CN105743540A - Signal processing system and method for ultra wide band system - Google Patents

Abstract

The invention provides a signal processing system and a signal processing method for an ultra wide band system. The system comprises a signal preprocessing unit, a multi-path channel and a signal combining unit, wherein the signal preprocessing unit performs branch preprocessing on a signal, and after passing through the multi-path channel, the processed signals are subjected to post-RAKE combination at a receiving end via the signal combining unit. In order to achieve the maximal output signal-to-noise ratio at the receiving end, a preprocessing diversity and post-RAKE parameter estimation method based on channel matrix feature vector is obtained through theoretical analysis.

Description

Signal processing system and method for ultra-wideband system

Technical Field

The invention belongs to the field of communication, and particularly relates to a signal processing system and a signal processing method for an ultra-wideband system.

Background

The ultra-wideband (UWB) technology is a new wireless communication technology, different from the traditional wireless communication technology, a UWB system does not use a carrier wave, but uses a narrow pulse with a very low duty ratio as an information carrier, and has the characteristics of high communication rate, low power consumption, simple realization, good confidentiality and the like.

The ultra-wideband system uses nanosecond or even picosecond pulse width pulses for communication, so that the ultra-wideband system has strong multipath resolution capability, and the characteristics ensure that a RAKE diversity receiver is used at a receiving end in most of transmitting and receiving schemes to capture and combine multipath energy, improve the receiving performance and improve the receiving efficiency. The traditional RAKE receiver aiming at the UWB system realizes multipath combination at a receiving end, and is never realized at a sending end, and in other communication systems such as a TDD-CDMA system, a scheme has been realized by realizing multipath separation and sending at the sending end, because channel models of sending an uplink channel and a downlink channel in the TDD system are consistent, parameters of uplink channel estimation can be directly used for the downlink channel, the scheme does not carry out multipath combination and direct receiving at the receiving end, and only needs to directly carry out sampling judgment on the maximum path, wherein the realization mode is called as Pre-RAKE (Pre-RAKE). Pre-Rake performs Pre-delay combination processing on signals at a sending end according to channel state information, wherein delay combination coefficients are equal to conjugates of channel fading factors.

At present, other communication systems only consider the multipath diversity of the transmitting end aiming at the Pre-Rake processing mode, and the receiving end directly makes a decision, and the implementation scheme is shown in fig. 1. In the design of the UWB system, the uplink channel parameter estimation value can be used for the downlink channel, and therefore, it is theoretically possible to use Pre-Rake in the signal processing of the UWB system.

First, the present PRE-RAKE scheme for UWB does not exist, and how to further combine the increased multipath energy at the receiving end is not considered, and how to further combine the multipath energy at the receiving end is found out an optimal multipath combining parameter estimation method in theoretical analysis by using a UWB channel model without increasing the complexity of the system.

Disclosure of Invention

Technical problem to be solved

It is an object of the present invention to provide a signal processing system and method for ultra-wideband systems that provides a significant improvement in bit error rate performance over conventional RAKE reception.

(II) technical scheme

The invention provides a signal processing system for an ultra-wideband system, wherein the ultra-wideband system comprises a sending end and a receiving end, and the signal processing system comprises:

the signal preprocessing unit is used for performing branch processing on the sending signal at the sending end to obtain signals of a plurality of branches;

a multipath channel for transmitting signals of a plurality of branches from a transmitting end to a receiving end;

a signal combining unit for combining the signals of the plurality of branches at a receiving end, wherein the signals can be post-RAKE combined, so the signal combining unit can adopt a post-RAKE receiver.

Wherein, the expression of the transmission signal is:

$\begin{array}{c}{s}_i\left(t\right)=\sqrt{P}\underset{i=0}{\overset{\mathrm{\∞}}{\Σ}}{b}_{i}w(t-{\mathrm{iT}}_s),\\ w\left(t\right)=\underset{n=0}{\overset{N_s-1}{\Σ}}{c}_{n}g(t-{\mathrm{nT}}_c),\end{array}$

wherein i represents the ith time point number, n represents the nth number in one symbol period, b_iIndicating the character signal numbered i, c_nFor spreading codes, P is the energy of the transmitted signal, and a symbol is formed by N_SA composition of pulses, g (T) being a function of the waveform of the transmitted pulse, T_sRepresents one symbol period;

the signal preprocessing unit is used for preprocessing the signal according to a time delay parameter T ═ tau_K，…，τ₁]The transmitting signal is processed in a branch way, and the branch time delay of the kth branch is T_s-τ_kWherein, τ_kRepresenting the kth multipath delay.

