CN1677885A - Diversity receiving method in spread spectrum communication system based on linear frequency regulation series - Google Patents

Abstract

The invention in use for wireless communication includes following steps: obtaining estimated value of channel from data signals of received baseband signals in training sequence signals of spread spectrum communication system; picking up a sample signal in spread spectrum sequence cycle from the said data signal; carrying out conjugate transformation for initial spread spectrum sequence; multiplying the transformed spread spectrum sequence by the said sampled signal, and then carrying out FFT; multiplying obtained Kth value by phase exp(- jk2Pi/N) so as to obtain sliding correlation value between received signal and initial spread spectrum sequence; based on estimated value of channel, carrying out multidiameter combination for sliding correlation value. The transmission data are obtained from spread spectrum sequence corresponding to maximum among combined values. Advantages are: reducing amount of calculation in channel estimation and Rake reception, increasing flexibility.

Description

Based on the diversity receiving method in the spread spectrum communication system of GCL sequences

Technical field

The present invention relates to a kind ofly, relate in particular to channel estimating and Rake method of reseptance in the diversity reception, belong to technical field of wireless communication based on the diversity receiving method in the spread spectrum communication system of GCL sequences.

Background technology

Signal usually will experience the influence of multipath channel in the wireless communication system, causes to exist between the signal of adjacent moment to interfere with each other.Different according to signal bandwidth and system works environment, multipath disturbs the influence that causes also to be not quite similar.Though multipath disturb to exist but its influence is very little in some cases, can not take special measure to resist its influence.Some situation then requires receiver to suppress the influence that multipath disturbs with the method for equilibrium or diversity reception.To the communication system based on spread spectrum, the diversity receiving method that inhibition multipath commonly used disturbs is that Rake receives, and it carries out the diversity merging to the multipath in the received signal, thereby improves the performance of communication system.Receive in the process that merges multipath at Rake, need know the channel gain in every footpath, this just need estimate channel.

In bag transmission wireless communication system, each packet all can have one section training sequence signal that receiver is known in the front when emission, utilize this segment signal receiver can adopt the least square scheduling algorithm to calculate channel estimation value.If training sequence has desirable autocorrelation performance, then use the slip related algorithm just can obtain channel estimation value.Though the slip related algorithm has had very big reduction than the amount of calculation of least-squares algorithm, carrying out utilizing the character of GCL sequences in the communication system of spread spectrum based on GCL sequences, can also further reduce the amount of calculation of channel estimating.In addition, channel estimating must just can be carried out later in the frequency deviation full remuneration in prior art.

Use the communication system of a plurality of frequency expansion sequences for data-modulated, in demodulation, Rake receives and need carry out around reason to each possible transmitting sequence, causes very big computation burden.

Summary of the invention

The objective of the invention is to propose a kind of based on the diversity receiving method in the spread spectrum communication system of GCL sequences, reducing the amount of calculation that channel estimating and Rake receive, and the flexibility when improving channel estimating and realizing.By utilizing the character of GCL sequences, channel estimating can be in received signal frequency deviation do not have to carry out under the situation of full remuneration, and can use fast fourier transform algorithm to reduce amount of calculation in the estimation procedure.And the character of cyclic shift between the frequency expansion sequence when utilizing data-modulated, Rake receives and also can use simpler structure and fast fourier transform algorithm to realize.

The present invention propose based on the diversity receiving method in the spread spectrum communication system of GCL sequences, may further comprise the steps:

(1) from spread spectrum communication system, obtains channel estimation value in the training sequence signal of baseband receiving signals;

(2) from spread spectrum communication system, take out the sampled signal in a frequency expansion sequence cycle in the data-signal of baseband receiving signals;

(3) the initial frequency expansion sequence to spread spectrum communication system carries out conjugater transformation, and the initial frequency expansion sequence after the conjugater transformation and the sampled signal in an above-mentioned frequency expansion sequence cycle are multiplied each other;

(4) the above-mentioned later signal that multiplies each other is carried out fast Fourier transform;

(5) k value that obtains with above-mentioned fast Fourier transform and phase place exp (jk ²π/N) multiply each other obtains the slip correlation of received signal and initial frequency expansion sequence, and wherein N is the length of frequency expansion sequence;

(6) according to above-mentioned channel estimation value above-mentioned slip correlation is carried out multipath and merge, and the value after relatively merging, obtain launching data by the pairing frequency expansion sequence of maximum.

