CN100512083C - Uplink receiving method for array antenna MT-CDMA system - Google Patents

Abstract

This invention relates to an up-link receiving method for an array antenna MT-CDMA system, which first of all separates sub-carrier signals from the received signals by every array element of the antenna, then de-spreads the signals and filters them to realize the de-spread of the user signal and separation of its sub-carrier multi-path signals to get the matched filter output of the multi-path signals, then, carries out 9 lines of vacuum domain merge to signals belonging to a same multi-path of a same sub-carrier of a same user, after that caries out time domain merge to multi-path signals of a same sub-carrier of a same user, finally judges the final judged variable to the merged user sub-carrier signal to get the judge result of the signal.

Description

A kind of array antenna MT-CDMA system method for receiving uplink

Technical field

The invention belongs to MT-CDMA cell mobile communication systems array antenna technical field.

Background technology

The scheme that multi-transceiver technology combines with CDMA technology mainly contains CDMA multiple carrier (MC-CDMA), multi-carrier direct sequence spectrum CDMA (multicarrier DS-CDMA) and three kinds of principal modes of multitone modulation CDMA (MT-CDMA).Usually, MC-CDMA is identical with the length of frequency expansion sequence with the sub-carrier number of multicarrier DS-CDMA scheme, and the MT-CDMA scheme adopts longer frequency expansion sequence and than the sub-carrier number of frequency expansion sequence length much less, two kinds of schemes of MT-CDMA scheme and other are compared to have and are suppressed that multiple access disturbs, performance is good and emission, advantage that the receiver complexity is low like this, are one of the most competitive schemes of future mobile communication system.

At present, the MT-CDMA system up-link also adopts single antenna to receive, and does not utilize the spatial domain redundant information, and the receptivity of system is very limited.

Summary of the invention

The present invention merges method of reseptance when having proposed a kind of array antenna MT-CDMA system uplink null for solving the problems of the technologies described above, this method merges by spatial domain, time domain and realized spatial domain, time domain diversity reception, has good receptivity.

The technical solution adopted in the present invention is that the signal that at first each array element of array antenna is received carries out the sub-carrier signal separation; Secondly, the sub-carrier signal of isolated each array element of institute is carried out despreading and matched filter processing, realize the despreading of subscriber signal and separating of user's subcarrier multipath signal, and the matched filtering that obtains user's subcarrier multipath signal is exported; Then, the signal that belongs to the same multipath of the same subcarrier of same user in the matched filter output signal on the different array elements is carried out the spatial domain merge, obtain the spatial domain diversity gain; Afterwards, each multipath signal that belongs to the same subcarrier of same user in the signal after more resultant spatial domain being merged carries out time domain and merges, and obtains the time domain diversity gain; At last, the conclusive judgement variable of resultant user's sub-carrier signal after spatial domain, time domain merge is adjudicated, obtain the court verdict of user's sub-carrier signal.

Below method of the present invention is discussed.

1, the separation of sub-carrier signal

Investigate the array antenna MT-CDMA system up link, array antenna is adopted in the base station, and travelling carriage adopts single antenna.Travelling carriage uplink scheme adopts traditional single antenna MT-CDMA transmission plan and QPSK modulation system.

K mobile subscriber arranged in certain cellular cell in the supposing the system.Each subcarrier of MT-CDMA transmission plan is carried independent user information, and the signal that l subcarrier of such k user launched can be expressed as:

$S_{k,l}\left(t\right)=\sqrt{p_{k,l}}(b_{k,l}^I\left(t\right)c_k^I\left(t\right)\cos \left({\mathrm{\ω}}_{l}t\right)+b_{k,l}^Q\left(t\right)c_k^Q\left(t\right)\sin \left({\mathrm{\ω}}_{l}t\right))$ [formula 1]

In the formula, P _{K, l}Represent k user l (signal power of individual carrier wave of 1≤l≤L), ω _lBe the angular frequency of l subcarrier, With

Be homophase and quadrature signal component, bit period is T _L,

With

Frequency expansion sequence for correspondence is expressed as follows:

$c_k^I\left(t\right)=\underset{q=1}{\overset{G}{\mathrm{\Σ}}}{c}_{k,q}^{I}g(t-q{T}_c)$ [formula 2]

$c_k^Q\left(t\right)=\underset{q=1}{\overset{G}{\mathrm{\Σ}}}{c}_{k,q}^{Q}g(t-q{T}_c)$

In the formula, c _{K, q}∈ 1 ,+1} (q=1 ..., G) being its spreading code, g (t) is that width is T _cCut general pulse, G is defined as G=T _L/ T _cSpreading gain, its frequency spreading wave has normalized energy, promptly ${\∫}_0^{T_L}{\left|c_k\left(t\right)\right|}^2\mathrm{dt}=1.$

Be without loss of generality, the assumed wireless channel is a frequency selective fading channels.Owing to adopt long frequency expansion sequence, therefore each subcarrier of arbitrary user experiences identical decline to the MT-CDMA transmission plan, and the multiple low-pass impulse response of k subscriber channel can be expressed as:

$h_{k,p}\left(t\right)=\underset{p=1}{\overset{P}{\mathrm{\Σ}}}{\mathrm{\ρ}}_{k,p}{e}^{-j{\mathrm{\φ}}_{k,p}}\mathrm{\δ}(t-{\mathrm{\τ}}_{k,p})$ [formula 3]

Wherein, P is the distinguishable multipath number of subscriber signal,

Be the multiple Gaussian random variable of zero-mean, variance is

τ _{K, p}=(p-1) T _c+ Δ _{K, p}Be the time delay of user k p bar multipath, the time delay of its p bar multipath all is identical concerning all carrier waves of same user, Δ _{K, p}Be the independent same distribution stochastic variable, [0, T _c) between evenly distribute.

Suppose that the base station adopts equidistantly linear battle array, like this base station array antenna n (=1 ..., N) signal that receives on the individual array element is:

[formula 4]

Wherein, $a_{k,p,n}=\exp [-j\frac{2\mathrm{\πd}}{\mathrm{\λ}}(n-1)\sin \left({\mathrm{\θ}}_{k,p}\right)]$ Be the array response of n array element, λ is a carrier wavelength, and d is the distance between the adjacent array element, θ _{K, p}Be the angle of arrival in p footpath of user k,

Be the phase shift in p footpath of user k, they all are identical concerning the same footpath of same all carrier waves of user.e _n(t) be additive white Gaussian noise, power spectral density is N ₀/ 2.

Like this array antenna received to resultant signal can be expressed as:

[formula 5]

Wherein, e (t)=[e ₁..., e _N] ^TBe corresponding noise vector, () ^TBe the transposition computing.The array vector of p multipath of k user is:

a _{K, p}=[a _{K, p, 1}, a _{K, p, 2}..., a _{K, p, N}] ^T[formula 6]

Each subcarrier of MT-CDMA transmission plan is carried independent user information, and the testing process of the arbitrary sub-carrier signal of arbitrary user all is identical, and therefore the detection of the arbitrary sub-carrier signal of arbitrary user only is discussed.Be without loss of generality, suppose the 1st user l ' (1≤l '≤L) individual subcarrier signals is a desired signal.

The signal times that each array element is received is with the subcarrier cos (ω corresponding with the transmitting terminal modulation system _{L '}T), sin (ω _{L '}T) combination just can realize the separation of each sub-carrier signal.To isolated the 1st the individual subcarrier signals of user l ' of QPSK modulation system be:

x(t)＝r(t)(cos(ω _l′t)+sin(ω _l′t))＝[x ₁(t)，x ₂(t)，…，x _N(t)] ^T

＝[r ₁(t)(cos(ω _l′t)+sin(ω _l′t))，r ₂(t)(cos(ω _l′t)+sin(ω _l′t))，…，r _N(t)(cos(ω _l′t)+sin(ω _l′t))] ^T

[formula 7]

2, despreading and matched filter processing

Isolated sub-carrier signal is carried out despreading and matched filter processing, realize the despreading of subscriber signal and separating of user's subcarrier multipath signal, obtain the matched filtering output of user's subcarrier multipath signal.Base station array antenna n (=1 ..., N) individual array element to the 1st the individual carrier wave of user l ' p ' (1≤p '≤P) matched filtering of i bit of individual multipath signal is output as:

${\mathrm{<figure-callout\; id="1"\; label="y"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">y</figure-callout>}}_{1,l\&prime;,p\&prime;,n}\left(i\right)={\&Integral;}_{(i-1)T_L+{\mathrm{\&tau;}}_{1,p\&prime;}}^{T_L+{\mathrm{\&tau;}}_{1,p\&prime;}}{r}_n\left(t\right)c_1^I(t-{\mathrm{\&tau;}}_{1,p\&prime;})(\cos \left({\mathrm{\&omega;}}_{l\&prime;}t\right)+\sin \left({\mathrm{\&omega;}}_{l\&prime;}t\right))\mathrm{dt}$

$+j{\&Integral;}_{(i-1)T_L+{\mathrm{\&tau;}}_{1,p\&prime;}}^{T_L+{\mathrm{\&tau;}}_{1,p\&prime;}}{r}_n\left(t\right)c_1^Q(t-{\mathrm{\&tau;}}_{1,p\&prime;})(\cos \left({\mathrm{\&omega;}}_{l\&prime;}t\right)+\sin \left({\mathrm{\&omega;}}_{l\&prime;}t\right))\mathrm{dt}$

[formula 8]

All array elements of base station to the 1st the individual carrier wave of user l ' p ' (1≤p '≤P) matched filtering of i bit of individual multipath signal is output as:

${\mathrm{<figure-callout\; id="1"\; label="y"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">y</figure-callout>}}_{1,l\&prime;,p\&prime;}\left(i\right)={[y_{1,l\&prime;,p\&prime;,1},\&CenterDot;\&CenterDot;\&CenterDot;,y_{1,l\&prime;,p\&prime;,N}]}^T$

$={\&Integral;}_{(i-1)T_L+{\mathrm{\&tau;}}_{1,p\&prime;}}^{T_L+{\mathrm{\&tau;}}_{1,p\&prime;}}r\left(t\right)c_1^I(t-{\mathrm{\&tau;}}_{1,p\&prime;})(\cos \left({\mathrm{\&omega;}}_{l\&prime;}t\right)+\sin \left({\mathrm{\&omega;}}_{l\&prime;}t\right))\mathrm{dt}$

$+j{\&Integral;}_{(i-1)T_L+{\mathrm{\&tau;}}_{1,p\&prime;}}^{T_L+{\mathrm{\&tau;}}_{1,p\&prime;}}r\left(t\right)c_1^Q(t-{\mathrm{\&tau;}}_{1,p\&prime;})(\cos \left({\mathrm{\&omega;}}_{l\&prime;}t\right)+\sin \left({\mathrm{\&omega;}}_{l\&prime;}t\right))\mathrm{dt}$ [formula 9]

$=\sqrt{{\mathrm{<figure-callout\; id="1"\; label="p"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">p</figure-callout>}}_{1,l\&prime;}}{T}_L({\mathrm{<figure-callout\; id="1"\; label="b"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">b</figure-callout>}}_{1,l\&prime;}^I\left(i\right)+{\mathrm{<figure-callout\; id="1"\; label="jb"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">jb</figure-callout>}}_{1,l\&prime;}^Q\left(i\right))\cos \left({\mathrm{\&phi;}}_{1,p\&prime;}\right){\mathrm{\&rho;}}_{1,p\&prime;}{a}_{1,l\&prime;}\left(i\right)+I_{1,l\&prime;}\left(i\right)+{\mathrm{<figure-callout\; id="1"\; label="n"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">n</figure-callout>}}_{1,l\&prime;}\left(i\right)$

Wherein, I _{1, l '}(i), n _{1, l '}(i) represent total interference and noise signal respectively.

