CN111711457A - A method for improving demodulation bandwidth through multi-channel parallel segment demodulation - Google Patents

Abstract

The invention discloses a method for improving a demodulation broadband through a multi-channel parallel segmented demodulation mode, which is based on a zero intermediate frequency structure, firstly sets the frequency of each path of radio frequency local oscillation signal, the cut-off frequency of a low-pass filter and the ADC sampling rate during orthogonal demodulation, then carries out multi-path orthogonal demodulation on an input signal, enables each path to output an analog baseband complex signal, obtains multi-path digital baseband complex signals after carrying out dual-path sampling through the ADC, and finally merges the multi-path baseband complex signals to realize the expansion of demodulation bandwidth.

Description

A method for improving demodulation bandwidth through multi-channel parallel segment demodulation

technical field

The invention belongs to the technical field of wireless communication, and more particularly, relates to a method for improving demodulation bandwidth through multi-channel parallel segment demodulation mode.

Background technique

Two frequency conversion schemes have been developed in modern wireless communication technology, namely super-heterodyne and zero-IF schemes. In essence, they both convert RF signals into IF signals that are easy to process. The difference is that the frequencies of the IF signals processed are different.

The superheterodyne structure of the first-order mixing is shown in Figure 1. The superheterodyne scheme first uses a band-pass filter to filter the input RF signal, and then uses the local oscillator signal to mix the input signal after low-noise amplification and image rejection filtering to down-convert the input signal to a preset signal. The frequency, after filtering and selection, is sampled for subsequent analog-to-digital conversion. First, this method has a large receiving dynamic range; second, it has high adjacent channel selectivity and receiving sensitivity; third, compared with the zero-IF scheme, it does not require a complex DC cancellation circuit; fourth, because this method generally One or more stages of intermediate frequency mixing are used to suppress the image frequency, so the hardware circuit is complex and the integration is not high; fifthly, this method will use a lot of relatively expensive and bulky filters; sixthly, it will consume a lot of money power. In general, superheterodyne receivers are difficult to integrate and have a large dynamic range. Generally, they are widely used in long-distance communications with high frequencies and weak signals.

The structure of the zero-IF receiver is shown in Figure 2. The zero-IF scheme firstly adjusts the amplitude of the input RF signal through a low-noise amplifier and an attenuator, filters it through a RF filter, and then directly converts the RF signal to a zero-frequency signal by mixing the RF signal with the local oscillator signal with the same frequency. After baseband filtering, it is sampled by subsequent circuits. In typical phase/amplitude modulation, the zero-IF architecture requires quadrature I and Q signals. The phase changes of the I and Q signals represent two sidebands, and the two sidebands contain different information. At the same time the presence of the IQ signal facilitates digital signal processing.

The advantage of the zero-IF scheme is that firstly, when an analog-to-digital converter chip with the same sampling rate is used, the bandwidth is twice that of the superheterodyne architecture. Second, the structure is simple and the volume is small, which is conducive to integrated design. With the increase in the complexity of the communication system in recent years and the design requirements of the project's high-IF bandwidth and high integration, the zero-IF solution can better solve the above problems.

In the basic block diagram of zero-IF RF reception shown in Figure 3, set the input RF signal as x(t)=cos(2πf _RF t), and the complex LO signal as x _L (t)=cos(2πf _L t)-jsin (2πf _L t), the cut-off frequency of the low-pass filter on the two channels is BW/2, the sampling rate of the analog-to-digital converter ADC is f _s and f _s ≥ BW. Suppose the ADC sampling output signal is x(n), then x(n) can be expressed as:

x(n)=(cos(2πf _IF n/f _s )+sin(2πf _IF n/f _s ))/2

Wherein, the baseband signal frequency f _IF =f _RF -f _L and |f _IF |≤BW/2. Since f _IF can be positive or negative, as long as the frequency f _L of the local oscillator signal is changed, the signal in the frequency range of f _L ±BW/2 in the input RF signal can be directly quadrature down-converted to the baseband, and the complex signal x(n ). In other words, using the structure shown in FIG. 3 , the input radio frequency signal can be directly quadrature demodulated to the baseband with f _L as the center frequency (or carrier frequency) and the information (or modulation signal) in the frequency range of BW, and the baseband frequency The range is [-BW/2, BW/2], where BW is called the demodulation bandwidth of the structure shown in FIG. 3 . The wider the demodulation bandwidth, the wider the frequency range that can be collected and analyzed at one time, the more information can be obtained, and the ability to observe certain transient signals can be greatly improved.