Further, the sampled output values of the signals of the multiple branches after passing through the multipath channel are represented by vectors as:

Y＝[y₁(i)，…，y_M(i)]^T＝Ab_iHB+N，

wherein, y_M(i) Denotes the sample output vector value of the M-th column with sample number i, a denotes the fixed gain of the system, H is a matrix of 2M × K, and M ═ T_S/T_b，

$H^T=\left(\begin{array}{c}0,\mathrm{....0...},\mathrm{0...},{\mathrm{\α}}_1,\mathrm{...0...},{\mathrm{\α}}_2,\mathrm{...0...}{\mathrm{\α}}_L\mathrm{...},0\\ 0,\mathrm{....}{\mathrm{\α}}_1\mathrm{...0...},{\mathrm{\α}}_2,\mathrm{....0},\mathrm{........}{\mathrm{\α}}_L\mathrm{...........},0\\ \mathrm{......},\mathrm{..........}\\ 0,\mathrm{...}{\mathrm{\α}}_{K-1}\mathrm{..0...}{\mathrm{\α}}_K,\mathrm{......0},\mathrm{..}{\mathrm{\α}}_L\mathrm{...0..........},0\end{array}\right),$

B＝[β_K，…，β₁]^TRepresenting a merged vector, β_KDenotes the K-th vector, α_lRepresenting the path fading parameters of the l-th path in a multi-path channel, α_lIn the case of a real number,l represents a total of L multipaths and N represents the quantization level of the noise.

Further, the signal combining unit combines the signals at the receiving end according to a branch combining weight W, which represents W ═ γ₁，…，γ_M]^T，γ_MThe mth vector weight value, numbered M, is represented, and the combined signal is represented as:

z(i)＝W^TY＝Ab_mW^THB+W^TN。

further, the system further comprises: a parameter estimation unit, configured to calculate a delay parameter T, a combining vector B, and a branch combining weight W, where the delay parameter T has an expression:

T＝argmax{||D||}，

wherein D is HH^TAnd according to the estimation of the characteristic value of the graph and the disc theorem, obtaining:

1≤||D||≤argmax{d_i，i+R′_i(A)}，

wherein d is_i，iDenotes the center of a circle, R'_i(A) The radius of the circle is represented, and,

$\mathrm{argmax}\left\{d_{i,i}\right\}=\underset{i=1}{\overset{K}{\Σ}}{{\stackrel{\‾}{\mathrm{\α}}}_i}^2,$

representation α₁，....，α_LThe numbers in which K absolute values are the largest and different from each other;

the branch merging weight W is a feature vector corresponding to the maximum feature value D;

the expression of the merging vector B is as follows:

B＝(W^TH)^T。

the invention also provides a signal processing method for the ultra-wideband system, which comprises the following steps:

s1, branching the sending signal at the sending end to obtain signals of a plurality of branches;

s2, transmitting the signals of multiple branches from the transmitting end to the receiving end;

s3, combining the signals of the plurality of branches at the receiving end.

Wherein, the expression of the transmission signal is:

$\begin{array}{c}{s}_i\left(t\right)=\sqrt{P}\underset{i=0}{\overset{\mathrm{\∞}}{\Σ}}{b}_{i}w(t-{\mathrm{iT}}_s),\\ w\left(t\right)=\underset{n=0}{\overset{N_s-1}{\Σ}}{c}_{n}g(t-{\mathrm{nT}}_c),\end{array}$

wherein i represents the ith time point number, n represents the nth number in one symbol period, b_iIndicating the character signal numbered i, c_nFor spreading codes, P is the energy of the transmitted signal, and a symbol is formed by N_SA composition of pulses, g (T) being a function of the waveform of the transmitted pulse, T_sRepresents one symbol period;

the signal preprocessing unit is used for preprocessing the signal according to a time delay parameter T ═ tau_K，…，τ₁]The transmitting signal is processed in a branch way, and the branch time delay of the kth branch is T_s-τ_kWherein, τ_kRepresenting the kth multipath delay.