The noise of system and interference characteristic to proposed empty the time merge method of reseptance performance have significant effects.Below the kind of system noise and interference is analyzed earlier.

1) noise

Noise item can provide as follows:

${\mathrm{<figure-callout\; id="1"\; label="n"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">n</figure-callout>}}_{1,l\&prime;}\left(i\right)={\mathrm{<figure-callout\; id="1"\; label="n"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">n</figure-callout>}}_{1,l\&prime;}^I\left(i\right)+\mathrm{<figure-callout\; id="1"\; label="j\; n"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">j</figure-callout>}{n}_{}^{}{}_{1,l\&prime;}^Q\left(i\right)$ [formula 10]

Homophase has identical variance with quadrature component

${\mathrm{\&sigma;}}_N^{I2}={\mathrm{\&sigma;}}_N^{Q2}={\mathrm{\&sigma;}}_N^2/2=\frac{N_0}{4T_L}$ [formula 11]

2) disturb

Total interference vector comprises homophase and quadrature component, and can separately be expressed as described different types of interference:

$I_{1,l\&prime;}\left(i\right)=I_{1,l\&prime;}^I\left(i\right)+j{I}_{1,l\&prime;}^Q\left(i\right)$ [formula 12]

$I_{1,l\&prime;}\left(i\right)={\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">m</figure-callout>}}_{1,l\&prime;}^I\left(i\right)+{\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">m</figure-callout>}}_{1,k}^I\left(i\right)+{\mathrm{<figure-callout\; id="1"\; label="s"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">s</figure-callout>}}_{1,l\&prime;}^I\left(i\right)+{\mathrm{<figure-callout\; id="1"\; label="s"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">s</figure-callout>}}_{1,k}^I\left(i\right)$ [formula 13]

$I_{1,l\&prime;}^Q\left(i\right)={\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">m</figure-callout>}}_{1,l\&prime;}^Q\left(i\right)+{\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">m</figure-callout>}}_{1,k}^Q\left(i\right)+{\mathrm{<figure-callout\; id="1"\; label="s"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">s</figure-callout>}}_{1,l\&prime;}^Q\left(i\right)+{\mathrm{<figure-callout\; id="1"\; label="s"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">s</figure-callout>}}_{1,k}^Q\left(i\right)$ [formula 14]

(1) from other user's of same carrier wave interference,

With

The multiple access that this interference is similar in traditional single carrier cdma system disturbs,

With

Can provide as follows:

[formula 15]

$\&CenterDot;\underset{q=1}{\overset{\mathrm{<figure-callout\; id="1"\; label="G\; c"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">G</figure-callout>}}{\mathrm{\&Sigma;}}}{c}_{}^{}{}_{1,q}^I[(b_{k,l\&prime;,(q-1)}^I+{\mathrm{jb}}_{k,l\&prime;,(q-1)}^Q)R_g\left({\mathrm{\&tau;}}_{k,p}^{\&prime;}\right)+(b_{k,l\&prime;,q}^I+{\mathrm{jb}}_{k,l\&prime;,q}^Q)R_g(T_c-{\mathrm{\&tau;}}_{k,p}^{\&prime;})]$

[formula 16]

$\&CenterDot;\underset{q=1}{\overset{\mathrm{<figure-callout\; id="1"\; label="G\; c"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">G</figure-callout>}}{\mathrm{\&Sigma;}}}{c}_{}^{}{}_{1,q}^Q[(b_{k,l\&prime;,(q-1)}^I+{\mathrm{jb}}_{k,l\&prime;,(q-1)}^Q)R_g\left({\mathrm{\&tau;}}_{k,p}^{\&prime;}\right)+(b_{k,l\&prime;,q}^I+{\mathrm{jb}}_{k,l\&prime;,q}^Q)R_g(T_c-{\mathrm{\&tau;}}_{k,p}^{\&prime;})]$

In the formula, R _gBe the partial auto correlation of chip waveform (τ), be defined as follows:

$R_g\left(\mathrm{\&tau;}\right)={\&Integral;}_0^{\mathrm{\&tau;}}g(t+T_c-\mathrm{\&tau;})g\left(t\right)\mathrm{dt},0\&le;\mathrm{\&tau;}\&le;T_c$ [formula 17]

Wherein,

P multipath of k user be with respect to the relative time delay of the individual multipath of first user p ',

It is the product of data bit and spreading code.

(2) from the interference on other carrier wave of other user,

With

To the MT-CDMA system,, introduced between subcarrier and disturbed because multipath fading has destroyed the orthogonality between the subcarrier.

With

Can provide as follows:

${\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">m</figure-callout>}}_{1,k}^I\left(i\right)=\underset{k=2}{\overset{K}{\mathrm{\&Sigma;}}}\underset{\underset{l\&NotEqual;l\&prime;}{l=1}}{\overset{L}{\mathrm{\&Sigma;}}}\underset{p=1}{\overset{P}{\mathrm{\&Sigma;}}}{a}_{k,p}\sqrt{P_{k,l}}{\mathrm{\&rho;}}_{k,p}\underset{q=1}{\overset{\mathrm{<figure-callout\; id="1"\; label="G\; c"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">G</figure-callout>}}{\mathrm{\&Sigma;}}}{c}_{}^{}{}_{1,q}^I[(b_{k,l,(q-1)}^I+{\mathrm{jb}}_{k,l,(q-1)}^Q)L_g\left({\mathrm{\&tau;}}_{k,p}^{\&prime;}\right)$ [formula 18]

$+(b_{k,l,q}^I+{\mathrm{jb}}_{k,l,q}^Q)L_g(T_c-{\mathrm{\&tau;}}_{k,p}^{\&prime;})]$

${\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">m</figure-callout>}}_{1,k}^Q\left(i\right)=\underset{k=2}{\overset{K}{\mathrm{\&Sigma;}}}\underset{\underset{l\&NotEqual;l\&prime;}{l=1}}{\overset{L}{\mathrm{\&Sigma;}}}\underset{p=1}{\overset{P}{\mathrm{\&Sigma;}}}{a}_{k,p}\sqrt{P_{k,l}}{\mathrm{\&rho;}}_{k,p}\underset{q=1}{\overset{\mathrm{<figure-callout\; id="1"\; label="G\; c"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">G</figure-callout>}}{\mathrm{\&Sigma;}}}{c}_{}^{}{}_{1,q}^Q[(b_{k,l,(q-1)}^I+{\mathrm{jb}}_{k,l,(q-1)}^Q)L_g\left({\mathrm{\&tau;}}_{k,p}^{\&prime;}\right)$ [formula 19]

$+(b_{k,l,q}^I+{\mathrm{jb}}_{k,l,q}^Q)L_g(T_c-{\mathrm{\&tau;}}_{k,p}^{\&prime;})]$

Wherein,

[formula 20]

(3) from the self-interference of same carrier wave,

With

With

Representative can provide as follows respectively from the self-interference of the multipath signal on the same sub-carrier:

[formula 21]

$\underset{q=1}{\overset{\mathrm{<figure-callout\; id="1"\; label="G\; c"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">G</figure-callout>}}{\mathrm{\&Sigma;}}}{c}_{}^{}{}_{1,q}^I[({\mathrm{<figure-callout\; id="1"\; label="b"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">b</figure-callout>}}_{1,l\&prime;,(q-1)}^I+{\mathrm{<figure-callout\; id="1"\; label="jb"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">jb</figure-callout>}}_{1,l\&prime;,(q-1)}^Q)R_g\left({\mathrm{\&tau;}}_{1,p}^{\&prime;}\right)+({\mathrm{<figure-callout\; id="1"\; label="b"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">b</figure-callout>}}_{1,l\&prime;,q}^I+{\mathrm{<figure-callout\; id="1"\; label="jb"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">jb</figure-callout>}}_{1,l\&prime;,q}^Q)R_g(T_c-{\mathrm{\&tau;}}_{1,p}^{\&prime;})]$

[formula 22]

$\underset{q=1}{\overset{\mathrm{<figure-callout\; id="1"\; label="G\; c"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">G</figure-callout>}}{\mathrm{\&Sigma;}}}{c}_{}^{}{}_{1,q}^Q[({\mathrm{<figure-callout\; id="1"\; label="b"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">b</figure-callout>}}_{1,l\&prime;,(q-1)}^I+{\mathrm{<figure-callout\; id="1"\; label="jb"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">jb</figure-callout>}}_{1,l\&prime;,(q-1)}^Q)R_g\left({\mathrm{\&tau;}}_{1,p}^{\&prime;}\right)+({\mathrm{<figure-callout\; id="1"\; label="b"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">b</figure-callout>}}_{1,l\&prime;,q}^I+{\mathrm{<figure-callout\; id="1"\; label="jb"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">jb</figure-callout>}}_{1,l\&prime;,q}^Q)R_g(T_c-{\mathrm{\&tau;}}_{1,p}^{\&prime;})]$

Wherein, It is the relative time delay of p multipath with respect to the individual multipath of p '.

