Disclosure of Invention
The technical problem to be solved by the present invention is to provide a method for acquiring a local dirty correlation template of a UWB receiver, which can acquire an accurate local dirty correlation template, suppress noise power in the local dirty correlation template to the maximum extent, improve demodulation performance, and effectively reduce the number of multipaths of the local dirty correlation template and further reduce the complexity of implementing the receiver.
In order to solve the technical problems, the invention adopts the technical scheme that: the method for acquiring the local dirty correlation template of the UWB receiver specifically comprises the following steps:
s1: on one hand, UWB digital signals processed by an analog-digital converter ADC are processed by a cross correlator, and on the other hand, convolution processing is simultaneously carried out;
s2: then, noise elimination processing is carried out through a frequency domain smoothing module, and then noise elimination is carried out in an accumulated mode to obtain a local dirty correlation template;
s3: and accumulating the local dirty correlation templates obtained in the step S2, and then performing channel multi-path screening on the accumulated local dirty correlation templates to obtain the local dirty correlation templates from the data larger than the set threshold.
As a preferred embodiment of the present invention, the method further includes step S4: and (8) performing SFD detection on the data subjected to the convolution processing in the step (S1), if the SFD is detected, obtaining a local dirty related template, and otherwise, continuously and circularly performing the convolution processing and the SFD detection. Wherein SFD is an abbreviation for start-of-frame delimiter frame start delimiter.
As a preferred embodiment of the present invention, the method further includes step S5: and processing the data which does not pass the set threshold comparison in the channel multipath screening in the step S3 by the method in the step S4, and finally obtaining the local dirty correlation template.
By adopting the technical scheme, the UWB signal is received by adopting the method of the local dirty related template, and the frequency domain smoothing noise elimination, the time domain accumulation noise elimination and the multipath voting method are adopted to reduce the multipath number of the local dirty related template when the local dirty related template is obtained, so that the precision of the local dirty related template is improved, the demodulation performance of the UWB is improved, and meanwhile, the realization complexity is further reduced due to the reduction of the multipath number of the local dirty related template. Wherein the single pulse correlation module: the method is used for single pulse matching correlation filtering out-of-band noise; a leader sequence correlation module: the device is used for carrying out correlation accumulation on the output of the single pulse correlation module to eliminate noise; a frequency domain smoothing module: further eliminating noise in a frequency domain and then accumulating; a multipath screening module: and screening the optimal multipath component and deleting the interference multipath.
As a preferred technical solution of the present invention, the cross correlator in step S1 includes a single-pulse correlation module and a preamble sequence correlation module, wherein the UWB digital signal processed by the ADC is processed by the single-pulse correlation module and then processed by the preamble sequence correlation module, and a formula for calculating a correlation result when the single-pulse correlation module is used for processing is as follows:
wherein b (m) is the correlation result; p (n) is a locally generated template pulse; p (n) is a conjugate function of p (n); x (n) is a sampling output signal of the receiving analog-digital converter ADC, which comprises K times of repeated preamble sequence S information; n is the pulse length; n and m are serial numbers of sampling signal output sampling points;
when a leader sequence correlation module is adopted for processing, the local leader sequence is recorded as: s (iL + j), where j is 0., L-1, i is 0., L-1, L is a sequence spreading factor; the calculation formula is as follows:
wherein
C (i) is a local random 3-element code sequence of length L, delta
L(i)=1,
Represents the Kronecker product; and b (iL + j) is a correlation result after being processed by a single-pulse correlation module.
As a preferred technical solution of the present invention, the specific method for performing noise cancellation processing by using the frequency domain smoothing module in step S2 includes:
s21: first convert d (j) to the frequency domain, denoted as: g ═ fft (d);
wherein d ═ d (0),. ·, d (L-1)]T,g=[g(0),...,g(L-1)]TL is a sequence spreading factor;
s22: then, a frequency domain filter is adopted to perform frequency domain denoising on the g to obtain a denoising frequency domain
The formula is as follows:
wherein W is a frequency domain filter coefficient matrix;
s23: at last handle
By inverse Fourier transform into time domain
The formula is as follows:
wherein
L is a sequence spreading factor.