Some major chip manufacturers at home and abroad have introduced quadrature demodulation chips that work in different frequency ranges and have different demodulation bandwidths, and most of them have the internal structure shown in Figure 3 (without ADC). However, in practical applications, there may be situations in which the demodulation chip cannot meet the requirements of frequency range and demodulation bandwidth at the same time.

SUMMARY OF THE INVENTION

The purpose of the present invention is to overcome the deficiencies of the prior art, and to provide a method for improving the demodulation bandwidth through a multi-channel parallel subsection demodulation method. Tuning bandwidth extension.

In order to achieve the above-mentioned purpose of the invention, a method for improving demodulation bandwidth through multi-channel parallel segmented demodulation mode of the present invention is characterized in that, comprising the following steps:

(1), the frequency f _Lk of each radio frequency local oscillator signal when setting quadrature demodulation;

Suppose the center frequency range of the input signal x(t) is f _L ±BW/2, and the entire demodulation range f _L ±BW/2 is equally divided into M equal parts, that is, the bandwidth of the demodulation BW/M of each channel, then The RF local oscillator signal frequency f _Lk input to the quadrature demodulation chip with narrow demodulation bandwidth of M channels satisfies:

f _Lk =f _L +(2k+1-M)BW/2M

Among them, k=0,1,2,...,M-1, M is an even number; BW is the demodulation bandwidth;

(2), set the cut-off frequency of the low-pass filter to BW/2M, the ADC sampling rate is f _s and f _s ≥ BW;

(3) Input the input signal x(t) to the quadrature demodulation chip with M channels of narrow demodulation bandwidth, and perform quadrature demodulation with M channels of RF local oscillator signal x _Lk (t)=exp(j2πf _Lk t) , and then pass the demodulated signal of each channel through the filtering process of the low-pass filter, so that each channel of low-pass filter outputs the analog baseband complex signal x _k (t );

(4), carry out dual-channel sampling by ADC for each channel of analog baseband complex signal x _k (t) to obtain M channels of digital baseband complex signal x _k (n);

(5) Perform digital complex mixing of each channel of digital baseband complex signal x _k (n) and the corresponding digital replica oscillator y _Lk (n)=exp(jω _Lk n), and then combine the real output signal of the complex mixing frequency. The part and the imaginary part are added correspondingly to obtain a digital complex signal with a sampling rate of f _s and an effective signal bandwidth of BW;

Among them, the frequency of the digital replica oscillator ω _Lk satisfies:

The purpose of the invention of the present invention is achieved in this way:

The present invention is a method for improving the demodulation bandwidth through multi-channel parallel subsection demodulation method. Based on the zero-IF structure, the frequency of each channel of radio frequency local oscillator signal, the cut-off frequency of the low-pass filter and the ADC during quadrature demodulation are set first. sampling rate, and then perform multi-channel quadrature demodulation on the input signal, so that each channel outputs an analog baseband complex signal. After dual-channel sampling by ADC, multiple digital baseband complex signals are obtained, and finally the multi-channel baseband complex signals are combined. Realize the expansion of demodulation bandwidth.

At the same time, a method for improving the demodulation bandwidth through the multi-channel parallel segment demodulation mode of the present invention also has the following beneficial effects:

(1), the present invention adopts the multi-channel parallel segmented demodulation mode to greatly reduce the requirement for the demodulation bandwidth of the quadrature demodulation chip;

(2) The present invention expands the frequency range by means of frequency mixing, thereby solving the problem that the working frequency range of the demodulation chip cannot meet the requirements;

(3) The present invention realizes wideband expansion through parallel demodulation of multiple demodulation chips with the same demodulation bandwidth, thereby solving the problem that the demodulation bandwidth cannot meet the requirements.