(4) from the self-interference of other subcarrier,

With

With

Representative can provide as follows respectively from the self-interference of the multipath signal on other subcarrier:

${\mathrm{<figure-callout\; id="1"\; label="s"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">s</figure-callout>}}_{1,k}^I\left(i\right)=\underset{\underset{l\&NotEqual;l\&prime;}{l=1}}{\overset{L}{\mathrm{\&Sigma;}}}\underset{p=1}{\overset{P}{\mathrm{\&Sigma;}}}{a}_{1,\mathrm{<figure-callout\; id="1"\; label="p\; P"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">p</figure-callout>}}\sqrt{P_{}}\sqrt{{}_{1,l}}{\mathrm{\&rho;}}_{1,p}\underset{q=1}{\overset{\mathrm{<figure-callout\; id="1"\; label="G\; c"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">G</figure-callout>}}{\mathrm{\&Sigma;}}}{c}_{}^{}{}_{1,q}^I[({\mathrm{<figure-callout\; id="1"\; label="b"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">b</figure-callout>}}_{1,l,(q-1)}^I+{\mathrm{<figure-callout\; id="1"\; label="jb"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">jb</figure-callout>}}_{1,l,(q-1)}^Q)L_g\left({\mathrm{\&tau;}}_{1,p}^{\&prime;}\right)$ [formula 23]

$+({\mathrm{<figure-callout\; id="1"\; label="b"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">b</figure-callout>}}_{1,l,q}^I+{\mathrm{<figure-callout\; id="1"\; label="jb"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">jb</figure-callout>}}_{1,l,q}^Q)L_g(T_c-{\mathrm{\&tau;}}_{1,p}^{\&prime;})]$

${\mathrm{<figure-callout\; id="1"\; label="s"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">s</figure-callout>}}_{1,k}^Q\left(i\right)=\underset{\underset{l\&NotEqual;l\&prime;}{l=1}}{\overset{L}{\mathrm{\&Sigma;}}}\underset{p=1}{\overset{P}{\mathrm{\&Sigma;}}}{a}_{1,\mathrm{<figure-callout\; id="1"\; label="p\; P"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">p</figure-callout>}}\sqrt{P_{}}\sqrt{{}_{1,l}}{\mathrm{\&rho;}}_{1,p}\underset{q=1}{\overset{\mathrm{<figure-callout\; id="1"\; label="G\; c"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">G</figure-callout>}}{\mathrm{\&Sigma;}}}{c}_{}^{}{}_{1,q}^Q[({\mathrm{<figure-callout\; id="1"\; label="b"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">b</figure-callout>}}_{1,l,(q-1)}^I+{\mathrm{<figure-callout\; id="1"\; label="jb"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">jb</figure-callout>}}_{1,l,(q-1)}^Q)L_g\left({\mathrm{\&tau;}}_{1,p}^{\&prime;}\right)$ [formula 24]

$+({\mathrm{<figure-callout\; id="1"\; label="b"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">b</figure-callout>}}_{1,l,q}^I+{\mathrm{<figure-callout\; id="1"\; label="jb"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">jb</figure-callout>}}_{1,l,q}^Q)L_g(T_c-{\mathrm{\&tau;}}_{1,p}^{\&prime;})]$

L _g(τ) such as formula 20 definition.

3, the spatial domain of signal merges

Below the correlation properties of disturbing are further analyzed, draw optimum and suboptimum merging weight that the received signal spatial domain merges.Disturb correlation properties to comprise auto-correlation and their cross correlation.

1) auto-correlation

(1) interference of other user's same carrier wave With

Array vector

With

Element

With

Can be regarded as a series of independent Gaussian stochastic variables.Therefore, their auto-correlation function can approximate evaluation be:

[formula 25]

Wherein, () ^*Be conjugate operation.δ (h) is the delta function.

The variance that can get homophase and quadrature component is:

${\mathrm{\&sigma;}}_{m,1,l\&prime;}^2=\underset{k=2}{\overset{K}{\mathrm{\&Sigma;}}}\underset{p=1}{\overset{P}{\mathrm{\&Sigma;}}}{P}_{k,l\&prime;}{\mathrm{\&sigma;}}_{k,p}^{2}G[R_g^2\left({\mathrm{\&tau;}}_{k,p}^{\&prime;}\right)+R_g^2(T_c-{\mathrm{\&tau;}}_{k,p}^{\&prime;})]$ [formula 26]

(2) interference of other other carrier wave of user,

With

Equally,

With

Auto-correlation be:

[formula 27]

The variance that can get homophase and quadrature component is:

${\mathrm{\&sigma;}}_{m,1,k}^2=\underset{k=2}{\overset{K}{\mathrm{\&Sigma;}}}\underset{\underset{l\&NotEqual;l\&prime;}{l=1}}{\overset{L}{\mathrm{\&Sigma;}}}\underset{p=1}{\overset{P}{\mathrm{\&Sigma;}}}G{\mathrm{\&sigma;}}_{k,p}^2{P}_{k,l}[L_g^2\left({\mathrm{\&tau;}}_{k,p}^{\&prime;}\right)+L_g^2(T_c-{\mathrm{\&tau;}}_{k,p}^{\&prime;})]$ [formula 28]

(3) self-interference of same carrier wave

With

With Auto-correlation be:

[formula 29]

The variance that can get homophase and quadrature component is:

${\mathrm{\&sigma;}}_{s,1,l\&prime;}^2=\underset{\underset{p\&NotEqual;p\&prime;}{p=1}}{\overset{P}{\mathrm{\&Sigma;}}}G{\mathrm{\&sigma;}}_{1,p}^2{\mathrm{<figure-callout\; id="1"\; label="P"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">P</figure-callout>}}_{1,l}[R_g^2\left({\mathrm{\&tau;}}_{1,p}^{\&prime;}\right)+R_g^2(T_c-{\mathrm{\&tau;}}_{1,p}^{\&prime;})]$ [formula 30]

(4) from the self-interference of other carrier wave

With

With

Auto-correlation function be:

[formula 31]

The variance that can get homophase and quadrature component is:

${\mathrm{\&sigma;}}_{s,1,k\&prime;}^2=\underset{\underset{l\&NotEqual;l\&prime;}{l=1}}{\overset{L}{\mathrm{\&Sigma;}}}\underset{p=1}{\overset{P}{\mathrm{\&Sigma;}}}G{\mathrm{\&sigma;}}_{1,p}^2{\mathrm{<figure-callout\; id="1"\; label="P"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">P</figure-callout>}}_{1,l}[L_g^2\left({\mathrm{\&tau;}}_{1,p}^{\&prime;}\right)+L_g^2(T_c-{\mathrm{\&tau;}}_{1,p}^{\&prime;})]$ [formula 32]

Suppose

With

All be independently and be evenly distributed on [0, T _c], g (t) is to be T the duration _cSquare wave, to formula 26,28,30 and 32 respectively [0, T _c] on ask average, can obtain:

${\mathrm{\&sigma;}}_{m,1,l\&prime;}^2=\underset{k=2}{\overset{K}{\mathrm{\&Sigma;}}}\underset{p=1}{\overset{P}{\mathrm{\&Sigma;}}}\frac{2}{3}{T}_c^{2}G{\mathrm{\&sigma;}}_{k,p}^2{P}_{k,l\&prime;}$ [formula 33]

${\mathrm{\&sigma;}}_{m,1,k}^2=\underset{k=2}{\overset{K}{\mathrm{\&Sigma;}}}\underset{\underset{l\&NotEqual;l\&prime;}{l=1}}{\overset{L}{\mathrm{\&Sigma;}}}\underset{p=1}{\overset{P}{\mathrm{\&Sigma;}}}\frac{{\mathrm{GT}}_c^2}{{6\mathrm{\&pi;}}^2{(l-l\&prime;)}^2}{\mathrm{\&sigma;}}_{k,p}^2{P}_{k,l}$ [formula 34]

${\mathrm{\&sigma;}}_{s,1,l\&prime;}^2=\underset{\underset{p\&NotEqual;p\&prime;}{p=1}}{\overset{P}{\mathrm{\&Sigma;}}}\frac{2}{3}G{T}_c^2{{\mathrm{\&sigma;}}_{1,p}^2\mathrm{<figure-callout\; id="1"\; label="P"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">P</figure-callout>}}_{1,l\&prime;}$ [formula 35]

${\mathrm{\&sigma;}}_{s,1,k}^2=\underset{\underset{l\&NotEqual;l\&prime;}{l=1}}{\overset{L}{\mathrm{\&Sigma;}}}\underset{p=1}{\overset{P}{\mathrm{\&Sigma;}}}\frac{{\mathrm{GT}}_c^2}{{6\mathrm{\&pi;}}^2{(l-l\&prime;)}^2}{\mathrm{\&sigma;}}_{1,p}^2{\mathrm{<figure-callout\; id="1"\; label="P"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">P</figure-callout>}}_{1,l}$ [formula 36]

Therefore, total variance sum of disturbing is:

${\mathrm{\&sigma;}}_I^2=2({\mathrm{\&sigma;}}_{m,1,l\&prime;}^2+{\mathrm{\&sigma;}}_{m,1,k}^2+{\mathrm{\&sigma;}}_{s,1,l\&prime;}^2+{\mathrm{\&sigma;}}_{s,1,k}^2)$ [formula 37]

2) cross-correlation

With

Between cross-correlation (perhaps exist

With Between cross-correlation) and With Between cross-correlation (perhaps exist

With

Between cross-correlation) be that the spatial domain of received signal on different array element n and n ' is relevant.It can be derived out by the cross-correlation function of linear array, and the cross-correlation function of linear array is expressed as follows:

$E\left\{a_{k,p}\left(t\right)a_{k,p}^H(t+\mathrm{\&tau;})\right\}=J_0\left(2\mathrm{\&pi;}{f}_v\mathrm{\&tau;}\right)R_{k,p}$ [formula 38]

Wherein, () ^HThe expression conjugate transpose, f _vIt is maximum Doppler frequency-shift.R _{K, p}Be k user p footpath of expression array vector a _{K, p}The N of correlation * N Hermitian Toeplitz matrix, its real part and imaginary part can be expressed as respectively:

$\mathrm{Re}\left\{R_{k,p}(n\&prime;,n)\right\}=J_0\left[r(n\&prime;,n)\right]+2\underset{u=1}{\overset{\&infin;}{\mathrm{\&Sigma;}}}{J}_{2u}\left[r(n\&prime;,n)\right]\&CenterDot;\cos \left({2\mathrm{u\&theta;}}_{k,p}\right)$ [formula 39]

$\mathrm{Im}\left\{R_{k,p}(n\&prime;,n)\right\}=2\underset{u=1}{\overset{\&infin;}{\mathrm{\&Sigma;}}}{J}_{2u+1}\left[r(n\&prime;,n)\right]\&CenterDot;\left[\sin \left({2u+1)\mathrm{\&theta;}}_{k,p}\right)\right]$ [formula 40]

J _uBe first kind u rank Bessel functions.