In the said method, obtain the method for channel estimation value, may further comprise the steps:

(1) from spread spectrum communication system, takes out the sampled signal in a frequency expansion sequence cycle in the training sequence signal of baseband receiving signals;

(2) initial frequency expansion sequence in the spread spectrum communication system is carried out conjugater transformation, the initial frequency expansion sequence after the conjugater transformation and the sampled signal in an above-mentioned frequency expansion sequence cycle are multiplied each other;

(3) the above-mentioned later signal that multiplies each other is carried out fast Fourier transform;

(4) a k+ Δ value and the phase place exp (jk that obtains with above-mentioned fast Fourier transform ²π/N) multiply each other obtains channel estimation value, wherein Δ=f * T _s, f is the frequency deviation of baseband receiving signals, T _sBe the frequency expansion sequence cycle, N is the length of frequency expansion sequence.

The present invention propose based on the diversity receiving method in the spread spectrum communication system of GCL sequences, have the following advantages:

Do not carry out when (1) channel estimating in the diversity reception can have full remuneration in the frequency deviation of received signal, therefore increased the flexibility of system design;

(2) channel estimating uses fast fourier transform algorithm to realize, has reduced the complexity that spread spectrum communication system is realized;

(3) Rake in the diversity reception receives and uses simpler structure and fast fourier transform algorithm, has lower amount of calculation.

Description of drawings

Fig. 1 is the impulse response schematic diagram of general multipath transmisstion channel.

Fig. 2 is the principle schematic that realizes channel estimating in the inventive method.

Fig. 3 realizes the principle schematic that Rake receives in the inventive method.

Embodiment

The present invention propose based on the diversity receiving method in the spread spectrum communication system of GCL sequences, at first from spread spectrum communication system, obtain channel estimation value in the training sequence signal of baseband receiving signals; From spread spectrum communication system, take out the sampled signal in a frequency expansion sequence cycle in the data-signal of baseband receiving signals; Initial frequency expansion sequence to spread spectrum communication system carries out conjugater transformation, and the initial frequency expansion sequence after the conjugater transformation and the sampled signal in an above-mentioned frequency expansion sequence cycle are multiplied each other; Later signal carries out fast Fourier transform to multiplying each other; K value that obtains with fast Fourier transform and phase place exp (jk ²π/N) multiply each other obtains the slip correlation of received signal and initial frequency expansion sequence, and wherein N is the length of frequency expansion sequence; According to above-mentioned channel estimation value above-mentioned slip correlation is carried out multipath and merge, and the value after relatively merging, obtain launching data by the pairing frequency expansion sequence of maximum.

In the said method, the method for obtaining channel estimation value is to take out the sampled signal in a frequency expansion sequence cycle from spread spectrum communication system in the training sequence signal of baseband receiving signals; Initial frequency expansion sequence in the spread spectrum communication system is carried out conjugater transformation, the initial frequency expansion sequence after the conjugater transformation and the sampled signal in an above-mentioned frequency expansion sequence cycle are multiplied each other; The above-mentioned later signal that multiplies each other is carried out fast Fourier transform; The k+ Δ value and the phase place exp (jk that obtain with above-mentioned fast Fourier transform ²π/N) multiply each other obtains channel estimation value, wherein Δ=f * T _s, f is the frequency deviation of baseband receiving signals, T _sBe the frequency expansion sequence cycle, N is the length of frequency expansion sequence.

In transmitting, training sequence is a GCL sequences, the initial frequency expansion sequence when it also is used as data-modulated.Other frequency expansion sequence of data-modulated is the cyclic shift of initial frequency expansion sequence, represents that wherein the decimal number of modulating data equals the figure place of cyclic shift.The GCL sequences that constitutes initial frequency expansion sequence is made up of N element, and N is an even number, wherein n element a _nExpression formula be:

$a_n=\exp \left(j\frac{\mathrm{M\π}{n}^2}{N}\right),$

Here n=0,1 ..., N-1, M are the integers relatively prime with N, equaling 1 with M in following each several part is that example describes.GCL sequences is the sequence of complex numbers of permanent envelope, and its phase place is nonlinear change with sampled point, and frequency is linear change with sampled point, so can regard it as the sample sequence of linear FM signal.

