed along the slow time axis, the pulse integration is also performed along the slow time axis, which includes the stack of received pulses. The FFT performs coherent integration of both signal and noise. For the coherent integration of N pulses containing both signal and noise, the power of the integrated signal component is N. 2 The noise power increases by N, and the integrated noise power increases by N. Therefore, the effective increase in SNR is N in dB units.

[0166] For any time-domain signal undergoing a Fourier transform into the frequency domain, discontinuities due to non-integer multiples of the waveform's period result in frequency components leaking into adjacent frequency bins. This leakage diffuses fine spectral lines in the FFT and is known as spectral leakage. Spectral leakage is minimized by a technique known as windowing. Windowing multiplies the measured time-domain discontinuous signal by an amplitude envelope that approaches zero at both ends. The signal-to-noise ratio (SNR) at the output of windowing is degraded by a factor known as the loss factor. The loss factor is defined as the ratio of the maximum achievable SNR (SNR without windowing) to the actual SNR (SNR with windowing). Depending on the application, there are various types of windowing functions, including Hamming windows, Hanning windows, and Blackman-Harris windows.

[0167] 5.2.1 Windowing, FFT, and Range-Doppler Plots in Simulations All radar models apply windowing to the time-domain signals before performing an FFT to generate range-Doppler plots. Following the steering radar, Blackman-Harris windowing was applied to the data processing chain of the radar models. SNR(SNR) at the output of the windowing process. W ) is given by the following equation,

[0168]

number

[0169] Here, LF is the loss coefficient of the Blackman-Harris window. The FFT was performed after the windowing process. The SNR at the output of the FFT is (SNR FFT ) is given by the following equation,

[0170]

number

[0171] Here, N is the number of pulses. Depending on the requirements, the FFT can be performed using any number of pulses. In all radar models, the FFT was performed using 2048 pulses (N=2048). Since only a single fixed test target was present in the simulation, the FFT using 2048 pulses significantly reduces the number of target peak occurrences in the histogram. Therefore, it is impossible to represent the signal power at the output of the FFT in the histogram plot. A range-Doppler plot with a fixed target at 600m is given in Figure 21. Figure 21 includes range-Doppler plots with and without Blackman-Harris windowing. The range-Doppler plot includes several distance bins on the vertical axis depending on the radar configuration. The number of Doppler bins on the horizontal axis is the same as the number of pulses used to generate the range-Doppler plot. The range-Doppler plot in Figure 21 includes 15 distance bins and a central Doppler bin equivalent to a range of 1400Hz.

[0172] The FFT of 2048 pulses yields an SNR gain of 33 dB. The range-Doppler plot without a Blackman-Harris window in Figure 21(a) has a target peak power of 145 dB and an average noise power of 56.5 dB, which is equivalent to an SNR of 88.5 dB. Since the SNR at the MF output was 55.5 dB as shown in Figure 20, the SNR of 88.5 dB at the FFT output is equivalent to the expected increase of 33 dB in SNR units. The range-Doppler plot with a Blackman-Harris window given in Figure 21(b) has a target peak power of 136 dB and an average noise power of 51 dB, which is equivalent to an SNR of 85 dB. The SNR of 85 dB at the FFT output with windowing is 3.5 dB lower than the value of 88.5 dB at the FFT output without windowing. Windowing follows the loss coefficient of the Blackman-Harris window. The loss factor of the Blackman-Harris windowing is between 3 and 3.5 dB, and the SNR value calculated from Figure 21 also shows a loss factor of 3.5 dB. In the simulation, the waveform in the Fourier transform using a fixed target contains only integer multiples of the period. Therefore, frequencies spread to adjacent frequency bins are not visible in the fixed target.

[0173] 6. Verification of the radar model Radar models can be validated by comparing the results of the radar model with both theoretical and real-world radar data. By comparing key metrics in the range-Doppler plot, it is possible to evaluate and validate simulated radar models against real-world radar in the best possible way. SNR and clutter-to-noise ratio (CNR) are two key metrics in the range-Doppler plot.

