CN100483963C - Orthogonal lower mixing frequency digital template matching pulse ultra wide band radio signal receiving method - Google Patents

Abstract

The invention belongs to the technical field of receiving wireless pulse signals, and is characterized in that it combines the technology of quadrature down-mixing and digital template matching, and adopts an analog-to-digital converter (ADC) with a high sampling rate after quadrature down-mixing. In the digital template matching, a flexible matching correlation template that can be adaptively adjusted and has different offset positions is used, so it has the following advantages: it can better adapt to various possible waveforms of the pulse in the chip after passing through the channel; The high baseband bandwidth is conducive to fast and high-precision frequency deviation correction; the hardware structure and implementation cost of the Rake receiver are simplified; at the same time, it supports pulse radio and UWB signal reception with carrier schemes, and has strong versatility.

Description

Pulse ultra-wideband wireless signal receiving method based on quadrature down-mixing and digital template matching

technical field

The invention relates to a pulse ultra-wideband wireless signal receiving method for orthogonal down-mixing digital template matching, and simultaneously supports pulse radio (IR) or carrier-modulated ultra-wideband pulse signals on the physical layer of a wireless personal area network The high-performance receiving and demodulation can carry out simple and practical circuit design and realization, and it is easy to integrate chips, which belongs to the technical field of wireless pulse signal reception.

Background technique

Narrow pulse ultra-wideband technology is a wireless communication technology that uses nanosecond-level non-sine wave narrow pulses to transmit data. Its signal spectrum range is very wide, usually above 500MHz, and it has strong confidentiality, strong anti-interference ability and high transmission rate. Advanced unique advantages. In order to avoid interference to other civilian equipment, FCC (Federal Communications Commission of the United States) stipulates that UWB (ultra-wideband) must work in the 3.1-10.6GHz frequency band, and the transmission power should not exceed -41.3dBm/Hz. For example, the MBOK-UWB (M-element Biorthogonal Keying-Ultra Wideband) scheme we proposed is based on the principle of narrow pulse radio technology. Its RF front-end uses discontinuous pulse signals as the medium for transmitting information. The designed signal spectrum In the range of 3.1-4.6GHz, the width exceeds 1GHz. At present, there are still some problems in the practical application of ultra-wideband wireless technology, mainly due to insufficient research on channel models, high cost and power consumption of transceiver chips, and technical standards such as 802.15.3a have not yet been finalized.

One of the traditional narrow-pulse ultra-wideband reception methods uses pure analog Rake correlation technology, that is, during the Rake reception of multipath signals, a series of analog signal templates with different delays are used to perform analog correlation with the antenna received signal, and then the correlation results Perform digital sampling and combining. This method requires high precision in the design of the analog delay line structure, and it is difficult to implement the chip, especially in the case of extremely high chip and data rates, and the performance is difficult to guarantee; research shows that the delay of each analog signal In the case of large time jitter, the accumulation of analog-related errors will affect the performance of reception.

Another traditional receiving method uses pure digital correlation technology, that is, AD sampling is performed on the signal at a very high sampling rate at the radio frequency, and then digital down conversion (DDC), correlation reception and other processing are performed. For high-speed data applications, this method has extremely high requirements on the sampling rate of the analog-to-digital converter (ADC), requires a large hardware implementation cost, and brings huge power consumption. Therefore, this kind of scheme also has great limitations for high-speed data applications.

Contents of the invention

The purpose of the present invention is to propose a pulse ultra-wideband wireless signal receiving method for orthogonal down-mixing digital template matching, which adopts the combination of I/Q two-way analog frequency mixing and digital matching related Rake reception, which can be simple and practical The circuit design and implementation are easy for chip integration; because it supports the reception and demodulation of UWB pulse signals in two forms of pulse radio and carrier, it greatly enhances its versatility for different signal transmission schemes.

The present invention is characterized in that the method comprises the following steps in sequence:

Step 1 receives the signal and obtains the zero-IF analog baseband signal through spectrum shifting after being sequentially composed of an antenna, a low-noise amplifier, an I/Q two-way analog mixer and a low-pass filter in series. The frequency of the local oscillator signal of the /Q two-way analog mixer is determined in the following way: according to the piconet architecture formulated by the ultra-wideband communication system scheme and the center frequency point selected by the transmitting user, a sine wave signal of the same frequency is selected as the The local oscillator signal of the mixer, the I/Q two-way local oscillator signal maintains orthogonality;

Step 2: The zero-IF analog baseband signal output by the analog front-end is sent to an ADC whose sampling rate is higher than twice the chip rate to sample the zero-IF baseband signal to generate I/Q two-way digital baseband signals;

Step 3 sends the I/Q two-way high-speed digital baseband signal that step 2 obtains into a digital matching correlation Rake receiver integrated on a digital integrated circuit chip, and in the digital domain, perform the following steps on the I/Q two-way high-speed digital The baseband signal performs frequency and timing estimation and correction, digital matching correlation, Rake reception and frame synchronization capture, and the obtained output signal is sequentially despreaded by spreading code and channel decoded to restore the original transmitted signal:

Step 3.1 Use a frequency correction circuit to perform frequency correction on the input I/Q two-way high-speed digital baseband signals according to the following steps:

Step 3.1.1 Send I/Q two-way digital baseband signals I'(n), Q'(n) to the phase detector in the carrier recovery loop including the polar type carrier recovery loop, according to the following algorithm Extract the phase error signal ε _c (n):

ε _c (n)=I'(n)×Q'(n), n is the sampling point label;

