CN110138474B

CN110138474B - A zoom panorama microwave spectrum monitoring method

Info

Publication number: CN110138474B
Application number: CN201910312696.4A
Authority: CN
Inventors: 游飞; 徐茂加; 王鹏; 游冠雄
Original assignee: University of Electronic Science and Technology of China
Current assignee: University of Electronic Science and Technology of China
Priority date: 2019-04-18
Filing date: 2019-04-18
Publication date: 2021-07-06
Anticipated expiration: 2039-04-18
Also published as: CN110138474A

Abstract

The invention discloses a zoom-type panoramic microwave spectrum monitoring method, which belongs to the technical field of radio frequency microwave communication, and is a zoom-type full-band microwave spectrum monitoring system. Use the MWC compressed sensing receiver to sense the signal distribution on the entire spectrum, and then use the adaptive local oscillator to realize zoom observation, achieve non-uniform frequency focusing ability, reduce the signal power of the local oscillator in the no-signal frequency band, and enhance the frequency mixing of the frequency band of interest gain, thereby improving the receiver sensitivity and suppressing the pressure of the interference signal on the back end of the receiver. The Fourier coefficient of the local oscillator signal is measured by the low-cost integrated online test method, and the signal recovery algorithm is added to realize the recovery of the signal of the selected frequency band.

Description

Zoom type panoramic microwave spectrum monitoring method

Technical Field

The invention belongs to the technical field of radio frequency microwave communication, and relates to a zoom type full-band microwave spectrum monitoring system.

Background

The consumption of DC-6GHz spectrum resources is almost exhausted, and communication systems such as 5G and 5G beyond are bound to move to the millimeter wave frequency band. Various systems such as communication, radar, navigation and the like can lead to radio environment becoming ever more complex, and requirements for the specification of millimeter wave spectrum resources, monitoring and positioning of interference signals and the like are gradually increased. Because the traditional receivers are all in a narrow-band receiver mode, a large amount of time is consumed for completing full-band scanning, and serious challenges are brought to real-time dynamic spectrum acquisition, detection and tracking. If a broadband receiver mode of directly and digitally sampling microwave signals is adopted, ADC devices with extremely high sampling rates are needed, on one hand, the cost is high, the ADC devices are in a forbidden list, and even if the ADC devices can be realized, time domain sampling needing to wait for a long time is needed to meet the technical requirement of spectrum resolution. The compressed sensing receiver is a better compromise scheme appearing in recent years, and under the condition of sacrificing a certain sensitivity, the scheme can effectively reduce the speed requirement of an ADC (analog-to-digital converter) device and meet the requirement of ultra-wideband real-time receiving. However, the millimeter wave communication frequency band may extend to around 60GHz, and the single carrier frequency band width may exceed 1GHz, which is still a serious challenge for the implementation of the classical compressed sensing receiver. Because its classical architecture still requires the signal to meet the sparsity requirement while the effective signal bandwidth of the processing is limited, e.g. 200MHz total bandwidth. And compared with a narrow-band receiver, the signal sensitivity of the classical compressed sensing receiver is greatly reduced, and the reduction amplitude is consistent in the full frequency band. For example, the theoretical limit of signal sensitivity of a DC-3GHz compressed sensing receiver is about-75 dBm, which can greatly influence the detection and demodulation of weak signals. Under the technical characteristic requirement of millimeter wave communication, the ultra-wideband frequency spectrum monitoring and the capability of demodulating and receiving large-bandwidth signals can be realized, and the method is an important mark for breaking through the current microwave frequency spectrum monitoring problem.

Mwc (modulated Wideband converter) compressed sensing receiver is a sub-nyquist sampling receiver proposed in 2010, which can be applied to multiband signals. The MWC receiver performs spectrum aliasing on an input signal by utilizing a pseudorandom periodic sequence at the front end of the system, moves high-frequency information of the signal to a baseband frequency, and then performs low-pass filtering and sampling on a mixing signal. Because a plurality of sampling channels are adopted, each sampling channel carries out different frequency mixing on the signals, and therefore, relatively sparse signals can be recovered from more frequency mixing signal sampling values in the following digital signal processing.

The classic MWC compressed sensing receiver can basically and uniformly mix all signals of a receiving frequency band to a low frequency, linear superposition is carried out, and finally signal restoration is carried out through a signal restoration algorithm. The local oscillation signals adopted by the method are random sequences, and the power level of the local oscillation signals on each discrete frequency point is uniform in statistics, so that the distribution of all signals on the whole frequency band can be sensed, and the method is favorable for uniformly receiving and processing the signals of all frequency bands. However, the scheme of uniform reception is not favorable for improving the sensitivity of the receiver, and the implementation of the high-speed random local oscillator is difficult, the frequency mixing gain of the millimeter wave frequency band signal is low, the signal sensitivity is extremely poor, and the influence of noise and interference is large.

Disclosure of Invention

The invention provides a rapid spectrum monitoring system in a DC-60GHz frequency range based on an MWC compression sensing receiver, which can sense the spectrum distribution of each signal in the frequency band and carry out zoom type observation.