[0174] The entire radar model described herein benefits from its ability to simulate targets and clutter in the environment at a given distance, velocity, and RCS. The entire radar model is validated by comparing the simulated results with experimentally measured results from an L-band steering radar. Validation of the radar model was facilitated by generating simulated data equivalent to the entire coherent processing interval (CPI). The CPI included durations equivalent to 2048 PRIs. Actual range-Doppler plots from radar trials were compared to simulated range-Doppler plots to validate the model. Several distance bins and 2048 Doppler bins constitute the actual range-Doppler plot.

[0175] Three levels of validation were performed to verify the validity of the entire simulated radar model. In the first level of validation, system-level signal power and SNR values ​​at the outputs of various blocks of the simulation were compared to expected values. In the second stage of validation, the fundamental parameters from the generated range-Doppler plots were compared to the fundamental parameters from the real radar. In the third stage of validation, the CNR, SNR, and thermal noise floor from the simulated range-Doppler plots were compared to those from the real range-Doppler plots.

[0176] 6.1 Comparison of System-Level Signal Power and SNR The first level of verification was performed by comparing the SNR and system-level signal power at the outputs of different building blocks in the transmit chain, receive chain, and data processing chain. In the first stage of verification, a single fixed clutter was simulated at a distance of 600m and an arbitrary RCS value. Simulation time equivalent to the total CPI was performed. Signal power, noise power, and SNR values ​​were measured using test probes placed at the outputs of building blocks in the radar model. It is convenient to divide the first stage of verification into two sections. Signal power is compared in the first section, and SNR values ​​are compared in the second section.

[0177] [Table 1]

[0178] In the first section, the simulated signal power at the outputs of the transmitting amplifier, transmitting antenna, and receiving antenna, given in Figures 7 and 8, is compared to the expected values. The expected values ​​of the signal power at the outputs of the transmitting amplifier and transmitting antenna were taken from the steering radar's power budget. The expected value of the signal power at the output of the receiving antenna was calculated from the radar distance equation. The comparison between the simulated and expected values ​​of the signal power is summarized in Table 1. Table 1 shows the expected values ​​of the signal power at the outputs of the transmitting amplifier and transmitting antenna that are exactly the same as the expected values. The simulated value of the signal power at the output of the receiving antenna is offset by a negligible value of 0.1 dB.

[0179] [Table 2]

[0180] The simulated SNR at the output of each building block in the receiving chain and data processing chain was compared to the expected SNR in Section 2. The simulated SNR values ​​were determined using the signal power and thermal noise power at the output of each building block. Radar theory and power budget reference values ​​were used to calculate the expected SNR values. The simulated and expected SNR values ​​at the outputs of different radar building blocks are summarized and compared in Table 2. Table 2 shows a good agreement between the simulated and expected SNR values ​​at the output of any building block. A constant offset of 0.1 dB between the expected and simulated SNR exists for all SNR values ​​starting from the receiving antenna. The 0.1 dB difference in Table 2 is due to small differences in parameters for target modeling. The SNR at the output with matched filters and FFT in Table 2 is offset by only -0.4 dB. Along with a constant offset of 0.1 dB present across all building blocks, the -0.4 dB difference is the net effect of the 0.5 dB difference between the expected and simulated values ​​of the SNR gain in the matched filtering process. The high parity between the simulated and expected values ​​of system-level signal power and SNR, shown in Tables 1 and 2, indicates the fidelity of the building blocks of the simulated radar model with respect to the theoretical and steering radar power budgets.

[0181] 6.2 Comparison of distance and basic Doppler parameters After comparing system-level signal power and SNR in the first level of verification in Section 6.1, fundamental parameters from range-Doppler plots of both the simulation and the steering radar are compared in the second level of verification. A moving target at a distance of 1500m and an arbitrary RCS, along with a fixed clutter at 600m and an arbitrary RCS, was also simulated in the second level of verification. The RCS and distance values ​​were arbitrary and did not follow any particular criteria. The target was moving at 251Hz, equivalent to a velocity of 30m / s at the radar's specific L-band transmit frequency. Range-Doppler plots with thermal noise were simulated using the full radar model, including both clutter and the target.