Further, the sampled output values of the signals of the multiple branches after passing through the multipath channel are represented by vectors as:

Y＝[y₁(i)，…，y_M(i)]^T＝Ab_iHB+N，

y_M(i) denotes the sample output vector value of the M-th column with sample number i, a denotes the fixed gain of the system, H is a matrix of 2M × K, and M ═ T_S/T_b，

$H^T=\left(\begin{array}{c}0,\mathrm{....0...},\mathrm{0...},{\mathrm{\α}}_1,\mathrm{...0...},{\mathrm{\α}}_2,\mathrm{...0...}{\mathrm{\α}}_L\mathrm{...},0\\ 0,\mathrm{....}{\mathrm{\α}}_1\mathrm{...0...},{\mathrm{\α}}_2,\mathrm{....0},\mathrm{........}{\mathrm{\α}}_L\mathrm{...........},0\\ \mathrm{......},\mathrm{..........}\\ 0,\mathrm{...}{\mathrm{\α}}_{K-1}\mathrm{..0...}{\mathrm{\α}}_K,\mathrm{......0},\mathrm{..}{\mathrm{\α}}_L\mathrm{...0..........},0\end{array}\right),$

B＝[β_K，...，β₁]^TRepresenting a merged vector, β_KDenotes the K-th vector, α_lRepresenting the path fading parameters of the l-th path in a multi-path channel, α_lIn the case of a real number,l represents a total of L multipaths and N represents the quantization level of the noise.

Further, step S3 combines the signals at the receiving end according to a branch combining weight W, which represents that W ═ γ₁，…，γ_M]^T，γ_MThe mth vector weight value, numbered M, is represented, and the combined signal is represented as:

z(i)＝W^TY＝Ab_mW^THB+W^TN。

further, the method further comprises: s0, calculating a delay parameter T, a merging vector B and a branch merging weight W, wherein the expression of the delay parameter T is as follows:

T＝argmax{||D||}，

wherein D is HH^TAnd according to the estimation of the characteristic value of the graph and the disc theorem, obtaining:

1≤||D||≤argmax{d_i，i+R′_i(A)}，

wherein d is_i，iDenotes the center of a circle, R'_i(A) The radius of the circle is represented, and,

$\mathrm{argmax}\left\{d_{i,i}\right\}=\underset{i=1}{\overset{K}{\Σ}}{{\stackrel{\‾}{\mathrm{\α}}}_i}^2,$

representation α₁，....，α_LThe numbers in which K absolute values are the largest and different from each other;

the branch merging weight W is a feature vector corresponding to the maximum feature value D;

the expression of the merging vector B is as follows:

B＝(W^TH)^T。

(III) advantageous effects

The invention provides a signal processing system and a signal processing method of an ultra-wideband system based on the thoughts of diversity and combination, which can utilize multipath and improve the receiving performance. In addition, the invention carries out parameter estimation based on the combination of the maximum eigenvalue eigenvector of the channel, and the obtained signal-to-noise ratio is always higher than the performance of the traditional maximum ratio combination RAKE receiver. Finally, the invention is a matrix maximum eigenvalue estimation method based on the disk theorem, thereby obtaining the optimal time delay estimation method, and the method can also be applied to other communication fields.

Drawings

Fig. 1 is a PRE-RAKE diversity processing scheme in the prior art.

Fig. 2 is a schematic diagram of a signal processing system for an ultra-wideband system provided by the present invention.

Figure 3 is a block diagram of a signal processing system for an ultra-wideband system provided by the present invention.

FIG. 4 is a schematic diagram of a matrix maximum eigenvalue estimation method based on the disk theorem according to the present invention.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to specific embodiments and the accompanying drawings.

As shown in fig. 2 and 3, the present invention provides a signal processing system including a pre-processing unit at a transmitting end and a post-RAKE receiver at a receiving end.