Step 3.1.2 Input the phase error signal ε _c (n) into the second-order digital loop filter in the carrier recovery loop for filtering, and its timing is:

The phase error accumulation signal φ(n+1) of the n+1 sampling point is:

φ(n+1)=φ(n)+K ₁ f(n)+K ₂ ε _c (n),

The frequency error accumulation signal f(n+1) of the n+1 sampling point is:

f(n+1)=f(n)+ _εc (n),

Wherein, constant K ₁ and K ₂ are the filter coefficients of this second-order digital loop filter, when the sampling rate f of the ADC _is much greater than the natural frequency of the loop:

K ₂ =2ηω _n T _s $K_1={\mathrm{<figure-callout\; id="2"\; label="K"\; filenames="C200610165248E00192.png"\; state="\{\{state\}\}">K</figure-callout>}}_2^2/4{\mathrm{\&eta;}}^2={\mathrm{\&omega;}}_{\mathrm{no}}{T}_{\mathrm{the\; s}},$

ω _n =2πf _n is the natural angular frequency of the loop, η is the damping coefficient, T _s =1/f _s ;

Step 3.1.3 takes the phase error cumulative signal φ(n+1) obtained through the loop filtering as the input of the local digitally controlled oscillator in the loop to adjust the frequency of the local oscillator signal;

Step 3.1.4 Multiply the I/Q two-way digital baseband signals I(m), Q(m) input to the frequency correction loop with the output cosφ(n) and sinφ(n) of the local numerically controlled oscillator, and perform phase Rotate to get frequency-corrected I/Q two-way digital baseband signals I'(m), Q'(m):

I'(m)=I(m)×cosφ(n)+Q(m)×sinφ(n)

Q'(m)=Q(m))×cosφ(n)-I(m)×sinφ(n)

The difference between m and n is the operation delay of the entire carrier recovery loop;

Step 3.2 uses a timing correction loop to correct the timing of the output signal of the frequency correction circuit according to the following steps:

Step 3.2.1 utilizes the main path position output by the I channel digital baseband signal I'(n) through frequency correction and the channel estimator to obtain the main path signal;

Step 3.2.2 inputs the main path signal into a main path correlator to obtain the main path correlation value Corr(n);

Step 3.2.3 sends the local template signal and the main path signal into an early correlator to obtain an early path correlation value Corr(n-1), which is the local template signal and one unit earlier than the main path signal The cross-correlation value of the signal at the sampling interval, the local template signal is generated by the local template signal generator, its initial value is the standard signal transmitted by the transmitter in a known single chip, and the phase delay or advance is carried out according to the template adjustment signal;

Step 3.2.4 sends the local template signal described in step 3.2.3 and the main path signal into a late correlator to obtain a late path correlation value Corr(n+1), which is the local template signal The cross-correlation value of the signal with a unit sampling interval later than the main path signal;

Step 3.2.5 Send the obtained main path correlation value Corr(n), early path correlation value Corr(n-1), and late path correlation value Corr(n+1) into a timing error extractor, and calculate according to the following formula Timing error signal ε _t (n):

ε _t (n)=sign(Corr(n))*[Corr(n+1)-Corr(n-1)]:

Step 3.2.6 sends the timing error signal ε _t (n) obtained in step 3.2.5 into the first-order digital low-pass loop filter to obtain the accumulated timing error value:

The timing is: E _t (n) = E _t (n-1) + K ₁ ε _t (n),

The constant K ₁ is the filter coefficient of the filter, K ₁ <1;

Step 3.27 uses a comparator to judge whether the cumulative error value obtained in step 3.2.6 reaches the set threshold, and if it reaches the set threshold, the cumulative error value is cleared, and a template adjustment signal is sent at the same time to control the generation of the local template signal The device delays or advances the phase of the local template signal by one unit sampling interval, and outputs the template adjustment signal to the subsequent digital matching related Rake receiving circuit to achieve timing correction; the advance or delay depends on the polarity of the accumulated error, if it is positive, then advance, otherwise, delay;

Step 3.3 sends the template adjustment signal obtained in step 3.2.7 and the I channel digital baseband signal I'(n) through frequency correction into a digital matching correlation and Rake receiving circuit, and calculates the output of the Rake correlation receiving circuit according to the following steps :

Step 3.3.1 Send the digital baseband signal and several signal templates with different offset positions output by the local template signal generator to the channel estimator, and search for a main path position corresponding to the maximum correlation value and The secondary path positions corresponding to several next largest correlation values successively, the correlation operation refers to: calculate the product of two input signals by each sampling point and accumulate, the obtaining process of the main path position is the process of symbol synchronization; step 3.2. 7. The obtained template adjustment signal adjusts the template signals of the main path and selected secondary paths within the time interval of two consecutive channel estimations;

Step 3.3.2 According to the positions of the main path and each secondary path obtained in step 3.3.1, the frequency-corrected I channel digital baseband signal I'(n) and the template corresponding to the obtained main path and each selected secondary path The signals are respectively sent to a correlator group, and the correlation operation is carried out respectively according to the method described in step 3.3.1 to obtain the digitally matched signal, which is sent to the Rake combiner;

Step 3.3.3 The Rake combiner combines at least one secondary path according to the set plan, and there is no upper limit;

Step 3.3.4 Rake combiner The multipath energy formed by the correlation results corresponding to the main path and each secondary path is weighted and accumulated according to the maximum combination ratio criterion or the equal gain combination ratio criterion to obtain a total energy, that is, the Rake correlation receiving circuit The output of I"(n);

Step 4 sends the output signal I"(n) of the Rake correlation reception that step 3.3.4 obtains to a frame synchronization capture device, and outputs the Rake correlation reception signal after the frame synchronization according to the following steps:

Step 4.1 Send the output signal I"(n) obtained in step 3.3.4 to a frame synchronization circuit, use the polarity of the signal as the information data bit, and search for the output corresponding to the information data bit to find the frame separator. So as to locate the frame header and locate the starting position of the data information;

Step 4.2 Send the result I"'(n) obtained in step 4.1 to the subsequent spreading code despreading module and channel decoder.