GCL sequences has desirable cycle autocorrelation performance, and promptly two identical GCL sequences only just have the correlation of non-zero in alignment, and correlation all is zero under other situation.This characteristic can be formulated as,

$R\left(k\right)=\underset{n=0}{\overset{N-1}{\∑}}{a}_{(n+k)\%N}{a}_n^{*}=\left\{\begin{array}{c}N,k=0\\ 0,k=1,\·\·\·,N-1\end{array}\right.$

Autocorrelation value when R in the formula (k) expression displacement is k, N is the length of frequency expansion sequence, and % represents to ask modular arithmetic, and promptly after n+k surpassed N, since 0 calculating, this was equivalent to GCL sequences is done circulative shift operation the subscript of a again.

At 1/T _sThe influence of frequency deviation of integral multiple under, GCL sequences produces the effect of cyclic shift, and the figure place of cyclic shift and 1/T _sMultiple equate that this characteristic can be formulated as,

$\exp \left(j\frac{{\mathrm{\&pi;n}}^2}{N}\right)\&CenterDot;\exp \left(\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="A20051007194100082.png"\; state="\{\{state\}\}">j</figure-callout>}\frac{2\mathrm{\&pi;k}}{N}n\right)=\exp (-\mathrm{<figure-callout\; id="2"\; label="j\; k"\; filenames="A20051007194100082.png"\; state="\{\{state\}\}">j</figure-callout>}\frac{k^{}}{}\frac{{}^2\mathrm{\&pi;}}{N})\&CenterDot;\exp \left(j\frac{\mathrm{\&pi;}(n+k{)}^2}{N}\right)$

Here T _sBe the frequency expansion sequence cycle, first expression in left side GCL sequences, second expression 1/T _sK frequency deviation doubly; But first expression in the right be a relevant fixed skew that have nothing to do with n with k, and second expression circulates left and move GCL sequences behind the k position.It will be appreciated that below, utilize these two character of GCL sequences can obtain simple channel estimation methods and Rake method of reseptance.

Shown in Figure 1 is the impulse response schematic diagram of general multipath transmisstion channel, and wherein abscissa is represented the time delay of multipath, and ordinate is represented the average power of multipath, and the average power of different multipaths is index decreased with time delay.In general multipath transmisstion channel, every multipath all is a multiple Gaussian random variable, and its amplitude is a rayleigh distributed, and phase place is evenly to distribute.The difference of power in different applied environments between multipath has a great difference, and for example, when having the direct projection path between the transmitter and receiver, article one that is to say that directly the footpath that arrives the earliest can be very strong, other footpath of causing by reflection can be relatively a little less than; And if do not have the direct projection path between transmitter and receiver, or exist obstacle to cause very big energy attenuation between the direct projection path, then reflection path may be a lot of by force.

Before channel estimating, suppose only remaining 1/T in the received signal _sThe frequency deviation of integral multiple, i.e. 1/T _sThe frequency deviation of little several times was compensated.The method of channel estimating is, takes out the sampled signal in a frequency expansion sequence cycle from the training sequence signal that receives, and allows conjugate sequence corresponding points of itself and initial frequency expansion sequence multiply each other, and multiplied result carried out the fast Fourier transform line phase of going forward side by side proofread and correct.

Describe the principle of channel estimation methods below in detail.When supposing not have frequency deviation, the training sequence signal in the received signal is r, because the desirable autocorrelation of GCL sequences, the slip correlation of training sequence signal r and initial frequency expansion sequence a is exactly a channel estimation value, promptly

$R\left(k\right)=\underset{n=0}{\overset{N-1}{\&Sum;}}r\left[n\right]a^{*}[(n+k)\%N].$

In received signal, there is Δ/T _sFrequency deviation the time, Δ is an integer, supposes that the training sequence signal is

Then channel estimation value is

$R\left(k\right)=\underset{n=0}{\overset{N-1}{\mathrm{\&Sigma;}}}r\left[n\right]a^{*}[(n+k)\%N]=\underset{n=0}{\overset{N-1}{\mathrm{\&Sigma;}}}\tilde{r}\left[n\right]\exp (-\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="A20051007194100082.png"\; state="\{\{state\}\}">j</figure-callout>}\frac{2\mathrm{\&Delta;\&pi;}}{N}n)a^{*}[(n+k)\%N].$