Because in the UWB system, the signal employs DS-BPSK modulation, the transmission signal can be expressed as:

$s_i\left(t\right)=\sqrt{P}\underset{i=0}{\overset{\mathrm{\∞}}{\Σ}}{b}_{i}w(t-{\mathrm{iT}}_s)---\left(1\right)$

$w\left(t\right)=\underset{n=0}{\overset{N_s-1}{\Σ}}{c}_{n}g(t-{\mathrm{nT}}_c)---\left(2\right)$

wherein s is_i∈ { -1, +1} represents modulated binary information, c_nFor spreading codes, P is the energy of the transmitted signal, and a symbol is formed by N_SThe pulse composition, g (t) is a waveform function of the transmitted pulse, and the function uses a 5 th order Gaussian function in order to meet the definition of the ultra-wideband signal:

$g\left(t\right)=K_2(-15\frac{t}{\mathrm{\σ}}+10\frac{t^3}{{\mathrm{\σ}}^3}-\frac{t^5}{n^5})e^{\frac{t^2}{2{\mathrm{\σ}}^2}}---\left(3\right)$

as shown in fig. 3, a transmission signal s_i(T) first, the branch processing is carried out by the preprocessing unit, and the branch time delay is T_s-τ_k，(τ_k≤T_s) Let the merge vector be B ═ β_K，…，β₁]^TTo maintain normalization, I B I calculation is required²1. UWB signals are subject to severe multipath fading in indoor transmissions, the UWB channel model is different from the rayleigh and rice channel models, and according to the latest description of the IEEE802.15.3TG3A working group on the UWB channel model, the UWB indoor channel model can be described by the impulse response of a tapped delay line structure:

$h\left(t\right)=\underset{l=1}{\overset{L}{\Σ}}{\mathrm{\α}}_l\mathrm{\δ}(t-{\mathrm{\τ}}_l)---\left(4\right)$

α_lrepresenting path fading parameters of the l-th path, α_lIn the case of a real number,τ_lrepresenting the signal delay, L is the number of multipaths in the multipath channel. The signal time delays are distinguished according to clusters and paths and respectively have independent distribution functions.

The signal passes through a multipath channel, and the received signal at the receiving end is:

$r_i\left(t\right)=\sqrt{P}\underset{l=1}{\overset{L}{\Σ}}\underset{k=1}{\overset{K}{\Σ}}{\mathrm{\β}}_k{\mathrm{\α}}_l{b}_{i}w(t+{\mathrm{\τ}}_k-{\mathrm{\τ}}_l+T_s)+n\left(t\right)---\left(5\right)$

wherein n (t) is white Gaussian noise, and the received signal is correlated and de-spread by a pulse matching filter to obtain output:

$y_m\left(i\right)={\∫}_{\left(i\right)T_s}^{(1+i)T_s}r(t-{\mathrm{\τ}}_m)*w\left(t\right)dt---\left(6\right)$

the matched and filtered signals are subjected to time delay sampling, and the sampling rate is R-1/T_bEach branch delay is T_s-τ_mThe output comprises M values y_m(i)，y_m(i) The self-interference term and the noise term are included, the self-interference term is ignored and only the Gaussian white noise term is considered on the assumption that the spread spectrum code has good self-correlation characteristics. The sampled output values of the branches before merging are vector-represented as: y ═ Y₁(i)，…，y_M(i)]^T＝Ab_iHB + n, where H is a matrix of 2M × K, M ═ T_S/T_b，

$H^T=\left(\begin{array}{c}0,\mathrm{....0...},\mathrm{0...},{\mathrm{\α}}_1,\mathrm{...0...},{\mathrm{\α}}_2,\mathrm{...0...}{\mathrm{\α}}_L\mathrm{...},0\\ 0,\mathrm{....}{\mathrm{\α}}_1\mathrm{...0...},{\mathrm{\α}}_2,\mathrm{....0},\mathrm{........}{\mathrm{\α}}_L\mathrm{...........},0\\ \mathrm{......},\mathrm{..........}\\ 0,\mathrm{...}{\mathrm{\α}}_{K-1}\mathrm{..0...}{\mathrm{\α}}_K,\mathrm{......0},\mathrm{..}{\mathrm{\α}}_L\mathrm{...0..........},0\end{array}\right)---\left(7\right)$

At the receiving end pairSignals are combined, and the branch combining weight vector represents W ═ gamma₁，…，γ_M]^TThe combined signal is expressed as:

z(i)＝W^TY＝Ab_mW^THB+W^TN(8)

the optimal parameter estimation algorithm is obtained by theoretical analysis.