[0182] The simulated range-Doppler plot included several distance bins. The PRF of the simulated range-Doppler plot was the same as that of the steering radar's range-Doppler plot. Figure 22 shows the simulated range-Doppler plot with distance bins on the vertical axis and Doppler frequency on the horizontal axis. Figure 30 clearly shows a target with a Doppler frequency of approximately 251 Hz (highlighted in red) and fixed clutter with a Doppler frequency of 0 Hz (highlighted in green), confirming the full radar model's ability to simulate targets and clutter at user-defined speeds.

[0183] The second level of validation compares distance and Doppler fundamental parameters, including distance bins, Doppler bins, distance resolution, and Doppler resolution, for both the simulated radar model and the L-band steering radar. Figure 22 was used to measure the distance and Doppler bins for both the target and clutter. The distance resolution and Doppler resolution for the entire simulated radar model were also calculated from Figure 22. There was excellent agreement between the distance resolution and Doppler resolution between the simulated radar model and the real radar. The comparison of fundamental parameters, including distance bins and Doppler bins, for the simulated radar model and the steering radar is summarized in Table 3.

[0184] [Table 3]

[0185] The simulated values ​​correspond to the values ​​measured from the simulated range-Doppler plot in Figure 22. The steering radar's distance resolution and Doppler resolution were used to calculate the expected values ​​of the distance bin and Doppler frequency bin for both clutter and target. Table 3 shows that the fundamental parameters from the simulation are the same as the expected parameters from the actual radar. The section concludes that the radar model has the ability to replicate the steering radar's distance and Doppler fundamental parameters in the simulation. The radar model described herein can be optimized for all fundamental parameters, including distance resolution, Doppler resolution, CPI, and PRF, for any configuration required by any user.

[0186] 6.3 Comparison of full-range Doppler plots The third stage of verification involves comparing and verifying the CNR, SNR, and thermal noise floor from a simulated range-Doppler plot with that of a real radar range-Doppler plot. Range-Doppler plots from actual radar experiments were used as a baseline for the third level of verification. An L-band steering radar was used for a real radar trial conducted at an airfield where a controlled drone (the target of interest) flew along a predetermined course. The real range-Doppler plot consisted of a single frame from a single beam at fixed azimuth and elevation angles centered on the direction of the target in the drone radar trial, equivalent to a full CPI. In the selected frame, the drone approached the radar at a specific distance and a radial velocity of 8.3 m / s. Despite the radar trial being conducted in a localized environment, the real range-Doppler plot included other unwanted targets in the radar's field of view.

[0187] The third stage of validation was carried out by simulating actual range-Doppler plots using a comprehensive radar model. The range-Doppler plots were initially simulated using CNRs from simulations that were close to each other and CNRs from actual range-Doppler plots. Because clutter power in each distance bin leaks to adjacent distance bins, the current model is not capable of simulating an exact replica of the CNR for each distance bin. For both real and simulated range-Doppler plots, the CNRs were calculated taking thermal noise into account.

[0188] [Table 4]

[0189] Table 4 compares the CNR from simulated range-Doppler plots with the CNR from real-world range-Doppler plots for different distance bins. Real-world radar environments contain a wide variety of clutter, each with a different RCS. In the simulation, clutter power is replicated by modeling a single clutter within each distance bin to represent a combination of real-world clutter. Also, in the simulation, clutter power from each distance bin leaks into adjacent distance bins. Therefore, a variation of a few dB in CNR levels between the simulation and real-world data is acceptable. The values ​​in Table 4 demonstrate a strong agreement between the simulated and real-world CNRs.

[0190] The second step in the final level of validation involved adding target information to the range-Doppler plot in the simulation. The target drone was represented in the simulation as an object at a distance equal to the actual distance to the target. The target's Doppler frequency was simulated at 69.8 Hz, equivalent to a drone speed of 8.3 m / s. The drone's RCS was modeled so that the SNR of both the simulation and the real-world data were comparable.