The pulse ultra-wideband wireless signal receiving method of quadrature down-mixing digital template matching proposed by the present invention mainly includes: combining quadrature down-mixing and digital template matching, thereby avoiding traditional analog correlation and Rake receiver Requirements for high-precision analog delay line structure design, while avoiding the requirements of traditional digital correlation and Rake receivers for high sampling rate ADCs, thereby simplifying the hardware structure and implementation costs, and improving the reliability of process implementation; high multiplier in the digital domain Oversampling combined with flexible matching correlation templates can better adapt to various possible waveforms of intra-chip pulses passing through the channel; it has a high baseband bandwidth, which is conducive to fast and high-precision frequency deviation correction; it also supports pulsed radio and The UWB signal reception with the carrier scheme has strong versatility; it has simple and practical circuit design and implementation with regular structure, and is easy to integrate chips.

Description of drawings:

Fig. 1 is a schematic block diagram of a pulse ultra-wideband wireless receiving method for quadrature down-mixing digital template matching.

Fig. 2 is a schematic circuit block diagram of a frequency correction loop in the pulse ultra-wideband wireless receiving method of quadrature down-mixing digital template matching.

Fig. 3 is a schematic block diagram of a timing correction loop in the pulse ultra-wideband wireless receiving method of quadrature down-mixing digital template matching.

Fig. 4 is a block diagram of digital matching correlation and Rake receiving circuit in the pulse ultra-wideband wireless receiving method of quadrature down-mixing digital template matching.

Fig. 5 is a block diagram of the circuit principle of frame synchronization capture in the pulse ultra-wideband wireless receiving method of quadrature down-mixing digital template matching.

Detailed ways:

The pulse ultra-wideband wireless signal receiving method of quadrature down-mixing digital template matching that the present invention proposes, its corresponding receiver is made up of radio frequency front-end, zero-IF baseband sampling circuit, digital matching correlation and Rake receiving three major parts (serial structure):

(1) The RF front-end is composed of antenna, low noise amplifier (LNA), I/Q two-way analog mixer and low-pass filter. The received signal is amplified at this stage and I/Q two-way analog down-mixing , low-pass filtering, etc., to obtain a zero-IF analog baseband signal after spectrum shifting;

(2) The zero-IF analog baseband signal output by the RF front-end enters the zero-IF baseband sampling circuit; the I/Q two-way zero-IF analog baseband signal is sent to an ADC with a sampling rate higher than twice the chip rate for sampling and quantization to generate I/ Q two digital baseband signals;

(3) The high-speed digital baseband signal output by the sampling circuit is sent to the digital matching correlation Rake receiver, and the signal is estimated and corrected in the digital domain for frequency and timing, digital matching correlation and Rake reception, frame synchronization capture, and the obtained output The signal is despread by spreading code and channel decoded sequentially to recover the original transmitted signal.

In the above structure, the method for determining the frequency of the local oscillator signal of the I/Q two-way analog mixer is:

According to the piconet architecture formulated by the ultra-wideband communication system scheme and the center frequency point selected by the transmitting user, the sine wave signal of the same frequency is selected as the local oscillator signal of the mixer to realize the spectrum shift of the signal and obtain the baseband signal. The I/Q local oscillator signals should maintain orthogonality.

In the above structure, the method for determining the sampling rate of the ADC is:

The sampling rate of the ADC should be much higher than the chip rate. The selection of the specific multiple should consider the characteristics of the transmitted pulse waveform (duty cycle, polarity change within a chip, etc.), for example, 4 times or 8 times. In this solution, the sampling rate of the ADC also determines the time resolution accuracy of the multipath Rake receiver.

For example, the chip rate is 500Mbps, and when the sampling rate of the ADC is 4 times the chip rate, the ADC sampling rate is required to be about 2Gsps, and the time resolution accuracy of the diameter is 1/(2Gsps)=0.5 nanoseconds.

In the above structure, the digital matching correlation Rake receiver can be implemented by FPGA (Field Programmable Gate Array), and can also be implemented on ASIC (Application Specific Integrated Circuit).