Description of drawings

Figure 1 is a schematic diagram of a superheterodyne structure of first-order mixing;

Figure 2 is a schematic diagram of the structure of a zero-IF receiver;

Figure 3 is a basic block diagram of zero-IF RF reception;

Fig. 4 is a kind of principle block diagram of improving demodulation broadband by multi-channel parallel segmented demodulation mode of the present invention;

Fig. 5 is a block diagram of the simulation structure of the two-way parallel demodulation circuit;

Fig. 6 is the output graph of channel 0 demodulation result;

Fig. 7 is the output diagram of channel 1 demodulation result;

Figure 8 is the output diagram of channel 0 digital complex mixing result;

Figure 9 is the output diagram of channel 1 digital complex mixing result;

Figure 10 is the total demodulation output diagram of the system when the input frequency is 2.6kHz and the local oscillator frequency is 2.2kHz;

Figure 11 is the total demodulation output diagram of the system when the input frequency is 1.8kHz and the local oscillator frequency is 2.2kHz;

Figure 12 is the total demodulation output diagram of the system when the input frequency is 2.1kHz and the local oscillator frequency is 2.2kHz;

Figure 13 is the total demodulation output diagram of the system when the input frequency is 2.4kHz and the local oscillator frequency is 2.2kHz.

Detailed ways

The specific embodiments of the present invention are described below with reference to the accompanying drawings, so that those skilled in the art can better understand the present invention. It should be noted that, in the following description, when the detailed description of known functions and designs may dilute the main content of the present invention, these descriptions will be omitted here.

Example

FIG. 4 is a schematic block diagram of the present invention for improving the demodulation bandwidth through a multi-channel parallel segmented demodulation method.

In the present embodiment, as shown in FIG. 4 , the demodulation broadband extension is realized by the parallel demodulation of M (M is an even number) narrow demodulation bandwidth quadrature demodulation chip, which specifically includes the following steps:

S1, the frequency f _Lk of each radio frequency local oscillator signal when setting quadrature demodulation;

Suppose the center frequency range of the input signal x(t) is f _L ±BW/2, and the entire demodulation range f _L ±BW/2 is equally divided into M equal parts, that is, the bandwidth of the demodulation BW/M of each channel, then The RF local oscillator signal frequency f _Lk input to the quadrature demodulation chip with narrow demodulation bandwidth of M channels satisfies:

f _Lk =f _L +(2k+1-M)BW/2M

Among them, k=0,1,2,...,M-1, M is an even number; BW is the demodulation bandwidth;

S2. Set the cut-off frequency of the low-pass filter to BW/2M, the ADC sampling rate is f _s and f _s ≥ BW;

In this embodiment, the frequency range of f _L ±BW/2 needs to be directly down-converted to the baseband [-BW/2, BW/2]. Due to the limitation of the demodulation bandwidth of the chip, the multi-channel quadrature as shown in Fig. 4 can be used. The demodulation structure is to divide the entire demodulation range f _L ±BW/2 into M equal parts, that is, the bandwidth of the demodulation BW/M of each channel, and the frequency range of the original RF signal corresponding to the demodulation of each channel is f _Lk ± BW/2M, it is necessary to properly set the frequency f _Lk of each RF local oscillator signal shown in FIG. 4 . According to the above requirements, the relationship between the RF local oscillator signal frequency f _Lk of each channel and the system center frequency f _L (that is, the RF local oscillator frequency of single channel demodulation) can be expressed as:

f _Lk =f _L +(2k+1-M)BW/2M

Among them, k=0,1,2,...,M-1.

For example, when M=4, it can be seen from the above formula that the frequency of each local oscillator signal can be expressed as f _Lk =f _L +(2k-3)BW/8(k=0,1,2,3), and 4 The corresponding RF signal frequency ranges of the demodulator are [f _L -BW/2,f _L -BW/4], [f _L -BW/4,f _L ], [f _L ,f _L +BW/4 ] and [f _L +BW/4,f _L +BW/2], that is, the total frequency range is [f _L -BW/2,f _L +BW/2], which is exactly the center frequency required by the system is f _L , the demodulation frequency range where the demodulation bandwidth is BW.

S3. Input the input signal x(t) to M channels of narrow demodulation bandwidth quadrature demodulation chips, and perform quadrature demodulation with M channels of RF local oscillator signals x _Lk (t)=exp(j2πf _Lk t), and then The demodulated signal of each channel is filtered by a low-pass filter, so that each channel of low-pass filter outputs an analog baseband complex signal x _k (t) with a frequency range of [-BW/2M, BW/2M];

S4. Perform dual-channel sampling on each channel of analog baseband complex signal x _k (t) through ADC to obtain M channels of digital baseband complex signal x _k (n);

In this embodiment, as shown in FIG. 4 , after each demodulation channel completes the quadrature demodulation of the frequency range f _Lk ±BW/2M in the corresponding original radio frequency signal, the average output frequency range is [-BW/2M ,BW/2M] analog baseband complex signal x _k (t), and then sampled by the ADC (each channel is dual-channel sampling, that is, the two ADCs sample the real and imaginary parts of the complex signal synchronously, and the sampling rate is f _s and f _s ≥ BW) yields a digital baseband complex signal x _k (n).