As can be seen from equation (7), the output SNRThe problem of finding the best parameter estimate is translated into finding the best parameter B ═ β_K，…，β₁]^T，W＝[γ₁，…，γ_M]^TAnd T ═ τ_K，…，τ₁]To maximize the output signal-to-noise ratio. In the estimation, | W is assumed first^T|²And T ═ τ_K，...，τ₁]I.e. H^TFixing, and solving an optimal vector B according to a Cauchy inequality:

$SNR=P_0\frac{|W^{T}HB{|}^2}{|W^T{|}^2}\≤P_0\frac{|W^T{\mathrm{HH}}^{T}W{|}^2}{|W^T{|}^2}---\left(9\right)$

when B is ═ W^TH)^TThe time inequality takes equal sign, and the output signal-to-noise ratio is maximum at the momentThen, the best W parameter estimation value is obtained, and the best parameter is obtainedIs determined by the following formula:

$\tilde{W}=\mathrm{argmax}\left\{\frac{W^T{\mathrm{HH}}^{T}W}{W^{T}W}\right\}---\left(10\right)$

as can be seen from equations (7) and (10), the matrix D ═ HH^TIs a 2M order positive definite Hermite matrix, and SNR/P₀In effect the rayleigh entropy of matrix D. According to the Rayleigh entropy eigenvalue most-valued theorem,wherein λ is₁≥λ₂≥…≥λ_mIs matrix D ═ HH^TIs determined by the characteristic value of (a),SNR/P when W is the eigenvector corresponding to the largest eigenvalue of the matrix D₀Maximum, i.e.: SNR_max＝P₀λ₁. Therefore, this parameter estimation algorithm is also referred to as a maximum eigenvalue merging algorithm.

At the time of analysis of optimum T ═ T [ tau ]_K，…，τ₁]Previously, one extreme case of signal-to-noise ratio was analyzed when K is 1, and when B is 1, there is no preprocessing branch, and W is [ α ] from the formulas (7), (9) and (10)₁，…，α_L]^TThis is exactly the optimal combining parameter for a maximal ratio combining RAKE receiver, when the output signal-to-noise ratio isWherein,σ²is the noise variance.

And when B is [ α ]_L，…，α₁]^T，W＝[0，…，1，…，0]^TThen, the mode is a Pre-Rake diversity mode, and the signal-to-noise ratio output under the Pre-Rake diversity modeIt is thus seen that the performance of Pre-Rake diversity is equal to the Rake reception performance.

When K > 1 was analyzed again, rank (HH) was found to be easily known from (7)^T) K because of λ₁≥λ₂≥…≥λ_mTherefore:

${\mathrm{K\λ}}_1\≥\underset{i=1}{\overset{K}{\Σ}}{\mathrm{\λ}}_i=K\underset{i=1}{\overset{L}{\Σ}}|{\mathrm{\α}}_i{|}^2---\left(11\right)$

from the equation, it can be seen that when K > 1, the output snr obtained based on the maximum eigenvector combination method is always greater than the performance of the conventional RAKE receiver, i.e., the lower bound of the output snr is the output snr of the conventional maximum ratio combining RAKE receiver. Obviously, the output signal-to-noise ratio, i.e. the matrix D ═ HH^TThe maximum eigenvalue of D exists at an upper bound, which is now analyzed by matrix eigenvalue estimation theory.

As can be seen from equation (7), the different arrangement order of matrix H results in D ═ HH^TThe maximum eigenvalue D varies. Then the best time delay parameter estimate T ═ τ_K，…，τ₁]Is determined by the following formula:

$\tilde{T}=\mathrm{argmax}\left\{\right|\left|D\right|\left|\right\}---\left(12\right)$

for solving T ═ τ_K，…，τ₁]The best parameter estimation value can be searched by a complex singular value decomposition algorithm. A simpler estimation method is given below. According to the characteristics of the UWB channel, D is a sparse matrix, and D is more than or equal to 1. According to the estimation of the characteristic value of the graph and the disc theorem: the value of D falls within the shaded area as shown in fig. 4.