The invention discloses a zoom type panoramic microwave spectrum monitoring method, which comprises the following steps:

step 1: the collected signals x (t) are simultaneously inputted into m channels, andwith local oscillator p in each channel_i(t) phase mixing, local oscillator signal p for each channel_i(t) is a pseudo-random sequence signal with the same period and different time domain waveforms;

step 2: the signals after frequency mixing in each channel in the step 1 enter a low-pass filter, and then the signals after low-pass filtering are sampled to obtain a digital signal y_i[n]；

And step 3: performing digital signal processing according to the obtained digital signal to complete spectrum sensing and obtain a frequency band where the expected signal is located;

step 4, according to the frequency spectrum distribution of the current signal, the local oscillation signal p in each channel is obtained_i(t) replacing with a self-adaptive local oscillator; the method for generating the self-adaptive local oscillator comprises the following steps:

the method comprises the following steps: establishing a self-adaptive local oscillator library in advance, wherein the self-adaptive local oscillator library comprises a plurality of different local oscillator signals, each group of local oscillator signals has obvious frequency selection characteristics, the frequency selection interest frequency ranges of the local oscillator signals of each group are different, and according to the frequency range of the expected signal sensed by the frequency spectrum in the step 3, selecting a corresponding group of local oscillator signals to perform frequency mixing with the acquired signal x (t) in all channels, and the same as the traditional MWC compressed sensing receiving;

the second method comprises the following steps: generating a plurality of single-tone signals in a corresponding frequency band by adopting a plurality of phase-locked loops according to the frequency band of the expected signal sensed by the frequency spectrum in the step 3, mixing the generated plurality of single-tone signals, performing power modulation on the mixed single-tone signals according to the power of the acquired signal to generate a group of local oscillation signals, and mixing the group of local oscillation signals with the acquired signal x (t) in all channels;

the third method comprises the following steps: directly and instantly generating by using an FPGA (field programmable gate array), determining the frequency spectrum characteristics of a local oscillation signal suitable for the current signal through the frequency spectrum perceived by the frequency spectrum in the step 3, reversely deducing the time domain form of the required local oscillation signal, generating a sequence group closest to the time domain form by the FPGA to be used as the local oscillation signal, and mixing the local oscillation signal with the acquired signal x (t) in all channels;

and 5: respectively enabling the signals subjected to frequency mixing in the m channels to enter a low-pass filter, and then sampling the signals subjected to low-pass filtering to obtain expected signal digital signals; and carrying out digital signal transmission processing on the acquired digital signal of the expected signal to finish signal recovery.

The MWC compressed sensing receiver is improved, the inherent defects of the MWC compressed sensing receiver are overcome, the sensing capability of the MWC compressed sensing receiver on a frequency spectrum is reserved, the MWC compressed sensing receiver can be transited to any interested frequency band range from a panorama, the real-time microwave signal monitoring capability can be provided in any interested frequency band range which does not exceed the specification limit, and the signal sensitivity can be adaptively optimized (zooming capability) along with the reduction of the bandwidth of a simultaneously captured frequency band.

The technical scheme of the invention is that an MWC compression sensing receiver is used for sensing signal distribution on the whole frequency spectrum, and then adaptive local oscillation is used for realizing zoom type observation, so that the non-uniform frequency focusing capability is realized, the local oscillation signal power of a signal-free frequency band is reduced, and the interest frequency band mixing gain is enhanced, thereby improving the sensitivity of the receiver and inhibiting the pressure of an interference signal on the rear end of the receiver. The Fourier coefficient of the local oscillation signal is measured by using a low-cost integrated online test method, and the signal recovery algorithm is added to realize the recovery of the signal of the selected frequency band.

Drawings

FIG. 1 is a system workflow;

FIG. 2 is a diagram of various signals distributed across a frequency spectrum;

FIG. 3 illustrates the operation of an MWC compressed sensing receiver;

FIG. 4 is a diagram of an adaptive local oscillator generator

FIG. 5 is a pseudo-random local oscillator signal spectrum;

FIG. 6 is a spectrum of a signal at an intermediate frequency of an MWC compressed sensing receiver;

FIG. 7 is a self-adaptive local oscillator spectrum;

fig. 8 is a signal spectrum at an intermediate frequency of the receiver of the present invention.

Detailed Description

The working principle of the entire zoom type spectrum monitoring system is described with reference to the accompanying drawings, and the working flow of the entire system is shown in fig. 1.

The input multi-band rf signal of the system is distributed with various signals over the whole frequency spectrum, and it is assumed that the multi-band rf signal x (t) is distributed with 3 different signals a, B, C in the frequency domain as shown in fig. 2. The first step of spectrum monitoring is to sense the whole spectrum, and the MWC receiver can complete the sensing of the spectrum.