R (n ', n)=2 π d|n '-n|/λ [formula 41]

3) total interference plus noise correlation matrix

Definition interference correlation matrix M (h) is:

$M\left(h\right)=E\{I_{1,l\&prime;}^I\left(i\right)I_{1,l\&prime;}^{\mathrm{IH}}(i+h)+I_{1,l\&prime;}^Q\left(i\right)I_{1,l\&prime;}^{\mathrm{QH}}(i+h)\}$ [formula 42]

$=M\mathrm{\&delta;}\left(h\right).$

Matrix M can be derived out by formula 33-38,

$M=2\underset{k=2}{\overset{K}{\mathrm{\&Sigma;}}}\underset{p=1}{\overset{P}{\mathrm{\&Sigma;}}}\frac{2}{3}{T}_c^{2}G{\mathrm{\&sigma;}}_{k,p}^2{P}_{k,l\&prime;}{a}_{k,p}{a}_{k,p}^H+2\underset{k=2}{\overset{K}{\mathrm{\&Sigma;}}}\underset{\underset{l\&NotEqual;l\&prime;}{l=1}}{\overset{L}{\mathrm{\&Sigma;}}}\underset{p=1}{\overset{P}{\mathrm{\&Sigma;}}}\frac{{\mathrm{GT}}_c^2}{{6\mathrm{\&pi;}}^2{(l-l\&prime;)}^2}{\mathrm{\&sigma;}}_{k,p}^2{P}_{k,l}{a}_{k,p}{a}_{k,p}^H$ [formula 43]

$+2\underset{\underset{p\&NotEqual;p\&prime;}{p=1}}{\overset{P}{\mathrm{\&Sigma;}}}\frac{2}{3}{\mathrm{GT}}_c^2{\mathrm{\&sigma;}}_{1,p}^2{\mathrm{<figure-callout\; id="1"\; label="P"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">P</figure-callout>}}_{1,l\&prime;}{a}_{1,p}{a}_{1,p}^H+2\underset{\underset{l\&NotEqual;l\&prime;}{l=1}}{\overset{L}{\mathrm{\&Sigma;}}}\underset{p=1}{\overset{P}{\mathrm{\&Sigma;}}}\frac{{\mathrm{GT}}_c^2}{{6\mathrm{\&pi;}}^2{(l-l\&prime;)}^2}{\mathrm{\&sigma;}}_{1,p}^2{\mathrm{<figure-callout\; id="1"\; label="P"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">P</figure-callout>}}_{1,l}{a}_{1,p}{a}_{1,p}^H$

The distribution function f (θ) that supposes the angle of arrival is equally distributed within [0,2 π], and formula 43 can further be expressed as:

$M=2({\mathrm{\&sigma;}}_{m,1,l\&prime;}^2+{\mathrm{\&sigma;}}_{m,1,k}^2+{\mathrm{\&sigma;}}_{s,1,l\&prime;}^2+{\mathrm{\&sigma;}}_{s,1,k}^2\mathrm{\&sigma;})\&CenterDot;{\&Integral;}_{\mathrm{\&theta;}}a\left(\mathrm{\&theta;}\right)a^H\left(\mathrm{\&theta;}\right)\&CenterDot;f\left(\mathrm{\&theta;}\right)d\left(\mathrm{\&theta;}\right)$ [formula 44]

$={\mathrm{\&sigma;}}_I^2{\&Integral;}_{\mathrm{\&theta;}}R\left(\mathrm{\&theta;}\right)f\left(\mathrm{\&theta;}\right)d\left(\mathrm{\&theta;}\right)$

R (θ) is the matrix in formula 39 and the formula 40, so the element of the row of the n ' row n in the matrix M is:

$m_{n\&prime;,n}={\mathrm{\&sigma;}}_I^2{J}_0\left[r(n\&prime;,n)\right]$ [formula 45]

Therefore, total interference plus noise correlation matrix is:

$R_T=E\left\{(I_{1,l\&prime;}+{\mathrm{<figure-callout\; id="1"\; label="n"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">n</figure-callout>}}_{1,l\&prime;}){(I_{1,l\&prime;}+{\mathrm{<figure-callout\; id="1"\; label="n"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">n</figure-callout>}}_{1,l\&prime;})}^H\right\}=M+{\mathrm{\&sigma;}}_N^{2}I$ [formula 46]

In the formula, I is a unit matrix.

R _TDetermining the optimum weight vector that merges in spatial domain of following formula, promptly the optimum weight vector in the spatial domain of the 1st the individual multipath of user l ' individual carrier wave p ' is:

${\mathrm{<figure-callout\; id="1"\; label="w"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">w</figure-callout>}}_{1,l\&prime;.p\&prime;,\mathrm{opt}}\left(i\right)=\mathrm{\&alpha;}{R}_T^{-1}{a}_{1,p\&prime;}^{*}\left(i\right)$ [formula 47]

Wherein,

Be the complex conjugate of desired user p ' footpath array response, α is a constant, can be taken as 1.

Work as R _TWhen approaching unit matrix, interference plus noise and white noise can be considered to sky the time, promptly have

$R_T=M+{\mathrm{\&sigma;}}_{1,l}^{2}I=({\mathrm{\&sigma;}}_I^2+{\mathrm{\&sigma;}}_N^2)I={\mathrm{\&sigma;}}^{2}I$ [formula 48]

Wherein,

Be defined in formula 37,

It is the noise variance that is shown in formula 11.Like this, can obtain suboptimum spatial domain merging weight vector is:

${\mathrm{<figure-callout\; id="1"\; label="w"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">w</figure-callout>}}_{1,l\&prime;,p\&prime;}^s\left(i\right)=[{\mathrm{<figure-callout\; id="1"\; label="w"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">w</figure-callout>}}_{1,l\&prime;,p\&prime;,1}^s\left(i\right),{\mathrm{<figure-callout\; id="1"\; label="w"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">w</figure-callout>}}_{1,l\&prime;,p\&prime;,2}^s\left(i\right),...,{\mathrm{<figure-callout\; id="1"\; label="w"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">w</figure-callout>}}_{1,l\&prime;,p\&prime;,N}^s\left(i\right)]=\mathrm{\&alpha;}{a}_{1,p\&prime;}^{*}\left(i\right)$ [formula 49]

The present invention only uses the suboptimum spatial domain of having simplified to merge weight, compare with the optimum merging, systematic function descends little, but because the suboptimum spatial domain merges the knowledge that weight only need be known user's multipath array vector, do not merge the contrary of the necessary interference plus noise correlation matrix of weight and do not need to calculate optimum spatial domain, greatly reduce the complexity of weight calculation.The multipath signal of the 1st the individual carrier wave of user l ' through the output vector that the spatial domain merges is:

z _{1, l '}(i)=[z _{1, l ', 1}(i), z _{1, l ', 2}(i) ..., z _{1, l ', P}(i)] ^T[formula 50]

In the following formula, the signal of the 1st the individual multipath of user l ' individual carrier wave p ' merges i the bit in back through the spatial domain and is output as:

${\mathrm{<figure-callout\; id="1"\; label="z"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">z</figure-callout>}}_{1,l\&prime;,p\&prime;}\left(i\right)={\mathrm{<figure-callout\; id="1"\; label="w"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">w</figure-callout>}}_{1,l\&prime;,p\&prime;}^s\left(i\right){\mathrm{<figure-callout\; id="1"\; label="y"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">y</figure-callout>}}_{1,l\&prime;,p\&prime;}\left(i\right)=\sqrt{{\mathrm{<figure-callout\; id="1"\; label="P"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">P</figure-callout>}}_{1,l\&prime;}}{T}_b{\mathrm{\&rho;}}_{1,p\&prime;}({\mathrm{<figure-callout\; id="1"\; label="b"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">b</figure-callout>}}_{1,l\&prime;}^I+{\mathrm{<figure-callout\; id="1"\; label="jb"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">jb</figure-callout>}}_{1,l\&prime;}^Q)\left|a_{1,p\&prime;}\left(i\right)\right|+{\mathrm{\&eta;}}_{1,p\&prime;}\left(i\right)$ [formula 51]

Wherein, $\left|a_{1,p\&prime;}\right|=\frac{a_{1,p\&prime;}^H\&CenterDot;a_{1,p\&prime;}}{\sqrt{a_{1,p\&prime;}^H\&CenterDot;a_{1,p\&prime;}}},$ η _{1, p '}(i) merge for the spatial domain that the back is disturbed and noise item with.