Firstly, the communication terminology involved in this method is given, and its explanation is as follows:

1) Chip: the unit signal used to represent information in digital communication, usually in the form of a standard pulse waveform or a standard digital signal sequence; each information bit or information symbol can be represented by several specially arranged chips;

2) Chip rate/symbol rate: The chip rate is usually several times the symbol rate, and the specific multiple is equal to the number of chips used to represent each symbol;

3) Digital matching correlation Rake reception: In digital communication, in order to maximize the receiving signal-to-noise ratio, matching correlation is usually used for signal reception processing; at the same time, in order to solve the energy spread caused by multi-path channels, a Rake structure (multi-path Parallel matching and correlation and weighted combination of their results) to efficiently collect multipath energy to improve demodulation performance;

4) Estimation of frequency and timing: using the received signal sequence, through the operation and processing of the corresponding algorithm, the relative frequency and timing deviation information of the received signal and the transmitted signal are obtained, so as to effectively correct the frequency and timing;

5) Synchronous capture of symbols and frames: The process of capturing the starting position of each symbol by the receiving end using an algorithm to capture the starting position of each symbol is the synchronous capturing of symbols; the receiving end uses an algorithm to capture the starting position of each signal frame The process of the start position is frame synchronous capture; the start position of the frame is marked by the frame separator;

6) Spreading code despreading: Spreading communication uses the polarity ('0' or '1') of multiple chips to represent an information bit or information symbol, which indicates that the sequence is a spreading code; The process of determining which information bit or information symbol is sent is the despreading of the spreading code;

7) Channel decoding: In digital communication, forward error correction channel coding is often performed on the transmitted information bits to increase redundant information to increase the reliability of communication; the corresponding operation at the receiving end (ie, removing redundancy and correcting errors) is called channel decoding;

8) The time resolution accuracy of the path: the minimum measure of the delay between each path when Rake receives multipath;

9) Baseband signal: signal without carrier modulation in digital communication;

10) Analog down-mixing: the process of multiplying the received analog signal with the local oscillator signal and low-pass filtering; this process realizes the process of shifting the frequency spectrum of the signal from the radio frequency to the baseband;

11) The local oscillator signal of the down-mixer: usually a single-frequency sinusoidal signal, and its frequency setting must be consistent with the center frequency of the transmitted signal;

12) Carrier recovery loop: The frequency estimation and correction algorithm is usually implemented in the form of a loop, which is called a carrier recovery loop; the loop is usually composed of a phase detector, a loop filter, an NCO (numerically controlled oscillator) )composition;

13) Phase detector: The circuit that uses the input signal to extract the phase error signal in the carrier recovery loop is called a phase detector;

14) Loop natural frequency: the inherent parameters of the carrier recovery loop, determined by parameters such as loop gain;

15) Correlator (correlation operation): realizes the corresponding multiplication and accumulation operation between the input digital signal and a group of digital signals equal in length; the latter is called the local template signal;

For example, the input signal is x(n), n=1, 2,...N, and the local template signal is y(n), n=1, 2,...N, then the output result of the correlator (correlation operation) (also known as mutual Correlation value) is: $\mathrm{Corr}=\underset{\mathrm{no}=1}{\overset{N}{\mathrm{\&Sigma;}}}x\left(\mathrm{no}\right)\&CenterDot;\mathrm{the\; y}\left(\mathrm{no}\right);$

16) Main diameter signal: through the correlation calculation between the input signal and the signal templates at different offset positions, and selecting the maximum value, the position of the main diameter of the signal can be obtained, thereby obtaining the main diameter signal;

17) Main path correlator: a correlator whose input signal is the main path signal; its output is called the main path correlation value;

18) Channel estimation: use the known transmission preamble sequence to train the receiver, perform correlation calculations between the input signal and signal templates at different offset positions, and select several maximum values to obtain several paths of the signal that need to be combined The position and gain of (main path and each secondary path); the information of these positions and gains is the result of channel estimation;

19) Multipath energy: The correlation results corresponding to the main path and each secondary path given by channel estimation are its multipath energy;

20) Combination of multipath energy: the multipath energy is weighted and accumulated by a certain criterion to obtain a total energy;

21) Criteria for the maximum combination ratio: the correlation value of the main path and each secondary path determines the weighted value of its energy combination; the path with greater energy has a greater combination weight (proportional relationship);

22) Equal-gain combination criterion: the gains of the main path and each secondary path are equal when energy is combined.

The following describes the content of the present invention in detail in conjunction with the accompanying drawings; when using the above-mentioned communication terminology, please refer to the above explanation if necessary:

Fig. 1 is a schematic block diagram of a pulse ultra-wideband wireless receiving method for quadrature down-mixing digital template matching. As shown in Figure 1, the signal is received by the antenna, amplified by the low-noise amplifier (LNA), and then mixed with the locally generated I/Q two-way local oscillator signal for analog down-mixing, and then low-pass filtered. The method of selecting the frequency of the local oscillator signal is as follows: according to the piconet architecture formulated by the ultra-wideband communication system scheme and the center frequency point selected by the transmitting user, a sine wave signal of the same frequency is selected as the local oscillator signal of the mixer to realize signal transmission. The frequency spectrum is shifted to obtain the baseband signal.

If the pulsed radio scheme is used when the signal is transmitted, the down-mixing operation is equivalent to spectrum shifting of the band-pass RF signal, and the corresponding baseband signal is obtained after filtering; if the signal is transmitted with a carrier modulation scheme, the down-mixing and low The combined effect of the pass filter is equivalent to shifting the spectrum of the signal modulated to the center frequency, and down-converting it to a baseband signal. Therefore, whether it is an IR or a carrier solution, after the RF front-end processing, the baseband data signal will be obtained and provided to the subsequent stage for processing.

The analog baseband signal output by the RF front-end is sampled by the ADC to obtain the baseband signal in the digital domain; the ADC sampling rate should be higher than the Nyquist sampling frequency to facilitate matching and correlation during digital baseband processing.

The sampled digital signal enters the baseband digital matching correlation Rake receiver for processing, and performs frequency and timing estimation and correction, digital matching correlation, Rake combination, and frame synchronization capture on the signal in the digital domain, and spreads the obtained output signal in sequence Code despreading and channel decoding to restore the original transmitted signal.