By the cyclic shift of the top GCL sequences that has provided and the relation of frequency deviation, promptly

$a[(n+k)\%N]=\exp \left(\mathrm{<figure-callout\; id="2"\; label="j\; k"\; filenames="A20051007194100082.png"\; state="\{\{state\}\}">j</figure-callout>}\frac{k^{}}{}\frac{{}^2\mathrm{\&pi;}}{N}\right)a\left[n\right]\exp \left(\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="A20051007194100082.png"\; state="\{\{state\}\}">j</figure-callout>}\frac{2\mathrm{k\&pi;}}{N}n\right),$

Can be written as by the channel estimation value when having frequency deviation in the received signal

$R\left(k\right)=\exp (-\mathrm{<figure-callout\; id="2"\; label="j\; k"\; filenames="A20051007194100082.png"\; state="\{\{state\}\}">j</figure-callout>}\frac{k^{}}{}\frac{{}^2\mathrm{\&pi;}}{N})\&CenterDot;\underset{n=0}{\overset{N-1}{\mathrm{\&Sigma;}}}\left(\tilde{r}\right[n\left]a^{*}\right[n\left]\right)\exp (-j\frac{2(k+\mathrm{\&Delta;})\mathrm{\&pi;}}{N}n)=\exp (-\mathrm{<figure-callout\; id="2"\; label="j\; k"\; filenames="A20051007194100082.png"\; state="\{\{state\}\}">j</figure-callout>}\frac{k^{}}{}\frac{{}^2\mathrm{\&pi;}}{N})F(k+\mathrm{\&Delta;}).$

F in the formula (k+ Δ) expression is to product

Do the k+ Δ output valve after the fast Fourier transform, and F (k+ Δ) be multiply by a phase place exp (jk ²π/N) has just obtained channel estimation value R (k).In fact, frequency deviation Δ/T before the channel estimating _sMay be unknown, so after obtaining the value of fast Fourier transform, also will wait frequency deviation to estimate to finish just to obtain final channel estimation value.The process of whole channel estimating as shown in Figure 2.

She Ji channel estimation methods can improve the flexibility of system design like this, because the value of F (k+ Δ) may be when time synchronized just calculate, so does not need to calculate separately during channel estimating again.Channel estimating only need be preserved the value of F (k+ Δ), and after frequency offset estimating was finished by the time, it was just passable according to the value of Δ F (k+ Δ) to be made a phasing.

Total N of the channel estimation value that obtains like this, that is to say the channel gain estimated value in N bar footpath, but in fact may there be so many footpaths in the channel at all, it is can be uncared-for owing to amplitude is too for a short time that some footpaths are perhaps arranged, select so after obtaining all channel estimation values, should do, select the bigger footpath of L bar and be used for doing the Rake merging.

Rake method of reseptance when data demodulates is described below.Because N possible frequency expansion sequence is arranged during data-modulated, and they are cyclic shifts of initial frequency expansion sequence, so can change order of operation that Rake receives to reduce the amount of calculation of demodulation.Specific practice is, get the received signal in a frequency expansion sequence cycle, allow its conjugation corresponding points multiply each other with initial frequency expansion sequence, multiplied result is done fast Fourier transform, value after the fast Fourier transform is done phasing, obtain N the value after the multipath merging according to channel estimation value with to the hypothesis of launching frequency expansion sequence, this N value is compared select a maximum value, be worth pairing frequency expansion sequence according to this and just can obtain launching data.