Corresponding homophase and quadrature judgment variables are:

${\mathrm{<figure-callout\; id="1"\; label="z"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">z</figure-callout>}}_{1,l\&prime;,p\&prime;}^I\left(i\right)=\mathrm{Re}\left\{{\mathrm{<figure-callout\; id="1"\; label="z"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">z</figure-callout>}}_{1,l\&prime;,p\&prime;}\left(i\right)\right\}=\sqrt{{\mathrm{<figure-callout\; id="1"\; label="P"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">P</figure-callout>}}_{1,l\&prime;}}{T}_L{\mathrm{\&rho;}}_{1,p\&prime;}{\mathrm{<figure-callout\; id="1"\; label="b"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">b</figure-callout>}}_{1,l\&prime;}^l\left|a_{1,p\&prime;}\left(i\right)\right|+{\mathrm{\&eta;}}_{1,p\&prime;}^I\left(i\right)$ [formula 52]

${\mathrm{<figure-callout\; id="1"\; label="z"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">z</figure-callout>}}_{1,l\&prime;,p\&prime;}^Q\left(i\right)=\mathrm{Im}\left\{{\mathrm{<figure-callout\; id="1"\; label="z"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">z</figure-callout>}}_{1,l\&prime;,p\&prime;}\left(i\right)\right\}=\sqrt{{\mathrm{<figure-callout\; id="1"\; label="P"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">P</figure-callout>}}_{1,l\&prime;}}{T}_L{\mathrm{\&rho;}}_{1,p\&prime;}{\mathrm{<figure-callout\; id="1"\; label="b"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">b</figure-callout>}}_{1,l\&prime;}^Q\left|a_{1,p\&prime;}\left(i\right)\right|+{\mathrm{\&eta;}}_{1,p\&prime;}^Q\left(i\right)$

4, the time domain of signal merges

The output vector z that the present invention merges through the spatial domain the P road multipath signal of each subcarrier _{1, l '}(i)=[z _{1, l ', 1}(i), z _{1, l ', 2}(i) ..., z _{1, l ', P}(i)] ^TMerge weight vector by time domain again $w_{k,l\&prime;}^t\left(i\right)=[w_{k,l\&prime;,1}^t\left(i\right),w_{k,l\&prime;,2}^t\left(i\right),...w_{k,l\&prime;,P}^t\left(i\right)]$ Do further time domain and merge, obtain the time domain diversity gain, and obtain k the final judgment variables of the individual carrier wave of user l '

Time domain merges is undertaken by the high specific merging criterion, and the time domain of the 1st the individual carrier wave of user l ' p ' footpath signal merges weight and is:

${\mathrm{<figure-callout\; id="1"\; label="w"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">w</figure-callout>}}_{1,l\&prime;,p\&prime;}^t\left(i\right)=\left|a_{1,p\&prime;}\left(i\right)\right|$ [formula 53]

Be that time domain merging weight vector is:

${\mathrm{<figure-callout\; id="1"\; label="w"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">w</figure-callout>}}_{1,l\&prime;}^t\left(i\right)=[\left|a_{\mathrm{1,1}}\left(i\right)\right|,\left|a_{\mathrm{1,2}}\left(i\right)\right|,...\left|a_{1,P}\left(i\right)\right|]$ [formula 54]

| a _{1, p '}| estimated when merging, so the weight that time domain merges does not need to reappraise in the signal spatial domain.Like this, can obtain the conclusive judgement variable is:

${\mathrm{<figure-callout\; id="1"\; label="Z"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">Z</figure-callout>}}_{1,l\&prime;}\left(i\right)={\mathrm{<figure-callout\; id="1"\; label="w"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">w</figure-callout>}}_{1,l\&prime;}^{\mathrm{<figure-callout\; id="1"\; label="t\; z"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">t</figure-callout>}}{z}_{}{}_{1,l\&prime;}\left(i\right)=\sqrt{{\mathrm{<figure-callout\; id="1"\; label="P"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">P</figure-callout>}}_{1,l\&prime;}}\underset{p\&prime;=1}{\overset{P}{\mathrm{\&Sigma;}}}{T}_L{\mathrm{\&rho;}}_{1,p\&prime;}({\mathrm{<figure-callout\; id="1"\; label="b"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">b</figure-callout>}}_{1,l\&prime;}^I+{\mathrm{<figure-callout\; id="1"\; label="jb"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">jb</figure-callout>}}_{1,l\&prime;}^Q){\left|a_{1,p\&prime;}\left(i\right)\right|}^2+{\mathrm{\&eta;}}_{1,p\&prime;}^{\&prime;}\left(i\right)$ [formula 55]

Wherein, For spatial domain, time domain merge that the back is disturbed and noise item and.

Corresponding homophase and quadrature judgment variables are:

${\mathrm{<figure-callout\; id="1"\; label="Z"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">Z</figure-callout>}}_{1,l}^I\left(i\right)=\mathrm{Re}\left\{{\mathrm{<figure-callout\; id="1"\; label="Z"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">Z</figure-callout>}}_{1,l\&prime;}\left(i\right)\right\}=\sqrt{{\mathrm{<figure-callout\; id="1"\; label="P"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">P</figure-callout>}}_{1,l\&prime;}}\underset{p\&prime;=1}{\overset{P}{\mathrm{\&Sigma;}}}{T}_L{\mathrm{\&rho;}}_{1,p\&prime;}{\mathrm{<figure-callout\; id="1"\; label="b"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">b</figure-callout>}}_{1,l\&prime;}^l{\left|a_{1,p\&prime;}\left(i\right)\right|}^2+{\mathrm{\&eta;}}_{1,p\&prime;}^{\&prime;I}\left(i\right)$ [formula 56]

${\mathrm{<figure-callout\; id="1"\; label="Z"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">Z</figure-callout>}}_{1,l\&prime;}^Q\left(i\right)=\mathrm{Im}\left\{{\mathrm{<figure-callout\; id="1"\; label="Z"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">Z</figure-callout>}}_{1,l\&prime;}\left(i\right)\right\}=\sqrt{{\mathrm{<figure-callout\; id="1"\; label="P"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">P</figure-callout>}}_{1,l\&prime;}}\underset{p\&prime;=1}{\overset{P}{\mathrm{\&Sigma;}}}{T}_L{\mathrm{\&rho;}}_{1,p\&prime;}{\mathrm{<figure-callout\; id="1"\; label="b"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">b</figure-callout>}}_{1,l\&prime;}^Q{\left|a_{1,p\&prime;}\left(i\right)\right|}^2+{\mathrm{\&eta;}}_{1,p\&prime;}^{\&prime;Q}\left(i\right)$

5, the judgement of signal

The present invention utilizes simple polarity decision method to adjudicate to resulting conclusive judgement variable, obtains the court verdict of user's sub-carrier signal.To the QPSK modulation system, be:

${\mathrm{<figure-callout\; id="1"\; label="b"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">b</figure-callout>}}_{1,l\&prime;}^I\left(i\right)=\left\{\begin{array}{cc}1,& {\mathrm{<figure-callout\; id="1"\; label="Z"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">Z</figure-callout>}}_{1,l\&prime;}^I\left(i\right)>0\\ -1,& {\mathrm{<figure-callout\; id="1"\; label="Z"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">Z</figure-callout>}}_{1,l\&prime;}^I\left(i\right)<0\end{array}\right.$ [formula 57]

${\mathrm{<figure-callout\; id="1"\; label="b"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">b</figure-callout>}}_{1,l\&prime;}^Q\left(i\right)=\left\{\begin{array}{cc}1,& {\mathrm{<figure-callout\; id="1"\; label="Z"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">Z</figure-callout>}}_{1,l\&prime;}^Q\left(i\right)>0\\ -1,& {\mathrm{<figure-callout\; id="1"\; label="Z"\; filenames="C200510012214E00221.png,C200510012214E00232.png"\; state="\{\{state\}\}">Z</figure-callout>}}_{1,l\&prime;}^Q\left(i\right)<0\end{array}\right.$ [formula 58]

Beneficial effect of the present invention:

At first, because isolated sub-carrier signal is carried out despreading and matched filter processing earlier, suppressed the multiple access interference to a great extent, to make the signal processing of receiver based on disturbing repressed signal to carry out, stability and performance that this can improve receiver greatly make the method for reseptance that is proposed have practical outstanding advantage.

Secondly, the spatial domain of signal merges has adopted suboptimum to merge weight, does not need to calculate optimum spatial domain and merges the contrary of the necessary interference plus noise correlation matrix of weight, greatly reduce the complexity of weight calculation, and systematic function descends very little; The time domain of carrying out afterwards merges the high specific merging method that adopted, and merging weight does not need to reappraise, and it is simple that method of reseptance is handled, and is easy to realize.

At last, proposed empty the time merge method of reseptance and merge by spatial domain, time domain and obtained spatial domain, time domain diversity gain, make system have good receptivity.

Description of drawings

Fig. 1 merges the reception structured flowchart when being the arbitrary subcarrier of the arbitrary user of array antenna MT-CDMA system up link (k) (l ') sky;

Fig. 2 is error rate of system theory and simulation result;

Fig. 3 is the not error performance of simultaneity factor of multipath number.

Embodiment

Below in conjunction with accompanying drawing method of the present invention is described in detail.

Method of the present invention is applicable to the array antennas mobile communication system of any employing MT-CDMA transmission plan.

Merge during with reference to the arbitrary subcarrier of arbitrary user (k) (l ') the array antenna MT-CDMA system uplink null of Fig. 1 and receive structured flowchart, a kind of concrete steps of array antenna MT-CDMA system method for receiving uplink comprise:

Step 1, the signal r that array antenna the 1st array element is received ₁(t), the signal r that received of the 2nd array element ₂(t) ..., the signal r that N array element is received _N(t) send into carrier signal separation module (1-1) respectively, (1-2) ..., (1-N) in, N carrier separation module finished the input signal subcarrier cos (ω corresponding with the transmitting terminal modulation system _{L '}T), sin (ω _{L '}T) make up the computing of multiplying each other, realize the separation of each sub-carrier signal, obtain output signal, x ₁(t)=r ₁(t) (cos (ω _{L '}T)+sin (ω _{L '}T)), x ₂(t)=r ₂(t) (cos (ω _{L '}T)+sin (ω _{L '}T)) ..., x _N(t)=r _N(t) (cos (ω _{L '}T)+sin (ω _{L '}T));

Step 2 is with signal x ₁(t) send into P despreading and matched filtering module (2-1-1) respectively, (2-2-1) ..., (2-P-1), carry out separating of despreading and subcarrier multipath signal, obtain the i bit signal:

$y_{k,l\&prime;,\mathrm{1,1}}\left(i\right)={\&Integral;}_{(i-1)T_L+{\mathrm{\&tau;}}_{k,1}}^{i{T}_L+{\mathrm{\&tau;}}_{k,1}}{r}_1\left(t\right)c_k^I(t-{\mathrm{\&tau;}}_{k,1})(\cos \left({\mathrm{\&omega;}}_{l}t\right)+\sin \left({\mathrm{\&omega;}}_{l}t\right))\mathrm{dt}+j{\&Integral;}_{(i-1)T_L+{\mathrm{\&tau;}}_{k,1}}^{{\mathrm{iT}}_L+{\mathrm{\&tau;}}_{k,1}}{r}_1\left(t\right)c_k^Q(t-{\mathrm{\&tau;}}_{k,1})(\cos \left({\mathrm{\&omega;}}_{l}t\right)+\sin \left({\mathrm{\&omega;}}_{l}t\right))\mathrm{dt}$