In the context of figure 4, it is shown,radius of circled_i，jIs the value of ith row and j column of the matrix D. Easily obtained from (7) and centered at d_i，jAlways less than 1, and because the maximum eigenvalue is a real number and | D | | | is always greater than 1 according to equation (11), | D | | | falls within the following range:

1≤||D||≤argmax{d_i，i+R′_i(A)}(13)

wherein D is a sparse matrix, R 'according to UWB channel characteristics'_i(A) Remains substantially unchanged, so the optimum T ═ τ_K，…，τ₁]One requirement for the parameters is:

$\mathrm{argmax}\left\{d_{i,i}\right\}=\underset{i=1}{\overset{K}{\Σ}}{{\stackrel{\‾}{\mathrm{\α}}}_i}^2---\left(14\right)$

whereinRepresentation α₁，…，α_LThe numbers in which K absolute values are largest and different from each other. To satisfy this condition, it is required that the matrix H in equation (7) has N < 2M in a certain column, and the column vector contains the K absolute values of the maximumAt this timeIs/are as followsExactly the time delay of the K strongest paths in the channel, i.e. a simple T ═ τ_K，…，τ₁]Provided is a parameter estimation method.

To this end, the optimal combining and delay parameter B ═ β_K，…，β₁]^T，W＝[γ₁，…，γ_M]^TAnd T ═ τ_K，…，τ₁]The estimation method can be obtained by simply summarizing the following steps:

1. the receiving end obtains channel information through channel estimation, finds out the time delay and parameters of K strongest paths

2. HH of the matrix D according to equations (7) and (10)^TSingular value decomposition is carried out to obtain a characteristic vector corresponding to the maximum characteristic value, namely the optimal parameter estimation value

3. According to formula (B) ═ W^TH)^TCalculating B to obtain the best estimated valueAnd will estimate the parametersAndand transmitting the data to a transmitting end for delay merging pretreatment.

4. The receiving end according to the time delay parameterCombining the channel estimation parameters to obtain the delay parameters and pass weightsCoefficient of weightAnd combining to obtain the final output.

Wherein the matrix singular value decomposition is obtained by SVD algorithm, and D is HH according to formula (8)^TDecomposing the mixture into D-U Λ V^TThe matrix Λ ═ diag [ λ [ ]₁，…，λ_K+L-1]Wherein λ is₁The parameter estimation value of W is the first column vector value of the matrix U. Because the UWB channel matrix D is a sparse matrix, singular value decomposition can be realized by a simplified SVD algorithm, and the complexity of operation is greatly reduced.

Meanwhile, it should be noted that in the singular value decomposition of actual parameter estimation, the estimated value may generate a "flipping effect" to cause misjudgment. I.e. obtaining an estimate from the SVD decomposition valuesBut the actual parameter estimate should beAnd the symbol transition is random, if the problem is not processed, a decision error occurs, and a symbol flipping effect is generated. This problem occurs because equation (9) takes an absolute value when the output signal-to-noise ratio is extremized, and thusAre solutions of equations, while the actual solution is only one. The solution is to estimate the valueMultiplied by a decision factorThe correct value is obtained by correction.

The above-mentioned embodiments are intended to illustrate the objects, technical solutions and advantages of the present invention in further detail, and it should be understood that the above-mentioned embodiments are only exemplary embodiments of the present invention, and are not intended to limit the present invention, and any modifications, equivalents, improvements and the like made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims (10)

1. A signal processing system for an ultra-wideband system, the ultra-wideband system including a transmitter and a receiver, the signal processing system comprising:

the signal preprocessing unit is used for performing branch processing on the sending signal at the sending end to obtain signals of a plurality of branches;

a multipath channel for transmitting signals of the plurality of branches from a transmitting end to a receiving end;

and the signal combining unit is used for combining the signals of the plurality of branches at a receiving end.