The processing flow of the digital Rake receiver is: the signal obtained by sampling needs to first estimate and correct the frequency and timing. Since the frequency of the local oscillator signal used by the down-mixer may deviate from the center frequency of the transmitted signal, the obtained baseband signal is not necessarily strictly zero-IF, and frequency correction is required. The speed of frequency correction is closely related to the baseband bandwidth. The higher the baseband bandwidth, the faster the relative speed of the frequency correction algorithm (correction can be realized in a small number of chips); and the baseband bandwidth of the signal corresponding to this structure is very high, so there is Useful for frequency correction. At the same time, the deviation of the reference clock crystal oscillator of the transmitter and the receiver will introduce timing deviation of sampling, which should also be corrected. The signal after frequency and timing correction enters the digital matching related Rake receiving module; by adaptively adjusting the digital matching template, it can handle the signal waveform changes caused by channel changes. According to the multipath distribution of channel estimation, the correlation calculation results of the largest and second largest receiving paths are sent to the Rake combiner to combine the energy of multipath signals to obtain effective diversity gain and improve the equivalent signal-to-noise ratio. The signal output by the digital matching related Rake receiving module needs to be captured synchronously with the frame. The frame-synchronized signal enters the spreading code despreading module and the channel decoder to perform spreading code despreading and channel decoding, thereby recovering the original sent data information.

The following is the algorithm description of each part and its specific implementation method:

Fig. 2 is a schematic circuit block diagram of a frequency correction loop in the pulse ultra-wideband wireless receiving method of quadrature down-mixing digital template matching. As shown in Figure 2, the frequency correction algorithm of this method can use a standard carrier recovery loop, and the specific implementation method is:

1) Send I/Q two-way digital baseband signals I′(n), Q′(n) to the phase detector of the carrier recovery loop, and extract the phase error signal ε _c (n) (the polarity can also be used here Type of carrier recovery loop, simplify the error extraction method); its extraction method is:

ε _c (n)=I'(n)×Q'(n), where n is the label of the sampling point;

2) The phase error signal ε _c (n) is input to the second-order digital loop filter in the carrier recovery loop for filtering, and its timing is:

The phase error accumulation signal φ(n+1) of the n+1 sampling point is:

φ(n+1)=φ(n)+K ₁ f(n)+K ₂ ε _c (n),

The frequency error accumulation signal f(n+1) of the n+1 sampling point is:

f(n+1)=f(n)+ _εc (n)

Among them, the constants K ₁ and K ₂ are the filter coefficients of the second-order digital loop filter. When the sampling rate f _s of the ADC is much greater than the natural frequency of the loop, these two parameters can be determined by the following formula:

K ₂ =2ηω _n T _s $K_1={\mathrm{<figure-callout\; id="2"\; label="K"\; filenames="C200610165248E00192.png"\; state="\{\{state\}\}">K</figure-callout>}}_2^2/4{\mathrm{\&eta;}}^2={\mathrm{\&omega;}}_{\mathrm{no}}{T}_{\mathrm{the\; s}},$

Among them, ω _n =2πf _n is the natural angular frequency of the loop, η is the damping coefficient, T _s =1/f _s ;

3) The phase error cumulative signal φ(n+1) obtained through loop filtering is used as the input of the local numerically controlled oscillator in the loop to adjust the frequency of the local oscillator signal;

4) Multiply the I/Q two-way digital baseband signals I(m), Q(m) input to the frequency correction loop with the output cosφ(n) and sinφ(n) of the local numerically controlled oscillator to perform phase rotation, Thus, the frequency-corrected I/Q two-way digital baseband signals I'(m), Q'(m) are obtained:

I'(m)=I(m)×cosφ(n)+Q(m)×sinφ(n)

Q'(m)=Q(m)×cosφ(n)-I(m)×sinφ(n)

The difference between m and n is the operation delay of the entire carrier recovery loop (related to the specific hardware implementation; since the phase rotation is a continuous process, there is no strict constraint on the operation delay).

Fig. 3 is a schematic block diagram of a timing correction loop in the pulse ultra-wideband wireless receiving method of quadrature down-mixing digital template matching. As shown in Figure 3, the timing correction algorithm of this method can use the "sooner or later ring" related method, and the specific implementation method is:

1) Utilize the I channel digital baseband signal I'(n) through frequency correction and the main path position output by the channel estimator to obtain the main path signal;

2) Input the main path signal into a main path correlator to obtain the main path correlation value Corr(n);

3) Send the local template signal and the main path signal into an early correlator to obtain an early path correlation value Corr(n-1), which is the signal of the local template signal and a unit sampling interval earlier than the main path signal The cross-correlation value of , the local template signal is generated by the local template signal generator, its initial value is the standard signal transmitted by the transmitter in a known single chip, and the phase delay or advance is carried out according to the template adjustment signal;

4) Send the local template signal and the main path signal described in 3) into a late correlator to obtain a late path correlation value Corr(n+1), which is the local template signal and the main path signal later than the main path signal. The cross-correlation value of a signal with a unit sampling interval;

5) Send the obtained main path correlation value Corr(n), early path correlation value Corr(n-1), and late path correlation value Corr(n+1) into a timing error extractor, and calculate the timing error according to the following formula Signal ε _t (n):

_εt (n)=sign(Corr(n))*[Corr(n+1)-Corr(n-1)];

6) Send the timing error signal ε _t (n) obtained in 5) into the first-order digital low-pass loop filter to obtain the accumulated timing error value: its timing is:

E _t (n) = E _t (n-1) + K ₁ ε _t (n);

Among them, the constant K ₁ is the filter coefficient, which is set by the system plan (usually no more than 1);

7) Use a comparator to judge whether the error obtained in 6) reaches the set threshold (set by the system plan, such as 1/2 of the main path correlation value), and if it reaches the set threshold, clear the accumulated error value , at the same time send a template adjustment signal to control the local template signal generator to delay or advance the phase of the local template signal by one unit sampling interval, and output the template adjustment signal to the subsequent digital matching related Rake receiving circuit to achieve timing correction;

Take advance or delay depending on the polarity of the accumulated error, if positive, advance, otherwise, delay;

Fig. 4 is a block diagram of digital matching correlation and Rake receiving circuit in the pulse ultra-wideband wireless receiving method of quadrature down-mixing digital template matching. As shown in Figure 4, the specific implementation methods of channel estimation, digital matching correlation and Rake merging in this method are as follows:

1) Send the frequency-corrected I channel digital baseband signal I'(n) and several signal templates with different offset positions output by the local template signal generator into the channel estimator, and search for a channel estimator with the maximum The main diameter position corresponding to the correlation value and the secondary diameter position corresponding to several second largest correlation values in turn, the process of obtaining the main diameter position is the process of symbol synchronization;

2) According to the positions of the main path and each secondary path obtained in 1), the frequency-corrected I channel digital baseband signal I'(n) and the template signal corresponding to the obtained main path and each selected secondary path are respectively sent into the A group of correlators, which respectively perform digital matching correlation operations, obtain digitally matched signals, and send them to the Rake combiner;

3) The selection of the number of combined secondary paths of the Rake combiner is set by the system plan, at least one, and no upper limit;

4) The multipath energy formed by the Rake combiner from the correlation results corresponding to the main path and each secondary path is weighted and accumulated according to the maximum combination ratio criterion or the equal gain combination ratio criterion to obtain a total energy, that is, the Rake correlation receiving circuit The output I"(n), which combining ratio criterion to choose is set by the system plan, and the receiving performance under the multipath channel is greatly improved by combining the multipath energy;

The sampling rate of the front-stage ADC is higher than twice the chip rate, and the time resolution accuracy of the Rake receiver is determined by the sampling rate of the ADC. For example, the chip rate is 500Mbps, and when the sampling rate of the ADC is 4 times the chip rate, the ADC sampling rate is required to be about 2Gsps, and the time resolution accuracy of the diameter is 1/(2Gsps)=0.5 nanoseconds.

As shown in Fig. 4, according to the result of channel estimation, this method flexibly adjusts the local template signal used in the digital matching correlation operation adaptively, and obtains the local template with adaptive effect that currently conforms to the intra-chip pulse of the channel effect Signal, the specific implementation method is:

1) The method of adaptively adjusting the template is the LMS (minimum mean square error) algorithm: the initial channel response is ω(0)=ω _est , that is, the response of the channel estimation result "tap line-delay" multipath model, and n= 0; Let x (n) be transmitter to send signal, y (n) be the signal that receives (use I road signal I ' (n) after frequency correction here); Update n=n+1, order:

e(n)=y(n)-ω ^T (n-1)x(n), that is, the amount of error;

ω(n)=ω(n-1)+μ(n)x(n)e(n), which is the update value of the channel response; where μ(n) is set by the system scheme, such as any constant or input signal Statistics;

After this process is updated for a period of time, the matching template with adaptive effect can be calculated: t(n)=ω ^T (n)x(n);

2) The template adjustment signal obtained by the timing correction loop adjusts the template signals of the main path and selected secondary paths within the time interval of two consecutive channel estimations;

Compared with the traditional zero-IF digital receiving method, this method has the advantages of high oversampling in the digital domain, multiple sampling points in a single chip, combined with flexible matching correlation templates, which can better adapt to the pulses in the chip passing through the channel. Various possible waveforms; while the baseband signals of traditional zero-IF transceivers are simple 1 or 0 information (BSPK polarity modulation), it is difficult to correctly handle the different waveform characteristics (duty duty) of the received UWB pulses ratio, polarity change within a chip, etc.).

As shown in Figure 4, this method uses a symbol synchronization acquisition algorithm that adaptively adjusts the correlation threshold, which can achieve a very low probability of false synchronization and missing synchronization, and achieve high-performance symbol synchronization. The specific implementation method is as follows:

1) In order to prevent accidental noise from searching for wrong main path positions, a confirmation mechanism can be introduced: only the main path correlation value exceeds the set threshold, and the positions corresponding to the main path correlation values are all in the same location in several consecutive searches. The position of the main diameter is considered to be correct only when the position is offset.

即 $\mathrm{Th}=1/\sqrt{2}{\mathrm{Th}}_{\mathrm{main}},$ 并送往信道估计器；如此则可有效地避免误同步的发生。2) In the process of symbol synchronization, in order to avoid the occasional "correlation value exceeding the threshold" which may cause false symbol synchronization; the threshold needs to be adaptively increased by the following method: when the position of the main path is captured, the correlation threshold is used by a correlation Threshold adjustment circuit, adjusted to the current main path correlation value

Right now $\mathrm{Th}=1/\sqrt{2}{\mathrm{Th}}_{\mathrm{main}},$ And sent to the channel estimator; in this way, the occurrence of false synchronization can be effectively avoided.