$y_{k,l\&prime;,\mathrm{2,1}}\left(i\right)={\&Integral;}_{(i-1)T_L+{\mathrm{\&tau;}}_{k,2}}^{i{T}_L+{\mathrm{\&tau;}}_{k,2}}{r}_1\left(t\right)c_k^I(t-{\mathrm{\&tau;}}_{k,2})(\cos \left({\mathrm{\&omega;}}_{l}t\right)+\sin \left({\mathrm{\&omega;}}_{l}t\right))\mathrm{dt}+j{\&Integral;}_{(i-1)T_L+{\mathrm{\&tau;}}_{k,2}}^{{\mathrm{iT}}_L+{\mathrm{\&tau;}}_{k,2}}{r}_1\left(t\right)c_k^Q(t-{\mathrm{\&tau;}}_{k,2})(\cos \left({\mathrm{\&omega;}}_{l}t\right)+\sin \left({\mathrm{\&omega;}}_{l}t\right))\mathrm{dt}$

$y_{k,l\&prime;,P,1}\left(i\right)={\&Integral;}_{(i-1)T_L+{\mathrm{\&tau;}}_{k,P}}^{i{T}_L+{\mathrm{\&tau;}}_{k,P}}{r}_1\left(t\right)c_k^I(t-{\mathrm{\&tau;}}_{k,P})(\cos \left({\mathrm{\&omega;}}_{l}t\right)+\sin \left({\mathrm{\&omega;}}_{l}t\right))\mathrm{dt}+j{\&Integral;}_{(i-1)T_L+{\mathrm{\&tau;}}_{k,P}}^{{\mathrm{iT}}_L+{\mathrm{\&tau;}}_{k,P}}{r}_1\left(t\right)c_k^Q(t-{\mathrm{\&tau;}}_{k,P})(\cos \left({\mathrm{\&omega;}}_{l}t\right)+\sin \left({\mathrm{\&omega;}}_{l}t\right))\mathrm{dt}$

Simultaneously with signal x ₂(t) send into P despreading and matched filtering module (2-1-2) respectively, (2-2-2) ..., (2-P-2), carry out separating of despreading and subcarrier multipath signal, obtain the i bit signal:

$y_{k,l\&prime;,\mathrm{1,2}}\left(i\right)={\&Integral;}_{(i-1)T_L+{\mathrm{\&tau;}}_{k,1}}^{i{T}_L+{\mathrm{\&tau;}}_{k,1}}{r}_1\left(t\right)c_k^I(t-{\mathrm{\&tau;}}_{k,1})(\cos \left({\mathrm{\&omega;}}_{l}t\right)+\sin \left({\mathrm{\&omega;}}_{l}t\right))\mathrm{dt}+j{\&Integral;}_{(i-1)T_L+{\mathrm{\&tau;}}_{k,2}}^{{\mathrm{iT}}_L+{\mathrm{\&tau;}}_{k,1}}{r}_2\left(t\right)c_k^Q(t-{\mathrm{\&tau;}}_{k,1})(\cos \left({\mathrm{\&omega;}}_{l}t\right)+\sin \left({\mathrm{\&omega;}}_{l}t\right))\mathrm{dt}$

$y_{k,l\&prime;,\mathrm{2,2}}\left(i\right)={\&Integral;}_{(i-1)T_L+{\mathrm{\&tau;}}_{k,2}}^{i{T}_L+{\mathrm{\&tau;}}_{k,2}}{r}_2\left(t\right)c_k^I(t-{\mathrm{\&tau;}}_{k,2})(\cos \left({\mathrm{\&omega;}}_{l}t\right)+\sin \left({\mathrm{\&omega;}}_{l}t\right))\mathrm{dt}+j{\&Integral;}_{(i-1)T_L+{\mathrm{\&tau;}}_{k,2}}^{{\mathrm{iT}}_L+{\mathrm{\&tau;}}_{k,2}}{r}_2\left(t\right)c_k^Q(t-{\mathrm{\&tau;}}_{k,2})(\cos \left({\mathrm{\&omega;}}_{l}t\right)+\sin \left({\mathrm{\&omega;}}_{l}t\right))\mathrm{dt}$

$y_{k,l\&prime;,P,2}\left(i\right)={\&Integral;}_{(i-1)T_L+{\mathrm{\&tau;}}_{k,P}}^{i{T}_L+{\mathrm{\&tau;}}_{k,P}}{r}_2\left(t\right)c_k^I(t-{\mathrm{\&tau;}}_{k,P})(\cos \left({\mathrm{\&omega;}}_{l}t\right)+\sin \left({\mathrm{\&omega;}}_{l}t\right))\mathrm{dt}+j{\&Integral;}_{(i-1)T_L+{\mathrm{\&tau;}}_{k,P}}^{{\mathrm{iT}}_L+{\mathrm{\&tau;}}_{k,P}}{r}_2\left(t\right)c_k^Q(t-{\mathrm{\&tau;}}_{k,P})(\cos \left({\mathrm{\&omega;}}_{l}t\right)+\sin \left({\mathrm{\&omega;}}_{l}t\right))\mathrm{dt}$

Simultaneously with signal x _N(t) send into P despreading and matched filtering module (2-1-N) respectively, (2-2-N) ..., (2-P-N), carry out separating of despreading and subcarrier multipath signal, obtain the i bit signal:

$y_{k,l\&prime;,1,N}\left(i\right)={\&Integral;}_{(i-1)T_L+{\mathrm{\&tau;}}_{k,1}}^{i{T}_L+{\mathrm{\&tau;}}_{k,1}}{r}_N\left(t\right)c_k^I(t-{\mathrm{\&tau;}}_{k,1})(\cos \left({\mathrm{\&omega;}}_{l}t\right)+\sin \left({\mathrm{\&omega;}}_{l}t\right))\mathrm{dt}+j{\&Integral;}_{(i-1)T_L+{\mathrm{\&tau;}}_{k,1}}^{{\mathrm{iT}}_L+{\mathrm{\&tau;}}_{k,1}}{r}_N\left(t\right)c_k^Q(t-{\mathrm{\&tau;}}_{k,1})(\cos \left({\mathrm{\&omega;}}_{l}t\right)+\sin \left({\mathrm{\&omega;}}_{l}t\right))\mathrm{dt}$

$y_{k,l\&prime;,2,N}\left(i\right)={\&Integral;}_{(i-1)T_L+{\mathrm{\&tau;}}_{k,2}}^{i{T}_L+{\mathrm{\&tau;}}_{k,2}}{r}_N\left(t\right)c_k^I(t-{\mathrm{\&tau;}}_{k,2})(\cos \left({\mathrm{\&omega;}}_{l}t\right)+\sin \left({\mathrm{\&omega;}}_{l}t\right))\mathrm{dt}+j{\&Integral;}_{(i-1)T_L+{\mathrm{\&tau;}}_{k,2}}^{{\mathrm{iT}}_L+{\mathrm{\&tau;}}_{k,2}}{r}_N\left(t\right)c_k^Q(t-{\mathrm{\&tau;}}_{k,2})(\cos \left({\mathrm{\&omega;}}_{l}t\right)+\sin \left({\mathrm{\&omega;}}_{l}t\right))\mathrm{dt}$

$y_{k,l\&prime;,P,N}\left(i\right)={\&Integral;}_{(i-1)T_L+{\mathrm{\&tau;}}_{k,P}}^{i{T}_L+{\mathrm{\&tau;}}_{k,P}}{r}_N\left(t\right)c_k^I(t-{\mathrm{\&tau;}}_{k,P})(\cos \left({\mathrm{\&omega;}}_{l}t\right)+\sin \left({\mathrm{\&omega;}}_{l}t\right))\mathrm{dt}+j{\&Integral;}_{(i-1)T_L+{\mathrm{\&tau;}}_{k,P}}^{{\mathrm{iT}}_L+{\mathrm{\&tau;}}_{k,P}}{r}_N\left(t\right)c_k^Q(t-{\mathrm{\&tau;}}_{k,P})(\cos \left({\mathrm{\&omega;}}_{l}t\right)+\sin \left({\mathrm{\&omega;}}_{l}t\right))\mathrm{dt}$

Step 3 is with N matched filtering output signal y _{K, l ', 1,1}(i), y _{K, l ', 1,2}(i) ..., y _{K, l ', 1, N}(i) send into the spatial domain and merge module (3-1), obtain vector y _{L, l ', 1}(i)=[y _{K, l ', 1,1}(i), y _{K, l ', 1,2}(i) ..., y _{K, l ', 1, N}(i)] ^T, merge weight vector by the spatial domain $w_{k,l\&prime;,1}^s\left(i\right)=[w_{k,l\&prime;,1,1}^s\left(i\right),w_{k,l\&prime;,1,2}^s\left(i\right),...,w_{k,l\&prime;,1,N}^s\left(i\right)],$ The spatial domain of finishing the 1st footpath of user k signal merges, and obtains the signal after the spatial domain merges $z_{k,l\&prime;,1}\left(i\right)=w_{k,l\&prime;,1}^s\left(i\right)\&CenterDot;y_{k,l\&prime;,1}\left(i\right);$ Simultaneously with N matched filtering output signal y _{K, l ', 2,1}(i), y _{K, l ', 2,2}(i) ..., y _{K, l ', 2, N}(i) send into the spatial domain and merge module (3-2), obtain vector y _{K, l ', 2}(i)=[y _{K, l ', 2,1}(i), y _{K, l ', 2,2}(i) ..., y _{K, l ', 2, N}(i)] ^T, merge weight vector by the spatial domain $w_{k,l\&prime;,2}^s\left(i\right)=[w_{k,l\&prime;,2,1}^s\left(i\right),w_{k,l\&prime;,2,2}^s\left(i\right),...,w_{k,l\&prime;,2,N}^s\left(i\right)],$ The spatial domain of finishing the 2nd footpath of user k signal merges, and obtains the signal after the spatial domain merges $z_{k,l\&prime;,2}\left(i\right)=w_{k,l\&prime;,2}^s\left(i\right)\&CenterDot;y_{k,l\&prime;,2}\left(i\right),$ , simultaneously with N matched filtering output signal y _{K, l ', P, 1}(i), y _{K, l ', P, 2}(i) ..., y _{K, l ', P, N}(i) send into the spatial domain and merge module (3-P), obtain vector y _{K, l ', P}(i)=[y _{K, l ', P, 1}(i), y _{K, l ', P, 2}(i) ..., y _{K, l ', P, N}(i)] ^T, merge weight vector by the spatial domain $w_{k,l\&prime;,P}^s\left(i\right)=[w_{k,l\&prime;,P,1}^s\left(i\right),w_{k,l\&prime;,P,2}^s\left(i\right),...,w_{k,l\&prime;,P,N}^s\left(i\right)],$ The spatial domain of finishing user k P bar footpath signal merges, and obtains the signal after the spatial domain merges $Z_{k,l\&prime;,P}\left(i\right)=w_{k,l\&prime;,P}^s\left(i\right)\&CenterDot;y_{k,l\&prime;,P}\left(i\right);$