As a preferred technical solution of the present invention, the step S2 further includes a step S24 of accumulating and denoising the signal after the noise cancellation process is performed by the frequency domain smoothing module, where the formula is as follows:
wherein K is 0.., K-1; k is the repeated sending times of the local leader sequence S in time;
as a preferred technical solution of the present invention, the method for screening channel multipath in step S3 includes:
s31: firstly, establishing a multipath voting pool vk=[vk(0),...,vk(L-1)]T,v-1=[0,...,0]T;
S32: then setting a threshold value delta, voting each path according to the following formula,
s33: finally, setting a threshold value gamma, and obtaining a local dirty correlation template h according to the following formula rulekThe formula is as follows:
the local dirty correlation template is hk=[hk(0),...,hk(L-1)]TAnd initialize h0=[1,0,0,...,0]T。
Compared with the prior art, the technical scheme has the beneficial effects that: the UWB signal is received by adopting a local dirty correlation template method, and the frequency domain smoothing noise elimination, the time domain accumulation noise elimination and the multipath voting method are adopted to reduce the multipath number of the local dirty correlation template when the local dirty correlation template is obtained, so that the precision of the local dirty correlation template is improved, the UWB demodulation performance is improved, and meanwhile, the realization complexity is further reduced due to the reduction of the multipath number of the local dirty correlation template.
Detailed Description
Example (b): as shown in fig. 1, the method for acquiring a local dirty correlation template of an UWB receiver specifically includes the following steps:
s1: on one hand, UWB digital signals processed by an analog-digital converter ADC are processed by a cross correlator, and on the other hand, convolution processing is simultaneously carried out;
the cross-correlator in step S1 includes a single-pulse correlation module and a preamble sequence correlation module, wherein the UWB digital signal processed by the ADC is processed by the single-pulse correlation module and then processed by the preamble sequence correlation module, and a formula for calculating a correlation result used when the single-pulse correlation module is used for processing is as follows:
wherein b (m) is the correlation result; p (n) is a locally generated template pulse; p (n) is a conjugate function of p (n); x (n) is a sampling output signal of the receiving analog-digital converter ADC, which comprises K times of repeated preamble sequence S information; n is the pulse length; n and m are serial numbers of sampling signal output sampling points;
when a leader sequence correlation module is adopted for processing, the local leader sequence is recorded as: s (iL + j), where j is 0., L-1, i is 0., L-1, L is a sequence spreading factor; the calculation formula is as follows:
wherein
C (i) is a local random 3-element code sequence of length L, delta
L(i)=1,
Represents the Kronecker product; b (iL + j) is a correlation result after being processed by a single-pulse correlation module;
s2: then, noise elimination processing is carried out through a frequency domain smoothing module, and then noise elimination is carried out in an accumulated mode to obtain a local dirty correlation template;
one possible local random 3-ary code sequence when length L is 31 is:
-0000+10-0+ + +0+ -000+ - ++++ 00- + 0-00; wherein-represents-1, + represents 1;
the specific method for performing noise elimination processing by using the frequency domain smoothing module in step S2 is as follows:
s21: first convert d (j) to the frequency domain, denoted as: g ═ fft (d);
wherein d ═ d (0),. ·, d (L-1)]T,g=[g(0),...,g(L-1)]TL is a sequence spreading factor;
s22: then, a frequency domain filter is adopted to perform frequency domain denoising on the g to obtain a denoising frequency domain
The formula is as follows:
wherein W is a frequency domain filter coefficient matrix;
s23: at last handle
By inverse Fourier transform into time domain
The formula is as follows:
wherein
L is a sequence spreading factor;
s24: after the noise elimination processing is carried out by adopting the frequency domain smoothing module, the accumulative noise elimination is carried out, and the formula is as follows:
wherein K is 0.., K-1; k is the repeated sending times of the local leader sequence S in time;
s3: and accumulating the local dirty correlation templates obtained in the step S2, and then performing channel multi-path screening on the accumulated local dirty correlation templates to obtain the local dirty correlation templates from the data larger than the set threshold.
The method for screening the channel multipath in the step S3 includes:
s31: firstly, establishing a multipath voting pool vk=[vk(0),...,vk(L-1)]T,v-1=[0,...,0]T;
S32: then setting a threshold value delta, voting each path according to the following formula,
s33: finally, setting a threshold value gamma, and obtaining a local dirty correlation template h according to the following formula rulekThe formula is as follows:
local dirty correlation templateIs hk=[hk(0),...,hk(L-1)]TAnd initialize h0=[1,0,0,...,0]T;
S4: performing SFD detection on the data subjected to convolution processing in the step S1, if the SFD is detected, obtaining a local dirty related template, and if the SFD is not detected, continuing to circularly perform convolution processing and SFD detection;
s5: and processing the data which does not pass the set threshold comparison in the channel multipath screening in the step S3 by the method in the step S4, and finally obtaining the local dirty correlation template.
While the embodiments of the present invention have been described in detail with reference to the drawings, the present invention is not limited to the above embodiments, and various changes can be made without departing from the spirit of the present invention within the knowledge of those skilled in the art.