Fig. 5 is a block diagram of the circuit principle of frame synchronization capture in the pulse ultra-wideband wireless receiving method of quadrature down-mixing digital template matching. As shown in Figure 5, the specific implementation method is as follows:

The output signal I"(n) received by Rake correlation is used as the input of the frame synchronization capture circuit; the polarity of the signal is used as the information data bit, and the frame separator can be found by searching the output corresponding to the information data bit, so as to locate the frame Header and locate the starting position of data information to achieve frame synchronization.

The output signal I'"(n) of Rake correlation reception after frame synchronization is sent to the subsequent spreading code despreading module and channel decoder. System scheme setting; the bit stream obtained after spreading code despreading and channel decoding is the final data, which enters the data interface and sends it to the upper layer for processing.

As mentioned above, according to the present invention, the pulse ultra-wideband wireless receiver implementation scheme of quadrature down-mixing digital template matching avoids the limitations of traditional pure analog correlation and pure digital correlation Rake receivers for high-speed data applications, and simplifies the hardware Structure and implementation cost; high oversampling in the digital domain combined with flexible matching correlation templates can better adapt to various possible waveforms of pulses passing through the channel in the chip; it is conducive to the correction of frequency deviation; it supports both pulse radio and carrier schemes UWB signal reception has strong versatility; it has simple, practical and structured circuit design and implementation, and is easy to integrate chips.

Claims (5)

1. the pulse ultra wide band radio signal receiving method of orthogonal lower mixing frequency digital masterplate coupling is characterized in that described method contains following steps successively:

Step 1 received signal through the radio-frequency front-end of forming by serial by antenna, low noise amplifier, I/Q two-way Analogue mixer and low pass filter successively after, obtain zero intermediate frequency analog baseband signal through frequency spectrum shift; The local oscillation signal frequency of described I/Q two-way Analogue mixer is determined by following mode: the piconet framework of being worked out according to the ultra-wideband communication system scheme and launch user-selected center frequency points of getting, choose the local oscillation signal of the sine wave signal of same frequency as this frequency mixer, this I/Q two-way local oscillation signal keeps orthogonality;

The zero intermediate frequency analog baseband signal of this AFE (analog front end) of step 2 output is delivered to the analog to digital converter (ADC) that sample rate is higher than the spreading rate twice and is carried out the sampling of zero intermediate frequency baseband signal, generates the digital baseband signal of I/Q two-way;

The I/Q two-way high-speed figure baseband signal that step 3 obtains step 2 is sent into a relevant Rake receiver of numeral coupling that is integrated on the digital integrated circuit chip, numeric field according to the following steps to I/Q two-way digital baseband signal carry out frequency and estimation regularly and correction, the numeral coupling is relevant and Rake reception and frame synchronization are caught, the output signal that obtains is carried out spreading code despreading and channel decoding successively, recovers original transmission signal:

Step 3.1 is carried out frequency correction to the I/Q two-way high-speed figure baseband signal of input successively according to the following steps with a frequency correction circuit:

Step 3.1.1 I/Q two-way digital baseband signal I ' (n), Q ' (n) sends into phase discriminator in the carrier recovery loop that comprises the polarity type carrier recovery loop, extracts phase error signal ε by following algorithm _c(n):

ε _c(n)=I ' (n) * Q ' (n), n is the sampled point label;

Step 3.1.2 is phase error signal ε _c(n) the second order digital loop filters of importing in this carrier recovery loop carries out filtering, and its sequential is:

The phase error accumulating signal φ (n+1) of n+1 sampled point is:

φ(n+1)＝φ(n)+K ₁f(n)+K ₂ε _c(n)

The frequency error accumulating signal f (n+1) of n+1 sampled point is:

f(n+1)＝f(n)+ε _c(n)，

Wherein, constant K ₁And K ₂Be the filter factor of this second order digital loop filters, when the sample rate f of described ADC _sDuring much larger than the natural frequency of loop:

$K_2=2{\mathrm{\&eta;\&omega;}}_n{T}_s,K_1=K_2^2/4{\mathrm{\&eta;}}^2={\mathrm{\&omega;}}_n{T}_s,$

ω _n=2 π f _nBe the natural angular frequency of loop, η is a damping coefficient, T _s=1/f _s

Step 3.1.3 is the input of the phase error accumulating signal φ (n+1) that obtains through described loop filtering as local digital controlled oscillator in the described loop, to adjust the frequency of local oscillation signal;

Step 3.1.4 multiplies each other the I/Q two-way digital baseband signal I (m), the Q (m) that are input to the frequency correction loop and output cos φ (n), the sin φ (n) of local digital controlled oscillator, carry out phase place rotation, thus obtain through the I/Q two-way digital baseband signal I ' of frequency correction (m), Q ' (m):

I′(m)＝I(m)×cosφ(n)+Q(m)×sinφ(n)

Q′(m)＝Q(m)×cosφ(n)-I(m)×sinφ(n)

The difference of m and n is the computing time-delay of described whole carrier recovery loop;

Step 3.2 is corrected loop with a timing and according to following steps the output signal of described frequency correction circuit is done regularly to correct:

Step 3.2.1 utilize through the I way word baseband signal I ' of frequency correction (n) with the main path position of channel estimator output, obtain main footpath signal;

Step 3.2.2 obtains main footpath correlation Corr (n) to a main footpath of signal input, described main footpath correlator;

Step 3.2.3 local masterplate signal and described main footpath signal send into one early correlator obtain a footpath correlation Corr (n-1) early, this morning, directly correlation was the cross correlation value of the local masterplate signal and the signal in main footpath signal unit sampling interval of morning, this this locality masterplate signal is produced by local masterplate signal generator, its initial value is the standard signal of transmitter emission in the known single chip, and adjusts signal according to masterplate and carry out phase delay or in advance;