Step 4 is with the P bar multipath signal z after the merging of spatial domain _{K, l ', 1}(i), z _{K, l ', 2}(i) ..., z _{K, l ', P}(i) send into time domain and merge module 4, obtain vector z _{K, l '}(i)=[z _{K, l ', 1}(i), z _{K, l ', 2}(i) ..., z _{K, l ', P}(i)] ^T, merge weight vector by time domain $w_{k,l\&prime;}^t\left(i\right)=[w_{k,l\&prime;,1}^t\left(i\right),w_{k,l\&prime;,2}^t\left(i\right),...w_{k,l\&prime;,P}^t\left(i\right)]$ The time domain of finishing multipath signal merges, and obtains the final judgment variables of user k $Z_{k,l\&prime;}\left(i\right)=w_{k,l\&prime;}^t\left(i\right)\&CenterDot;z_{k,l\&prime;}\left(i\right);$

Step 5 is with final judgment variables Z _{K, l '}(i) send into signal decision module 5 and carry out polarity decision, obtain the court verdict of the arbitrary subcarrier of arbitrary user (k) (l ') signal.

The spatial domain merges and to be based on suboptimum and to merge that weight vector carries out, and the weight vector of signal spatial domain, k user l ' individual carrier wave p ' footpath merging is $w_{k,l\&prime;,p\&prime;}^s\left(i\right)=\mathrm{\&alpha;}{a}_{k,p\&prime;}^{*}\left(i\right)$ (α is a constant, can be taken as 1).

Time domain merges is undertaken by the high specific merging criterion, and the time domain of k the individual carrier wave of user l ' merges weight vector and is $w_{k,l\&prime;}^t\left(i\right)=[\left|a_{k,1}\left(i\right)\right|,\left|a_{k,2}\left(i\right)\right|,...\left|a_{k,P}\left(i\right)\right|].$

Fig. 2 and Fig. 3 have provided the performance simulation result of a kind of array antenna MT-CDMA system method for receiving uplink that adopts the present invention's proposition.The array element distance of the even linear battle array that is adopted in emulation is crossed is λ/2, if there is not channel fading, all transmit a signal to and have identical power when reaching the base station so, and multipath power is obeyed evenly and distributed, and Normalized Signal/Noise Ratio is E _b/ N ₀=10dB, the spread processing gain is 64, the angle of arrival of all user's subcarrier Different Diameter is separate, and evenly distributes between [0,2 π].

Fig. 2 has provided theory and the simulation result that does not adopt the inventive method error rate of system when number of users simultaneously.Adopted 3 carrier waves, each carrier wave has 3 multipaths.As can be seen from Figure 2 the theory of error rate of system and simulation result are consistent, and show that employing suboptimum merging method is very little to the performance impact of system.Simultaneously, after adopting array antenna, the performance of system is greatly improved than single-antenna case.

Fig. 3 has provided the error performance that adopts the inventive method system when user's multipath number is respectively 2 and 3.As can be seen from Figure 3, along with the increase of array number and multipath number,, significantly improved the performance of system owing to obtained bigger spatial domain and time diversity gain.

Claims (3)

1. array antenna MT-CDMA system method for receiving uplink to the arbitrary subcarrier l ' of arbitrary user k, is characterized in that comprising following receiving step:

Step 1, the signal r that array antenna the 1st array element is received ₁(t), the signal r that received of the 2nd array element ₂(t) ..., the signal r that N array element is received _N(t) send into carrier signal separation module (1-1) respectively, (1-2) ..., (1-N) in, N carrier separation module finished the input signal subcarrier cos (ω corresponding with the transmitting terminal modulation system _{L '}T), sin (ω _{L '}T) make up the computing of multiplying each other, realize the separation of each sub-carrier signal, obtain output signal, x ₁(t)=r ₁(t) (cos (ω _{L '}T)+sin (ω _{L '}T)), x ₂(t)=r ₂(t) (cos (ω _{L '}T)+sin (ω _{L '}T)) ..., x _N(t)=r _N(t) (cos (ω _{L '}T)+sin (ω _{L '}T));

Wherein, ω _{L '}It is the angular frequency of the individual subcarrier of MT-CDMA signal l ';

Step 2 is with signal x ₁(t) send into P despreading and matched filtering module (2-1-1) respectively, (2-2-1) ..., (2-P-1), carry out separating of despreading and subcarrier multipath signal, obtain the i bit signal:

$y_{k,l\&prime;,\mathrm{1,1}}\left(i\right)={\&Integral;}_{(i-1)T_L+{\mathrm{\&tau;}}_{k,1}}^{{\mathrm{iT}}_L+{\mathrm{\&tau;}}_{k,1}}{r}_1\left(t\right)c_k^1(t-{\mathrm{\&tau;}}_{k,1})(\cos \left({\mathrm{\&omega;}}_{l\&prime;}t\right)+\sin \left({\mathrm{\&omega;}}_{l\&prime;}t\right))\mathrm{dt}+j{\&Integral;}_{(i-1)T_L+{\mathrm{\&tau;}}_{k,1}}^{{\mathrm{iT}}_L+{\mathrm{\&tau;}}_{k,1}}{r}_1\left(t\right)c_k^Q(t-{\mathrm{\&tau;}}_{k,1})(\cos \left({\mathrm{\&omega;}}_{l\&prime;}t\right)+\sin \left({\mathrm{\&omega;}}_{l\&prime;}t\right))\mathrm{dt},$

$y_{k,l\&prime;,\mathrm{2,1}}\left(i\right)={\&Integral;}_{(i-1)T_L+{\mathrm{\&tau;}}_{k,2}}^{{\mathrm{iT}}_L+{\mathrm{\&tau;}}_{k,2}}{r}_1\left(t\right)c_k^1(t-{\mathrm{\&tau;}}_{k,2})(\cos \left({\mathrm{\&omega;}}_{l\&prime;}t\right)+\sin \left({\mathrm{\&omega;}}_{l\&prime;}t\right))\mathrm{dt}+j{\&Integral;}_{(i-1)T_L+{\mathrm{\&tau;}}_{k,2}}^{{\mathrm{iT}}_L+{\mathrm{\&tau;}}_{k,2}}{r}_1\left(t\right)c_k^Q(t-{\mathrm{\&tau;}}_{k,2})(\cos \left({\mathrm{\&omega;}}_{l\&prime;}t\right)+\sin \left({\mathrm{\&omega;}}_{l\&prime;}t\right))\mathrm{dt},$

…，

$y_{k,l\&prime;,P,1}\left(i\right)={\&Integral;}_{(i-1)T_L+{\mathrm{\&tau;}}_{k,P}}^{{\mathrm{iT}}_L+{\mathrm{\&tau;}}_{k,P}}{r}_1\left(t\right)c_k^1(t-{\mathrm{\&tau;}}_{k,P})(\cos \left({\mathrm{\&omega;}}_{l\&prime;}t\right)+\sin \left({\mathrm{\&omega;}}_{l\&prime;}t\right))\mathrm{dt}+j{\&Integral;}_{(i-1)T_L+{\mathrm{\&tau;}}_{k,P}}^{{\mathrm{iT}}_L+{\mathrm{\&tau;}}_{k,P}}{r}_1\left(t\right)c_k^Q(t-{\mathrm{\&tau;}}_{k,P})(\cos \left({\mathrm{\&omega;}}_{l\&prime;}t\right)+\sin \left({\mathrm{\&omega;}}_{l\&prime;}t\right))\mathrm{dt},$

Simultaneously with signal x ₂(t) send into P despreading and matched filtering module (2-1-2) respectively, (2-2-2) ..., (2-P-2), carry out separating of despreading and subcarrier multipath signal, obtain the i bit signal:

$y_{k,l\&prime;,\mathrm{1,2}}\left(i\right)={\&Integral;}_{(i-1)T_L+{\mathrm{\&tau;}}_{k,1}}^{{\mathrm{iT}}_L+{\mathrm{\&tau;}}_{k,1}}{r}_2\left(t\right)c_k^1(t-{\mathrm{\&tau;}}_{k,1})(\cos \left({\mathrm{\&omega;}}_{l\&prime;}t\right)+\sin \left({\mathrm{\&omega;}}_{l\&prime;}t\right))\mathrm{dt}+j{\&Integral;}_{(i-1)T_L+{\mathrm{\&tau;}}_{k,1}}^{{\mathrm{iT}}_L+{\mathrm{\&tau;}}_{k,1}}{r}_2\left(t\right)c_k^Q(t-{\mathrm{\&tau;}}_{k,1})(\cos \left({\mathrm{\&omega;}}_{l\&prime;}t\right)+\sin \left({\mathrm{\&omega;}}_{l\&prime;}t\right))\mathrm{dt},$

$y_{k,l\&prime;,\mathrm{2,2}}\left(i\right)={\&Integral;}_{(i-1)T_L+{\mathrm{\&tau;}}_{k,2}}^{{\mathrm{iT}}_L+{\mathrm{\&tau;}}_{k,2}}{r}_2\left(t\right)c_k^1(t-{\mathrm{\&tau;}}_{k,2})(\cos \left({\mathrm{\&omega;}}_{l\&prime;}t\right)+\sin \left({\mathrm{\&omega;}}_{l\&prime;}t\right))\mathrm{dt}+j{\&Integral;}_{(i-1)T_L+{\mathrm{\&tau;}}_{k,2}}^{{\mathrm{iT}}_L+{\mathrm{\&tau;}}_{k,2}}{r}_2\left(t\right)c_k^Q(t-{\mathrm{\&tau;}}_{k,2})(\cos \left({\mathrm{\&omega;}}_{l\&prime;}t\right)+\sin \left({\mathrm{\&omega;}}_{l\&prime;}t\right))\mathrm{dt},$