S5. The effective bandwidth of the digital baseband complex signal x _k (n) is BW/M, and the effective bandwidth required by the actual system is BW, so it is necessary to combine M baseband complex signals with an effective bandwidth of BW/M inside the FPGA In order to realize the expansion of the demodulation bandwidth, the specific process is as follows:

Perform digital complex mixing on each channel of digital baseband complex signal x _k (n) and the corresponding digital replica oscillator y _Lk (n)=exp(jω _Lk n), and then combine the real part and imaginary part of the complex mixing output signal Correspondingly add up to obtain a digital complex signal with a sampling rate of f _s and an effective signal bandwidth of BW;

Among them, the frequency of the digital replica oscillator ω _Lk satisfies:

ω _Lk =2π(2k+1-M)BW/2Mf _s

=π(2k+1-M)BW/Mf _s

The purpose of complex mixing is to restore the frequency range [-BW/2M, BW/2M] corresponding to x _k (n) to a wide demodulation bandwidth for demodulating the direct input signal (the local oscillator is f _L , the demodulation range is f _L ±BW/2) and sampling the demodulated output baseband signal with f _s to obtain the frequency range, namely [(2k+M)BW/2M, (2k+2-M)BW/2M], the bandwidth is BW/M . By adding up multiple digital mixing results, a digital complex signal with a sampling rate of f _s and an effective signal bandwidth of BW can be obtained.

Based on the above analysis, the present invention uses multiple orthogonal demodulation chips with narrow demodulation bandwidth to realize the expansion of demodulation bandwidth in parallel demodulation, which greatly reduces the requirement for the demodulation bandwidth of the orthogonal demodulation chip.

Next, the present invention is verified by taking the parallel demodulation of two channels (M=2) as an example and combining with Simulink simulation.

As shown in Figure 5, a two-channel parallel demodulation structure is adopted, that is, two-channel 400Hz demodulation bandwidth quadrature demodulation realizes 800Hz quadrature demodulation bandwidth. Set the system local oscillator to 2200Hz, the demodulation local oscillator frequency of the two sub-channels is (2200+200(2k-1))Hz(k=0, 1), that is, the demodulation local oscillator frequency of channel 0 is 2.0kHz, channel 1 The demodulated local oscillator is 2.4kHz. The effective bandwidth of demodulation output of both channels is [-200Hz, 200Hz]. Due to the different local oscillator signals, the input signal frequency ranges processed by the two quadrature demodulation channels are also different, but the bandwidths are both 400Hz, that is, the processing frequency range of channel 0 is [1800, 2200] Hz, and the processing frequency of channel 1 is [1800, 2200] Hz. The range is [2200, 2600] Hz. The total processing frequency range of the system is [1800, 2600] Hz, which is exactly the frequency range processed by the single-channel quadrature demodulation corresponding to the local oscillator of 2.2 kHz and the demodulation bandwidth of 800 Hz.

The cutoff frequencies of the four sub-signals in the two baseband signals are all 200Hz, and the subsequent ADC sampling rate theoretically only needs to satisfy the Nyquist sampling theorem, such as 500Hz sampling rate. However, since the final demodulation bandwidth is 800Hz, that is, the real and imaginary cut-off frequencies of the complex signals obtained by the final two-channel synthesis are both 400Hz. If the ADC adopts 500Hz, before the two channels of digital baseband signals are synthesized in the FPGA , it needs to be double-interpolated first, otherwise the 800Hz demodulation result cannot be obtained. For simplicity, the ADC sampling rate is set to 1kHz here.