[0191] In any real-world radar, phase noise plays a significant role in target detection. Phase noise in a radar system causes the phase noise floor in a range-Doppler plot to be revealed from the thermal noise floor in distance bins with stronger clutter power. The overall increase in noise floor makes target detection more difficult. The introduction of phase noise in radar simulations and its effects on range-Doppler plots are discussed herein.

[0192] A comparison of simulated and real-world range-Doppler plots is given in Figure 23. Both range-Doppler plots in Figure 23 are normalized to the strongest clutter. Figure 23(a) shows the simulated range-Doppler plot, and Figure 23(b) shows the real-world range-Doppler plot. In Figure 23(b), the increase in the overall noise floor and the appearance of a phase noise floor above the thermal noise floor are visible at distances with extremely high clutter. Since phase noise is not added to the full radar model simulation, the simulated range-Doppler plot in Figure 23(a) includes only a uniform thermal noise floor. The real-world range-Doppler plot in Figure 23(b) also includes an additional peak corresponding to an unwanted target in the radar field of view.

[0193] Except for the phase noise floor and unwanted targets, the simulated range-Doppler plot and the actual range-Doppler plot are qualitatively similar in Figure 23. Clutter power per distance bin does not independently match because the CNR per distance bin cannot be accurately reproduced in the simulation. The target of interest is highlighted in a red box in both range-Doppler plots in Figure 33. In both the simulated and actual range-Doppler plots, the target appears in the 19th distance bin and the 19th Doppler bin. The signal-to-thermal noise ratio of the target is calculated from both the simulated and actual range-Doppler plots and is given in Table 5. Table 5 shows excellent agreement between the simulated and actual signal-to-thermal noise ratios. The high level of parity in the signal-to-thermal noise ratio confirms the ability of the comprehensive radar model to simulate range-Doppler plots with targets and thermal noise floors that replicate values ​​from the actual range-Doppler plot.

[0194] [Table 5]

[0195] Peak clutter power and thermal noise floor for each distance bin were taken from Figures 23(a) and 23(b) to generate comparison graphs. The thermal noise floor refers to the noise floor for distance bins far from the radar, where the effect of phase noise induced by clutter can be ignored. The graphs in Figure 24 compare the simulated range-Doppler plots and the real-world range-Doppler plots from Figures 23(a) and 23(b), respectively.

[0196] Figure 24 clearly shows the overlapping thermal noise floors for both the real-world range-Doppler plot and the simulated range-Doppler plot. The comparison graph also shows a good agreement between the simulated clutter power and the real-world clutter power. The clutter power for distance bin 21 from the real-world range-Doppler plot was an exception that could not be replicated in the simulation. Clutter in real radar includes complex objects and other weather phenomena in the radar field of view. In contrast, clutter in the simulation is represented by a single fixed clutter in each distance bin. Because the radar pulse width is longer than the distance bins, there is an additional effect of clutter power leaking into adjacent distance bins. Due to these two effects, clutter power in any distance bin cannot be replicated in the simulation with total parity. Figures 23(a), 23(b), and 24, along with Tables 4 and 5, show the reliability and fidelity of the comprehensive radar model for replicating the real-world steering radar trial in the simulation.

[0197] The range-Doppler plot of the simulated radar trial and the actual steering radar trial are given in Figure 25(c).

[0198] Figure 25(c) clearly shows good parity between the real range-Doppler plot with phase noise of a classical oscillator and the simulated range-Doppler plot. Figure 25(c) represents the ability of a comprehensive radar model to replicate a real steering radar trial at a very detailed level. The real range-Doppler plot in Figure 25(b) includes unwanted targets in the radar field of view along with the target of interest. At the same time, only the target of interest is modeled in the simulated range-Doppler plot in Figure 25(a). In both range-Doppler plots in Figure 25(c), the presence of a phase noise floor is clearly evident. A phase noise floor can be seen emerging from the thermal noise floor for distance bins with higher clutter power. Since the reflected power from the clutter is inversely proportional to the fourth power at the clutter distance, in almost all cases, the distance bins closer to the radar will have the highest clutter power. In Figure 7, it is clearly evident that the phase noise floor emerges from the thermal noise floor in the distance bins (distance bins 1-4) with high clutter power and close to the radar. The effect of phase noise is also visible for distance bins 6, 7, 9, and 10, which have relatively higher clutter power.