Step 3.2.4 sends into a slow correlator to described local masterplate signal of step 3.2.3 and described main footpath signal, obtain a footpath correlation Corr (n+1) late, this slow footpath correlation is local masterplate signal and the main directly cross correlation value of the signal in a slow unit sampling interval of signal;

Step 3.2.5 resulting main footpath correlation Corr (n), early footpath correlation Corr (n-1), footpath correlation Corr (n+1) sends into a timing error extractor late, calculates signal of timing error ε by following formula _t(n):

ε _t(n)＝sign(Corr(n))*[Corr(n+1)-Corr(n-1)]；

The signal of timing error ε that step 3.2.6 obtains step 3.2.5 _t(n) send into the timing error value that single order digital lowpass loop filter obtains accumulating:

Its sequential is: E _t(n)=E _t(n-1)+K ₁ε _t(n),

Constant K ₁Be the filter factor of this filter, K ₁＜1;

Whether step 3.2.7 reaches preset threshold with the accumulated error value that a comparator comes determining step 3.2.6 to obtain, if reach preset threshold then the zero clearing of accumulated error value, send a masterplate simultaneously and adjust signal, control local masterplate signal generator the phase delay of local masterplate signal or carry the previous unit sampling interval, and the output masterplate is adjusted the numeral coupling relevant Rake receiving circuit of signal after giving and is regularly corrected realizing; Take to shift to an earlier date or postpone to depend on the polarity of accumulated error,, then shift to an earlier date if just, otherwise, postpone;

The I way word baseband signal I ' that the masterplate that step 3.3 obtains step 3.2.7 is adjusted signal and process frequency correction (n) sends into a relevant and Rake receiving circuit of numeral coupling, calculates the output of this Rake correlation reception circuit according to the following steps:

Step 3.3.1 sends into channel estimator to this digital baseband signal with the signal masterplate of several different deviation posts of being exported by local masterplate signal generator, search a main path position corresponding and corresponding with the big correlation of several times successively inferior path position through the signal correction computing with maximum related value, described related operation is meant: calculate the product of two input signals and add up by each sampled point, the procurement process of main path position is the process of sign synchronization; The masterplate that step 3.2.7 obtains is adjusted signal and in the time interval of double channel estimating the masterplate signal that main footpath reaches selected each time footpath is adjusted;

The main footpath that step 3.3.2 obtains according to step 3.3.1 and the position in each time footpath, (n) sending into a correlator bank respectively with the directly corresponding masterplate signal of the main footpath that obtains and selected each time through the I way word baseband signal I ' of frequency correction, the described method of 3.3.1 is carried out related operation respectively set by step, obtain the signal after numeral is mated, be sent to the Rake combiner;

Step 3.3.3Rake combiner merges time number in footpath and is at least one, no maximum;

The multipath energy that step 3.3.4Rake combiner is made of the pairing correlated results in main footpath and each time footpath is weighted than criterion than criterion or equal gain combining according to the maximum merging and adds up, obtain a total energy, i.e. the output I ＂ (n) of this Rake correlation reception circuit;

Step 4 is sent to a frame synchronization grabber to the output signal I ＂ (n) of the Rake correlation reception that step 3.3.4 obtains, according to the following steps the Rake correlation reception signal of output frame after synchronously:

Step 4.1 is sent to a vertical sync circuit to the output signal I ＂ (n) that step 3.3.4 obtains, the polarity of this signal as the information data position, search can be found the frame separator corresponding to the output of this information data position, thus the original position of locating frame head and locator data information;

The output of the Rake correlation reception circuit after step 4.2 finishes step 4.1 frame synchronization I ＂ ' as a result (n) is sent to follow-up spreading code despreading module and channel decoder.

2. the pulse ultra wide band radio signal receiving method of orthogonal lower mixing frequency digital masterplate coupling according to claim 1, it is characterized in that, according to Minimum Mean Square Error LMS algorithm, the channel estimation results that obtains according to step 3.3.1 and through the I road signal I ' after the frequency correction (n), the self adaptation masterplate adjuster by this LMS algorithm of utilization obtains a matching stencil t (n) that the self adaptation effect is arranged.

3. the pulse ultra wide band radio signal receiving method of orthogonal lower mixing frequency digital masterplate coupling according to claim 1, it is characterized in that, in order to prevent that accidental noise from causing searching wrong main path position, after step 3.3.1, introduce following confirmation method: have only main footpath correlation to surpass preset threshold, and when the position of main footpath correlation correspondence is all at same deviation post in several times are searched for continuously, think that just main path position is correct.

4. the pulse ultra wide band radio signal receiving method of orthogonal lower mixing frequency digital masterplate coupling according to claim 3, it is characterized in that, for the situation of avoiding occurrent " main footpath correlation surpasses threshold value " causes the mistake symbol synchronous, threshold value needs to improve adaptively by the following method: after capturing correct main path position, dependent thresholds is adjusted circuit with dependent thresholds, be adjusted into current main footpath correlation

, and be sent to channel estimator.

5. the pulse ultra wide band radio signal receiving method of orthogonal lower mixing frequency digital masterplate coupling according to claim 1, it is characterized in that described received signal is that the Wireless Personal Network physical layer communicates the impulse radio that is utilized or the ultra-wideband impulse signal of carrier wave is arranged.

Images