2. The signal processing system for an ultra-wideband system of claim 1, wherein the transmitted signal is expressed by:

$\begin{array}{c}{s}_i\left(t\right)=\sqrt{P}\underset{i=0}{\overset{\mathrm{\&infin;}}{\mathrm{\&Sigma;}}}{b}_{i}w\left(t-{\mathrm{iT}}_s\right),\\ w\left(t\right)=\underset{n=0}{\overset{N_s-1}{\&Sigma;}}{c}_{n}g(t-{\mathrm{nT}}_c),\end{array}$

wherein i represents the ith time point number, n represents the nth number in one symbol period, b_iIndicating the character signal numbered i, c_nFor spreading codes, P is the energy of the transmitted signal, and a symbol is formed by N_sA composition of pulses, g (T) being a function of the waveform of the transmitted pulse, T_sRepresents one symbol period;

the signal preprocessing unit is used for preprocessing the signal according to a time delay parameter T ═ tau_K，…，τ₁]The transmitting signal is processed in a branch way, and the branch time delay of the kth branch is T_s-τ_kWherein, τ_kRepresenting the kth multipath delay.

3. The signal processing system for an ultra-wideband system of claim 2, wherein the sampled output values of the signals of the plurality of branches after passing through the multipath channel are vector represented as:

Y＝[y₁(i)，…，y_M(i)]^T＝Ab_iHB+N，

wherein, y_M(i) Denotes the sample output vector value of the M-th column with sample number i, a denotes the fixed gain of the system, H is a matrix of 2M × K, and M ═ T_s/T_b，

$H^T=\left(\begin{array}{c}0,\mathrm{....0...},\mathrm{0...},{\mathrm{\&alpha;}}_1,\mathrm{...0...},{\mathrm{\&alpha;}}_2,\mathrm{...0...}{\mathrm{\&alpha;}}_L\mathrm{...},0\\ 0,\mathrm{....}{\mathrm{\&alpha;}}_1\mathrm{...0...},{\mathrm{\&alpha;}}_2,\mathrm{....0},\mathrm{........}{\mathrm{\&alpha;}}_L\mathrm{...........},0\\ \mathrm{......},\mathrm{..........}\\ 0,\mathrm{...}{\mathrm{\&alpha;}}_{K-1}\mathrm{..0...}{\mathrm{\&alpha;}}_K,\mathrm{......0},\mathrm{..}{\mathrm{\&alpha;}}_L\mathrm{...0..........},0\end{array}\right),$

B＝[β_K，…，β₁]^TRepresenting a merged vector, β_KDenotes the K-th vector, α_lRepresenting the path fading parameters of the l-th path in a multi-path channel, α_lIn the case of a real number,l represents a total of L multipaths and N represents the quantization level of the noise.

4. The signal processing system according to claim 3, wherein the signal combining unit combines the signals at the receiving end according to a branch combining weight W, the branch combining weight vector representing W ═ γ₁，…，γ_M]^T，γ_MThe mth vector weight value, numbered M, is represented, and the combined signal is represented as:

z(i)＝W^TY＝Ab_mW^THB+W^TN。

5. the signal processing system for an ultra-wideband system of claim 4, further comprising:

a parameter estimation unit, configured to calculate a delay parameter T, a combining vector B, and a branch combining weight W, where the delay parameter T has an expression:

T＝argmax{||D||}，

wherein D is HH^TAnd according to the estimation of the characteristic value of the graph and the disc theorem, obtaining:

1≤||D||≤argmax{d_i，i+R′_i(A)}，

wherein,d_i，idenotes the center of a circle, R'_i(A) The radius of the circle is represented, and,

$\arg \max \left\{d_{i,i}\right\}=\underset{i=1}{\overset{K}{\&Sigma;}}{{\stackrel{\&OverBar;}{\mathrm{\&alpha;}}}_i}^2,$

representation α₁，…，α_LThe numbers in which K absolute values are the largest and different from each other;

the branch merging weight W is a feature vector corresponding to the maximum feature value D;

the expression of the merging vector B is as follows:

B＝(W^TH)^T。

6. a signal processing method for an ultra-wideband system, wherein the ultra-wideband system comprises a transmitting end and a receiving end, and the signal processing method comprises:

s1, branching the sending signal at the sending end to obtain signals of a plurality of branches;

s2, transmitting the signals of the multiple branches from the transmitting end to the receiving end;

and S3, combining the signals of the plurality of branches at the receiving end.