…，

$y_{k,l\&prime;,P,2}\left(i\right)={\&Integral;}_{(i-1)T_L+{\mathrm{\&tau;}}_{k,P}}^{{\mathrm{iT}}_L+{\mathrm{\&tau;}}_{k,P}}{r}_2\left(t\right)c_k^1(t-{\mathrm{\&tau;}}_{k,P})(\cos \left({\mathrm{\&omega;}}_{l\&prime;}t\right)+\sin \left({\mathrm{\&omega;}}_{l\&prime;}t\right))\mathrm{dt}+j{\&Integral;}_{(i-1)T_L+{\mathrm{\&tau;}}_{k,P}}^{{\mathrm{iT}}_L+{\mathrm{\&tau;}}_{k,P}}{r}_2\left(t\right)c_k^Q(t-{\mathrm{\&tau;}}_{k,P})(\cos \left({\mathrm{\&omega;}}_{l\&prime;}t\right)+\sin \left({\mathrm{\&omega;}}_{l\&prime;}t\right))\mathrm{dt},$

…，

Simultaneously with signal x _N(t) send into P despreading and matched filtering module (2-1-N) respectively, (2-2-N) ..., (2-P-N), carry out separating of despreading and subcarrier multipath signal, obtain the i bit signal:

$y_{k,l\&prime;,1,N}\left(i\right)={\&Integral;}_{(i-1)T_L+{\mathrm{\&tau;}}_{k,1}}^{{\mathrm{iT}}_L+{\mathrm{\&tau;}}_{k,1}}{r}_N\left(t\right)c_k^1(t-{\mathrm{\&tau;}}_{k,1})(\cos \left({\mathrm{\&omega;}}_{l\&prime;}t\right)+\sin \left({\mathrm{\&omega;}}_{l\&prime;}t\right))\mathrm{dt}+j{\&Integral;}_{(i-1)T_L+{\mathrm{\&tau;}}_{k,1}}^{{\mathrm{iT}}_L+{\mathrm{\&tau;}}_{k,1}}{r}_N\left(t\right)c_k^Q(t-{\mathrm{\&tau;}}_{k,1})(\cos \left({\mathrm{\&omega;}}_{l\&prime;}t\right)+\sin \left({\mathrm{\&omega;}}_{l\&prime;}t\right))\mathrm{dt},$

$y_{k,l\&prime;,2,N}\left(i\right)={\&Integral;}_{(i-1)T_L+{\mathrm{\&tau;}}_{k,2}}^{{\mathrm{iT}}_L+{\mathrm{\&tau;}}_{k,2}}{r}_N\left(t\right)c_k^1(t-{\mathrm{\&tau;}}_{k,2})(\cos \left({\mathrm{\&omega;}}_{l\&prime;}t\right)+\sin \left({\mathrm{\&omega;}}_{l\&prime;}t\right))\mathrm{dt}+j{\&Integral;}_{(i-1)T_L+{\mathrm{\&tau;}}_{k,2}}^{{\mathrm{iT}}_L+{\mathrm{\&tau;}}_{k,2}}{r}_N\left(t\right)c_k^Q(t-{\mathrm{\&tau;}}_{k,2})(\cos \left({\mathrm{\&omega;}}_{l\&prime;}t\right)+\sin \left({\mathrm{\&omega;}}_{l\&prime;}t\right))\mathrm{dt},$

…，

$y_{k,l\&prime;,P,N}\left(i\right)={\&Integral;}_{(i-1)T_L+{\mathrm{\&tau;}}_{k,P}}^{{\mathrm{iT}}_L+{\mathrm{\&tau;}}_{k,P}}{r}_N\left(t\right)c_k^1(t-{\mathrm{\&tau;}}_{k,P})(\cos \left({\mathrm{\&omega;}}_{l\&prime;}t\right)+\sin \left({\mathrm{\&omega;}}_{l\&prime;}t\right))\mathrm{dt}+j{\&Integral;}_{(i-1)T_L+{\mathrm{\&tau;}}_{k,P}}^{{\mathrm{iT}}_L+{\mathrm{\&tau;}}_{k,P}}{r}_N\left(t\right)c_k^Q(t-{\mathrm{\&tau;}}_{k,P})(\cos \left({\mathrm{\&omega;}}_{l\&prime;}t\right)+\sin \left({\mathrm{\&omega;}}_{l\&prime;}t\right))\mathrm{dt},$

During step 2 was various, the bit period of MT-CDMA signal was T _L, With Be respectively the pairing frequency expansion sequence of QPSK modulation system homophase and quadrature signal component, τ _{K, 1}τ _{K, P}It is respectively user k the 1st time delay to P bar multipath;

Step 3 is with N matched filtering output signal y _{K, l ', 1,1}(i), y _{K, l ', 1,2}(i) ..., y _{K, l ', 1, N}(i) send into the spatial domain and merge module (3-1), obtain vector y _{K, l ', 1}(i)=[y _{K, l ', 1,1}(i), y _{K, l ', 1,2}(i) ..., y _{K, l ', 1, N}(i)] ^T, merge weight vector by the spatial domain $w_{k,l\&prime;,1}^s\left(i\right)=[w_{k,l\&prime;,\mathrm{1,1}}^s\left(i\right),w_{k,l\&prime;,\mathrm{1,2}}^s\left(i\right),...,w_{k,l\&prime;,1,N}^s\left(i\right)],$ The spatial domain of finishing the 1st footpath of user k signal merges, and obtains the signal after the spatial domain merges $z_{k,l\&prime;,1}\left(i\right)=w_{k,l\&prime;,1}^s\left(i\right)\&CenterDot;y_{k,l\&prime;,1}\left(i\right);$ Simultaneously with N matched filtering output signal y _{K, l ', 2,1}(i), y _{K, l ', 2,2}(i) ..., y _{K, l ', 2, N}(i) send into the spatial domain and merge module (3-2), obtain vector y _{K, l ', 2}(i)=[y _{K, l ', 2,1}(i), y _{K, l ', 2,2}(i) ..., y _{K, l ', 2, N}(i)] ^T, merge weight vector by the spatial domain $w_{k,l\&prime;,2}^s\left(i\right)=[w_{k,l\&prime;,\mathrm{2,1}}^s\left(i\right),w_{k,l\&prime;,\mathrm{2,2}}^s\left(i\right),...,w_{k,l\&prime;,2,N}^s\left(i\right)],$ The spatial domain of finishing the 2nd footpath of user k signal merges, and obtains the signal after the spatial domain merges $z_{k,l\&prime;,2}\left(i\right)=w_{k,l\&prime;,2}^s\left(i\right)\&CenterDot;y_{k,l\&prime;,2}\left(i\right),$ , simultaneously with N matched filtering output signal y _{K, l ', P, 1}(i), y _{K, l ', P, 2}(i) ..., y _{K, l ', P, N}(i) send into the spatial domain and merge module (3-P), obtain vector y _{K, l ＇, P}(i)=[y _{K, l ＇, P, 1}(i), y _{K, l ', P, 2}(i) ..., y _{K, l ', P, N}(i)] ^T, merge weight vector by the spatial domain $w_{k,l\&prime;,P}^s\left(i\right)=[w_{k,l\&prime;,P,1}^s\left(i\right),w_{k,l\&prime;,P,2}^s\left(i\right),...,w_{k,l\&prime;,P,N}^s\left(i\right)],$ The spatial domain of finishing user k P bar footpath signal merges, and obtains the signal after the spatial domain merges $z_{k,l\&prime;,P}\left(i\right)=w_{k,l\&prime;,P}^s\left(i\right)\&CenterDot;y_{k,l\&prime;,P}\left(i\right);$

Step 4 is with the P bar multipath signal z after the merging of spatial domain _{K, l ', 1}(i), z _{K, l ', 2}(i) ..., z _{K, l ', P}(i) send into time domain and merge module (4), obtain vector z _{K, l '}(i)=[z _{K, l ', 1}(i), z _{K, l ', 2}(i) ..., z _{K, l ', P}(i)] ^T, merge weight vector by time domain $w_{k,l\&prime;}^t\left(i\right)=[w_{k,l\&prime;,1}^t\left(i\right),w_{k,l\&prime;,2}^t\left(i\right),...,w_{k,l\&prime;,P}^t\left(i\right)]$ The time domain of finishing multipath signal merges, and obtains the final judgment variables of user k $Z_{k,l\&prime;}\left(i\right)=w_{k,l\&prime;}^s\left(i\right)\&CenterDot;z_{k,l\&prime;}\left(i\right);$

Step 5 is with final judgment variables Z _{K, l '}(i) send into signal decision module (5) and carry out polarity decision, obtain the court verdict of the arbitrary subcarrier l ' signal of arbitrary user k.

2. according to the described a kind of array antenna MT-CDMA system method for receiving uplink of claim 1, it is characterized in that the spatial domain merges to be based on suboptimum and to merge that weight vector carries out that the weight vector of signal spatial domain, k user l ' individual carrier wave p ' footpath merging is $w_{k,l\&prime;,p\&prime;}^s\left(i\right)={\mathrm{\&alpha;a}}_{k,p\&prime;}^{*}\left(i\right);$

Wherein, a _{K, p '}Be the array vector of the individual multipath of user k p ', () ^*Be conjugate operation, α is a constant, is taken as 1.

3. according to the described a kind of array antenna MT-CDMA system method for receiving uplink of claim 1, it is characterized in that time domain merges is undertaken by the high specific merging criterion, and the time domain merging weight vector of k the individual carrier wave of user l ' is $w_{k,l\&prime;}^t\left(i\right)=[\left|a_{k,1}\left(i\right)\right|,\left|a_{k,2}\left(i\right)\right|,...,|a_{k,P}\left(i\right)|];$

Wherein, a _{K, 1}(i), a _{K, 2}(i) ..., a _{K, P}(i) be user k the 1st, 2 respectively ..., the array response vector of P multipath.

Images