[0199] A better comparison of the relationship between simulated range-Doppler plots with phase noise for a classical oscillator and real-world range-Doppler plots is given in Figure 26. The comparison graph in Figure 26 shows the peak clutter power, thermal noise floor, and overall noise floor for both the simulated range-Doppler plot and the real-world range-Doppler plot in Figure 25(c). The thermal noise floor refers to the noise floor for distance bins that are far from the radar, where the effect of phase noise induced by clutter can be ignored. The comparison graph in Figure 26 is an extended version of the comparison graph described earlier, with the inclusion of the overall noise floor in each distance bin. The overall noise floor consists of both the thermal noise floor and the phase noise floor.

[0200] Figure 27 schematically illustrates the method according to an exemplary embodiment.

[0201] This method simulates a RADAR system defined by a corresponding RADAR model comprising a transmission chain, a reception chain, and a data processing chain. This method is implemented by a computer comprising a processor and a memory, and this method transmitting, by means of the transmission chain, a radio frequency (RF) signal into an environment including a target and optionally clutter (2702); calculating, for each reflected RF signal from the target and from optionally present clutter, each reflected RF signal of the transmitted RF signal (2704); receiving, by means of the reception chain, each reflected RF signal (2706); detecting a target using each received reflected RF signal by means of the data processing chain (2708); and determining a simulated output of the RADAR system using the result of detecting the target (2710).

[0202] FIG. 28 schematically shows a method according to an exemplary embodiment.

[0203] This method is a method for validating a RADAR model comprising a transmission chain, a reception chain, and a data processing chain corresponding to a RADAR system. This method is implemented by a computer comprising a processor and a memory, and this method obtaining an output collected from the RADAR system for an environment including a target and optionally clutter (2802); creating a RADAR model corresponding to the RADAR system (2804); determining a simulated output of the RADAR model for an environment including a target and optionally clutter by simulating the RADAR model (2806); and validating the RADAR model by comparing the simulated output with the collected output (2808). It is equipped with.

[0204] 7 Conclusion This specification presents the development of a comprehensive radar model in simulation and its validation using data from real-world radar. The radar model, including all fundamental building blocks in the transmit, receive, and data processing chains, is a powerful tool for testing virtual and real-world radar scenarios, along with targets and clutter in the environment. Such a comprehensive radar model was developed from scratch, focusing on the behavior of radar hardware building blocks. The radar model was validated by comparing the results generated in simulation with actual experimental results from L-band steering radar. The radar model was capable of replicating real-world radar parameters at various levels.

[0205] Detailed studies characterizing the performance of different classical and quantum oscillators will enable comprehensive radar models to analyze the limitations of phase noise in conventional radar oscillators and explore the potential advantages of quantum oscillators in radar systems.

[0206] While preferred embodiments have been illustrated and described, those skilled in the art will understand that various changes and modifications can be made without departing from the scope of the invention, as defined in the appended claims and described above.

[0207] Attention is drawn to all papers and documents filed concurrently with or prior to this specification in connection with this application and made available to the public together with this specification, and the contents of all such papers and documents are incorporated herein by reference.

[0208] All features disclosed herein (including any appended claims and drawings) and / or all steps of any method or process disclosed herein may be combined in any combination, except for any combination in which at most some of such features and / or steps are mutually exclusive.

[0209] Each feature disclosed herein (including any attached claims and drawings) may be replaced by an alternative feature serving the same, equivalent, or similar purpose unless otherwise specified. Therefore, unless otherwise specified, each disclosed feature is merely an example of a general set of equivalent or similar features.