7. The signal processing method for an ultra-wideband system of claim 6, wherein the expression of the transmitted signal is:

$\begin{array}{c}{s}_i\left(t\right)=\sqrt{P}\underset{i=0}{\overset{\mathrm{\&infin;}}{\mathrm{\&Sigma;}}}{b}_{i}w\left(t-{\mathrm{iT}}_s\right),\\ w\left(t\right)=\underset{n=0}{\overset{N_s-1}{\&Sigma;}}{c}_{n}g(t-{\mathrm{nT}}_c),\end{array}$

wherein i represents the ith time point number, n represents the nth number in one symbol period, b_iIndicating the character signal numbered i, c_nFor spreading codes, P is the energy of the transmitted signal, and a symbol is formed by N_sA composition of pulses, g (T) being a function of the waveform of the transmitted pulse, T_sRepresents one symbol period;

the signal preprocessing unit is used for preprocessing the signal according to a time delay parameter T ═ tau_K，…，τ₁]The transmitting signal is processed in a branch way, and the branch time delay of the kth branch is T_s-τ_kWherein, τ_kRepresenting the kth multipath delay.

8. The signal processing method for an ultra-wideband system of claim 7, wherein sampled output values of signals of the plurality of branches after passing through the multipath channel are represented as:

Y＝[y₁(i)，…，y_M(i)]^T＝Ab_iHB+N，

y_M(i) denotes the sample output vector value of the M-th column with sample number i, a denotes the fixed gain of the system, H is a matrix of 2M × K, and M ═ T_s/T_b，

$H^T=\left(\begin{array}{c}0,\mathrm{....0...},\mathrm{0...},{\mathrm{\&alpha;}}_1,\mathrm{...0...},{\mathrm{\&alpha;}}_2,\mathrm{...0...}{\mathrm{\&alpha;}}_L\mathrm{...},0\\ 0,\mathrm{....}{\mathrm{\&alpha;}}_1\mathrm{...0...},{\mathrm{\&alpha;}}_2,\mathrm{....0},\mathrm{........}{\mathrm{\&alpha;}}_L\mathrm{...........},0\\ \mathrm{......},\mathrm{..........}\\ 0,\mathrm{...}{\mathrm{\&alpha;}}_{K-1}\mathrm{..0...}{\mathrm{\&alpha;}}_K,\mathrm{......0},\mathrm{..}{\mathrm{\&alpha;}}_L\mathrm{...0..........},0\end{array}\right),$

B＝[β_K，…，β₁]^TRepresenting a merged vector, β_KDenotes the K-th vector, α_lRepresenting the path fading parameters of the l-th path in a multi-path channel, α_lIn the case of a real number,l represents a total of L multipaths and N represents the quantization level of the noise.

9. The signal processing method for an ultra-wideband system according to claim 8, wherein the step S3 combines signals at a receiving end according to a branch combining weight W, the branch combining weight vector representing W ═ γ₁，…，γ_M]^T，γ_MThe mth vector weight value, numbered M, is represented, and the combined signal is represented as:

z(i)＝W^TY＝Ab_mW^THB+W^TN。

10. the signal processing method for an ultra-wideband system of claim 9, further comprising:

s0, calculating a delay parameter T, a merging vector B and a branch merging weight W, wherein the expression of the delay parameter T is as follows:

T＝argmax{||D||}，

wherein D is HH^TAnd according to the estimation of the characteristic value of the graph and the disc theorem, obtaining:

1≤||D||≤argmax{d_i，i+R′_i(A)}，

wherein d is_i，iDenotes the center of a circle, R'_i(A) The radius of the circle is represented, and,

$\arg \max \left\{d_{i,i}\right\}=\underset{i=1}{\overset{K}{\&Sigma;}}{{\stackrel{\&OverBar;}{\mathrm{\&alpha;}}}_i}^2,$

representation α₁，…，α_LThe numbers in which K absolute values are the largest and different from each other;

the branch merging weight W is a feature vector corresponding to the maximum feature value D;

the expression of the merging vector B is as follows:

B＝(W^TH)^T。