Describe this Rake recipient's ratio juris below in detail.The received signal of supposing a current frequency expansion sequence cycle of getting is r, and wherein contained frequency expansion sequence is a _m, i.e. the m time of initial frequency expansion sequence displacement then carried out despreading and carries out the multipath merging obtaining to received signal,

${\mathrm{\&rho;}}_m=\underset{l=0}{\overset{L-1}{\mathrm{\&Sigma;}}}\left\{c_l\underset{n=0}{\overset{N-1}{\mathrm{\&Sigma;}}}r\right[n\left]a_m\right[(n-l)\%N\left]\right\}$

C in the formula _lBe the l bar estimated value directly that the front channel estimation process obtains, L is the bar number of multipath.Which frequency expansion sequence in order to judge contained in the received signal is, need be to possible frequency expansion sequence a ₀～a _N-1Suppose one by one, and calculate ρ ₀～ρ _N-1Value.Because a ₀～a _N-1Be the 0th～N-1 time displacement of initial frequency expansion sequence, and a _m[(n-l) %N] is the l time displacement of m sequence, calculating ρ as can be seen ₀～ρ _N-1Process in need to use received signal r and all a ₀～a _N-1Correlation, and each correlation all will be repeated to calculate L time.So received signal r and frequency expansion sequence a in the method for the invention, ₀～a _N-1Correlation calculated in advance, and use fast fourier transform algorithm, therefore whole Rake receives the complexity that realizes and is greatly diminished.

Use the fast fourier transform algorithm realization received signal principle front relevant to carry out explanation with frequency expansion sequence, here it should be noted that, because the difference of sequence direction of displacement, the corrected value of phase place is also inequality after the fast Fourier transform, and this difference can be formulated as:

$\underset{n=0}{\overset{N-1}{\mathrm{\&Sigma;}}}r\left[n\right]a_m[(n-l)\%N]=\exp (-j\frac{(m+l{)}^2\mathrm{\&pi;}}{N})\&CenterDot;F[N-(m+l)]=\tilde{F}[N-(m+l)]$

The value that F in the formula (m) expression fast Fourier transform obtains, Value behind the expression phasing.Therefore, to different frequency expansion sequence hypothesis, it is right that Rake only needs when receiving Value merge, this process can be expressed as with formula as shown in Figure 3,

${\mathrm{\&rho;}}_m=\underset{l=0}{\overset{L-1}{\mathrm{\&Sigma;}}}\left\{c_l\tilde{F}\right[N-(m+l)\left]\right\}.$

Utilize said method to calculate after N the ρ value, therefrom choose a maximum value, its pairing frequency expansion sequence just is considered to the frequency expansion sequence launched, thereby can be launched data accordingly.

Claims (2)

1, a kind of based on the diversity receiving method in the spread spectrum communication system of GCL sequences, it is characterized in that this method may further comprise the steps:

(1) from spread spectrum communication system, obtains channel estimation value in the training sequence signal of baseband receiving signals;

(2) from spread spectrum communication system, take out the sampled signal in a frequency expansion sequence cycle in the data-signal of baseband receiving signals;

(3) the initial frequency expansion sequence to spread spectrum communication system carries out conjugater transformation, and the initial frequency expansion sequence after the conjugater transformation and the sampled signal in an above-mentioned frequency expansion sequence cycle are multiplied each other;

(4) the above-mentioned later signal that multiplies each other is carried out fast Fourier transform;

(5) k value that obtains with above-mentioned fast Fourier transform and phase place exp (jk ²π/N) multiply each other obtains the slip correlation of received signal and initial frequency expansion sequence, and wherein N is the length of frequency expansion sequence;

(6) according to above-mentioned channel estimation value above-mentioned slip correlation is carried out multipath and merge, and the value after relatively merging, obtain launching data by the pairing frequency expansion sequence of maximum.

2, diversity receiving method as claimed in claim 1 is characterized in that the wherein said method of obtaining channel estimation value, may further comprise the steps:

(1) from spread spectrum communication system, takes out the sampled signal in a frequency expansion sequence cycle in the training sequence signal of baseband receiving signals;

(2) initial frequency expansion sequence in the spread spectrum communication system is carried out conjugater transformation, the initial frequency expansion sequence after the conjugater transformation and the sampled signal in an above-mentioned frequency expansion sequence cycle are multiplied each other;

(3) the above-mentioned later signal that multiplies each other is carried out fast Fourier transform;

(4) a k+ Δ value and the phase place exp (jk that obtains with above-mentioned fast Fourier transform ²π/N) multiply each other obtains channel estimation value, wherein Δ=f * T _s, f is the frequency deviation of baseband receiving signals, T _sBe the frequency expansion sequence cycle, N is the length of frequency expansion sequence.

Images