[0210] The present invention is not limited to the details of the embodiments described above. The present invention extends to any novel one or any novel combination of features disclosed herein (including any appended claims and drawings), or any novel one or any novel combination of any steps of any method or process so so disclosed.

Claims

1. A method for simulating a RADAR system defined by a corresponding RADAR model comprising a transmission chain, a reception chain, and a data processing chain, which is implemented by a computer comprising a processor and memory. The aforementioned transmission chain transmits radio frequency (RF) signals to an environment that includes a target and, optionally, clutter. Calculating the reflected RF signals from the target and from the optional clutter, The receiving chain receives each of the reflected RF signals, The data processing chain detects the target using each of the received reflected RF signals, The results of detecting the aforementioned target are used to determine the simulated output of the RADAR system. A method for providing this.

2. The method according to claim 1, wherein transmitting the RF signal to the environment including the target and optionally the clutter by the transmission chain comprises generating a continuous RF signal having a given amplitude envelope, optionally converting the continuous RF signal into a pulsed, amplitude-modulated RF signal, and optionally amplifying the pulsed, amplitude-modulated RF signal.

3. The method according to claim 1 or 2, wherein the target comprises a moving target or a stationary target, and / or is a moving target or a stationary target, and optionally the clutter comprises a stationary clutter, and / or is a stationary clutter.

4. The method according to any one of claims 1 to 3, wherein calculating the respective reflected RF signals of the transmitted RF signals from the target and the optional clutter is further comprising:

5. The method according to any one of claims 1 to 4, wherein the receiving chain receives the respective reflected RF signals, the receiving chain comprises receiving the respective reflected RF signals of the transmitted RF signals from the target and the optional clutter, optionally adding thermal noise to the received, respective reflected RF signals, optionally amplifying the received, respective reflected RF signals, optionally down-converting the received, respective reflected RF signals to, for example, a first intermediate frequency (IF1) signal, optionally filtering the IF1 signal, and optionally down-converting the IF1 signal to a second intermediate frequency (IF2) signal.

6. The method according to any one of claims 1 to 5, wherein the detection chain detects the target using each of the received reflected RF signals, and the data processing chain processes each of the received reflected RF signals, which includes one or more of the following: performing a Hilbert transform on each of the received reflected RF signals; performing a match filter on each of the received reflected RF signals; performing a window on each of the received reflected RF signals; performing a Fast Fourier Transform (FFT) on each of the received reflected RF signals; and generating a range-Doppler plot.

7. The method according to any one of claims 1 to 6, comprising adapting the RADAR model, for example, the transmission chain, the reception chain, and the data processing chain.

8. The method according to claim 7, wherein updating the RADAR model comprises updating the RADAR model using the simulated output of the RADAR system.

9. Amounts to modify the target and / or the optional clutter, Herein, calculating the respective reflected RF signals of the transmitted RF signal from the target and the optional clutter comprises calculating the respective reflected RF signals of the transmitted RF signal from the modified target and the optional modified clutter. The method according to any one of claims 1 to 8.

10. The method according to any one of claims 1 to 9, wherein the transmission block comprises a quantum oscillator.

11. A method for providing a RADAR system, which is implemented at least in part by a computer having a processor and memory, To simulate a RADAR model corresponding to the aforementioned RADAR system, comprising a transmission chain, a reception chain, and a data processing chain, The results of simulating the aforementioned RADAR model are used to provide the RADAR system. A method for providing this.

12. The method according to claim 11, further comprising updating the RADAR model based on the results of verifying the RADAR model.

13. The method according to claim 12, comprising simulating the updated RADAR model.

14. A RADAR system provided by the method according to any one of claims 11 to 13.

15. A computer comprising a processor and memory configured to implement the method according to any one of claims 1 to 13; a computer program comprising instructions that, when executed by the computer comprising the processor and memory, cause the computer to perform the method according to any one of claims 1 to 13; or a non-temporary computer-readable storage medium comprising instructions that, when executed by the computer comprising the processor and memory, cause the computer to perform the method according to any one of claims 1 to 13.

Images