The operating principle of the MWC receiver is shown in fig. 3. x (t) enter m channels simultaneously and are respectively connected with local oscillators p of all channels_i(t) phase mixing, local oscillator signal p for each channel_iAnd (t) are all pseudo-random sequence signals with the same period, but the time domain waveforms of each local oscillator are different. The mixed signal enters a low-pass filter and is then sampled by the ADC into a digital signal y_i[n]. Since the local oscillator is a pseudo-random sequence, the local oscillator is represented in a frequency domain as a multi-tone form with different amplitudes and equal intervals, as shown in fig. 5. After x (t) is mixed, due to the characteristic of the local oscillator signal in the frequency domain, the signals a, B, and C distributed in different frequency bands in x (t) are shifted to the intermediate frequency and subjected to linear aliasing, and the spectrum of the mixed and overlapped intermediate frequency signal is shown in fig. 6. For passing different local oscillators p_i(t) mixed x (t), which output different intermediate frequency signals. Due to the output signal y_i[n]The linear superposition of all the frequency spectrum components of the input signal is included, and the frequency spectrum sensing of all the signals can be completed through digital signal processing, so that the positioning of the signals on a frequency domain is realized.

In order to realize the zoom type observation, after the MWC compressed sensing receiver senses the frequency spectrum, the original pseudo-random sequence local oscillator is not used any more, and an adaptive local oscillator is used instead, which is also the key point of the invention. Since the information about the distribution of the signal in the frequency domain has been obtained in the preceding sensing stage, it becomes possible to observe signals in different frequency bands by adjusting the local oscillator. The local oscillator used is generated by an adaptive local oscillator generator, as shown in fig. 4. The frequency domain of the self-adaptive local oscillator is different from the original pseudo-random local oscillator, all signals do not need to be mixed, and only the signals which are interested by people need to be mixed to the intermediate frequency. For example, we want to receive signals a, B and are not interested in C, then the frequency domain of the adaptive local oscillator can be as shown in fig. 7. The intermediate frequency spectrum after using this adaptive local oscillator is shown in fig. 8. By adjusting the self-adaptive local oscillator, radio frequency signals in different frequency bands can be moved to the intermediate frequency, and meanwhile, signals in other frequency bands can be prevented from being moved to the intermediate frequency as far as possible. The method can increase the frequency mixing gain of the frequency band of interest, thereby improving the sensitivity of the receiver, inhibiting the pressure of interference signals on the rear end of the receiver and realizing the focusing of observation frequency.

The accurate test of the local oscillator signal Fourier coefficient is a key technology for realizing signal reduction, when the self-adaptive local oscillator is controlled and changed, the workload of testing the Fourier coefficient in advance is huge, and the influence of circuit non-ideal factors is larger. And then the useful signal is restored by applying a signal monitoring and restoring algorithm in a classical compressed sensing receiver.

Claims

1. A zoom type panoramic microwave spectrum monitoring method, the method comprising:

Step 1: Input the collected signal x(t) into m channels at the same time, and mix it with the local oscillator p _i (t) in each channel, when the local oscillator signal p _i (t) of each channel is the same period Pseudo-random sequence signals with different domain waveforms;

Step 2: Enter the mixed signal in each channel of Step 1 into a low-pass filter, and then sample the low-pass filtered signal to obtain a digital signal y _i [n];

Step 3: perform digital signal processing according to the obtained digital signal, complete spectrum sensing, and obtain the frequency band where the desired signal is located;

Step 4: According to the spectrum distribution of the current signal, replace the local oscillator signal p _i (t) in the original channels with an adaptive local oscillator; the generation method of the adaptive local oscillator is:

An adaptive local oscillator library is established in advance. The adaptive local oscillator library includes a plurality of different local oscillator signals. The local oscillator signals of each group have obvious frequency selection characteristics, and each group of local oscillator signals has different frequency bands of interest for frequency selection. According to the frequency band of the desired signal detected by the spectrum in step 3, select a corresponding group of local oscillator signals to mix with the acquisition signal x(t) in all channels, which is the same as the traditional MWC compressed sensing reception;

Step 5: Enter the mixed signals in the m channels in step 4 into the low-pass filter, and then sample the low-pass filtered signal to obtain the desired digital signal; digitally transmit the obtained desired digital signal. Signal processing to complete signal recovery.

2. A zoom type panoramic microwave spectrum monitoring method, the method comprising:

According to the frequency band of the desired signal detected by the spectrum in step 3, multiple phase-locked loops are used to generate multiple single-tone signals in the corresponding frequency band, and the generated multiple single-tone signals are mixed. Perform power modulation to generate a group of local oscillator signals, and use this group of local oscillator signals to mix with the acquisition signal x(t) in all channels;

3. A zoom type panoramic microwave spectrum monitoring method, the method comprising:

Use the FPGA to generate it directly and instantly, determine the spectral characteristics of the local oscillator signal suitable for the current signal through the frequency band of the desired signal detected by the spectrum in step 3, and invert the time domain shape of the required local oscillator signal, and the FPGA generates the closest time domain shape. The sequence group is used as the local oscillator signal, and this group of local oscillator signals is mixed with the acquisition signal x(t) in all channels;