After the sampling is completed, the baseband signal of channel 0 is subjected to complex mixing with a digital local oscillator frequency of 2π(-200)/1000, and its frequency range is moved from [-200, 200] Hz to [-400, 0] Hz ; For channel 1 baseband signal, realize complex mixing with digital local oscillator frequency of 2π(200)/1000, and move its frequency range from [-200, 200] Hz to [0, 400] Hz. The 800Hz demodulation result in the frequency range [-400, 400]Hz can be obtained by adding the two digital complex mixing results.

Set the input signal frequency to 2.6kHz, perform quadrature demodulation with the local oscillator of 2.2kHz and the demodulation bandwidth of 800Hz, and the obtained signal frequency should be 400Hz. Figure 6 shows the output of channel 0 demodulation results. Since 2.6kHz is outside the demodulation range of channel 0, there is no signal output from channel 0. Figure 7 shows the demodulation result output of channel 1. Since 2.6kHz is within the demodulation range of channel 1, and the local oscillator frequency of channel 1 is 2.4kHz, the frequency of the complex signal output by channel 1 demodulation is 200Hz. Figure 8 shows the output result of channel 0 after complex mixing with digital local oscillator frequency 2π(-200)/1000, and its effective frequency range is moved to [-400,0]Hz. Figure 9 shows the output result of channel 1 after complex mixing with the digital local oscillator frequency 2π(200)/1000. The effective frequency range is moved to [0, 400]Hz, and the input frequency of 2.6kHz corresponds to a frequency of 400Hz. Figure 10 is the signal obtained by adding the two baseband signals after complex mixing, that is, the total quadrature demodulation output of the system. It can be seen that the effective frequency range is [-400, 400]Hz, and the demodulation output corresponding to the input signal 2.6kHz The frequency is 400Hz.

Figure 11 shows the system demodulation output spectrum when the input signal frequency is 1.8kHz, and the output frequency is 1.8kHz-2.2kHz=-400Hz.

Figure 12 shows the system demodulation output spectrum when the input signal frequency is 2.1kHz, and the output frequency is 2.1kHz-2.2kHz=-100Hz.

Figure 13 shows the system demodulation output spectrum when the input signal frequency is 2.4kHz, and the output frequency is 2.4kHz-2.2kHz=200Hz.

In summary, the system can achieve demodulation under different input frequency signals, and improve the demodulation bandwidth.

Although illustrative specific embodiments of the present invention have been described above to facilitate understanding of the present invention by those skilled in the art, it should be clear that the present invention is not limited to the scope of the specific embodiments. For those skilled in the art, As long as various changes are within the spirit and scope of the present invention as defined and determined by the appended claims, these changes are obvious, and all inventions and creations utilizing the inventive concept are included in the protection list.

Claims (1)

1. a method for improving demodulation broadband by multi-channel parallel segmented demodulation mode, is characterized in that, comprises the following steps: (1), the frequency f _Lk of each radio frequency local oscillator signal when setting quadrature demodulation; Suppose the center frequency range of the input signal x(t) is f _L ±BW/2, and the entire demodulation range f _L ±BW/2 is equally divided into M equal parts, that is, the bandwidth of the demodulation BW/M of each channel, then The RF local oscillator signal frequency f _Lk input to the quadrature demodulation chip with narrow demodulation bandwidth of M channels satisfies: f _Lk =f _L +(2k+1-M)BW/2M Among them, k=0,1,2,...,M-1; BW is the demodulation bandwidth; (2), set the cut-off frequency of the low-pass filter to BW/2M, the ADC sampling rate is f _s and f _s ≥ BW; (3) Input the input signal x(t) to the quadrature demodulation chip with M channels of narrow demodulation bandwidth, and perform quadrature demodulation with M channels of RF local oscillator signal x _Lk (t)=exp(j2πf _Lk t) , and then pass the demodulated signal of each channel through the filtering process of the low-pass filter, so that each channel of low-pass filter outputs the analog baseband complex signal x _k (t ); (4), carry out dual-channel sampling by ADC for each channel of analog baseband complex signal x _k (t) to obtain M channels of digital baseband complex signal x _k (n); (5) Perform digital complex mixing of each channel of digital baseband complex signal x _k (n) and the corresponding digital replica oscillator y _Lk (n)=exp(jω _Lk n), and then combine the real output signal of the complex mixing frequency. The part and the imaginary part are added correspondingly to obtain a digital complex signal with a sampling rate of f _s and an effective signal bandwidth of BW; Among them, the frequency of the digital replica oscillator ω _Lk satisfies:

Images