CN103441968A - Improved Jakes channel estimation method based on chaos random phase - Google Patents

Abstract

The invention discloses an improved Jakes channel estimation method based on a chaos random phase in the technical field of wireless communication. An in-phase component signal and an orthogonal component signal in a communication system are generated according to a phase signal sequence of an improved Jakes model system, and an actual value signal estimated value is further generated; path attenuation coefficient values of multiple paths are estimated according to the least mean square error criterion, the obtained actual value signal estimated value and an actually measured received signal. The improved Jakes channel estimation method based on the chaos random phase is used for generating a generalized stable signal under multi-path conditions and conducting channel estimation, is particularly suitable for a wireless communication system and a visible light communication system, and is simple in generation method and easy to implement.

Description

Modified model Jakes channel estimation methods based on the chaos random phase

Technical field

What the present invention relates to is a kind of method of wireless communication technology field, be specifically related to a kind of modified model Jakes channel estimation methods based on the chaos random phase, for Channel Modeling analysis and the channel estimating when the communication system under multi-path environment is transmitted for signal, the requirement of broad sense smooth performance be can meet, Analysis of channel property in wireless communication system, visible light communication system and reception and the detection of modulation signal are specially adapted to.

Background technology

In the research of at present radio communication and visible light communication, Clarke model and simulator thereof are a kind of common reference model (Zhang Min, " Clarke model emulation system and realization. " Changsha Vocationl Technical College's journals of communicating by letter, vol.5, no.4, pp.17-21, Dec.2006).The signal statistics of Clarke model requires to meet following feature: the probability-distribution function of (1) envelope is a wide-sense stationary process; (2) PHASE DISTRIBUTION is evenly distributed; (3) auto-correlation function is only relevant with the time difference; (4) cross-correlation function is 0.

The Clarke reference model is an ideal model, and in the various simulation analysis and application of current radio communication and visible light communication, often by introducing the Jakes model, carry out the above-mentioned Clarke reference model of matching (yellow relaxing, Yan Yunbao, Li Yu, " research and the improvement of wireless channel Jakes model. " Chuxiong Normal University's journal, vol.24, no.12, pp.42-44, Dec.2009).In the Jakes model, the field intensity real number form after the various plane waves that receiver receives superpose mutually can be expressed as follows:

$R\left(t\right)=\sqrt{\frac{2}{N_0}}{\mathrm{<figure-callout\; id="0"\; label="E"\; filenames="HDA0000376273010000011.png,HDA0000376273010000012.png"\; state="\{\{state\}\}">E</figure-callout>}}_0\underset{n=1}{\overset{N_0}{\mathrm{\&Sigma;}}}{C}_n\&CenterDot;\cos ({\mathrm{\&omega;}}_{c}t+{\mathrm{\&omega;}}_{m}t\cos {\mathrm{\&alpha;}}_n+{\mathrm{\&phi;}}_n),$ Wherein: N ₀the total number of paths that means multipath, E ₀mean normalization field intensity amplitude, C _n(n=1,2 ..., N ₀) mean the attenuation coefficient of n paths, ω _cmean carrier frequency, ω _mmean maximum doppler frequency, α _n(n=1,2 ..., N ₀) mean the angle of arrival of n paths, φ _n(n=1,2 ..., N ₀) mean fixed phase drift additional after the n of path.And its plural form is:

wherein: $T\left(t\right)=\sqrt{\frac{2}{N_0}}{\mathrm{<figure-callout\; id="0"\; label="E"\; filenames="HDA0000376273010000011.png,HDA0000376273010000012.png"\; state="\{\{state\}\}">E</figure-callout>}}_0\underset{n=1}{\overset{N_0}{\mathrm{\&Sigma;}}}{C}_n\&CenterDot;e^{j({\mathrm{\&omega;}}_{m}t\cos {\mathrm{\&alpha;}}_n+{\mathrm{\&phi;}}_n)}.$

Analyzed and studied by the statistical property to the Jakes model known, although the phase place that receives signal (0,2 π] obey in scope and be uniformly distributed, can meet the phase place requirement of Clarke reference model; But calculating the auto-correlation function that receives signal can obtain:

$\begin{array}{c}{R}_{\mathrm{RR}}({\mathrm{<figure-callout\; id="1"\; label="t"\; filenames="HDA0000376273010000011.png,HDA0000376273010000012.png"\; state="\{\{state\}\}">t</figure-callout>}}_1,t_2)=E\left[R\left(t_1\right)R\left(t_2\right)\right]\\ ={\mathrm{<figure-callout\; id="0"\; label="E"\; filenames="HDA0000376273010000011.png,HDA0000376273010000012.png"\; state="\{\{state\}\}">E</figure-callout>}}_0^2\cos [{\mathrm{\&omega;}}_c(t_2-t_1)+{\mathrm{\&omega;}}_m(t_2-t_1)\cos {\mathrm{\&alpha;}}_n]+{\mathrm{<figure-callout\; id="0"\; label="E"\; filenames="HDA0000376273010000011.png,HDA0000376273010000012.png"\; state="\{\{state\}\}">E</figure-callout>}}_0^2\cos [{\mathrm{\&omega;}}_c({\mathrm{<figure-callout\; id="1"\; label="t"\; filenames="HDA0000376273010000011.png,HDA0000376273010000012.png"\; state="\{\{state\}\}">t</figure-callout>}}_1+t_2)+{\mathrm{\&omega;}}_m({\mathrm{<figure-callout\; id="1"\; label="t"\; filenames="HDA0000376273010000011.png,HDA0000376273010000012.png"\; state="\{\{state\}\}">t</figure-callout>}}_1+t_2)\cos {\mathrm{\&alpha;}}_n]\end{array}$

The auto-correlation function that above formula shows signal except with time difference (t ₂-t ₁) relevant outside, also with time and (t ₁+ t ₂) relevant.The signal that this proof Jakes model produces not is generalized stationary random process, and can not meet the requirement of Clarke reference model to auto-correlation function, to channel estimating, can have a negative impact.

By top analysis, can find that the signal that the Jakes simulator produces is not to experience each state, therefore the signal that its produces is not just the broad sense said before this paper signal stably yet, therefore sort signal disabled time mean value replaces the mean value of its statistics, also becomes the importance of its application examples of restriction as channel estimating.

Through the retrieval to prior art, find, international monopoly document number WO94/21065, open day 1994-09-15, disclose a kind of " method and apparatus for Code And Decode with the generation coherent communication ".This technology is when coding, and fiducial mark is inserted in the input data symbol stream to form the reference coded stream of an input data symbol.Then, thus be ready to the reference coded stream of transmission by the reference coded stream of using extended code expansion input data symbol before transmission.When decoding, the signal of communication that goes expansion to receive by extended code is to derive a benchmark sample stream and a data sample stream.By using the benchmark sample stream, channel response is estimated out.By using the estimated channel response to detect the estimation data symbol from data sample value current.But the defect that the prior art is compared with the present invention and deficiency are that the prior art algorithm complex is higher, larger to the overhead requirements of hardware device like this, be unfavorable for the requirement of device miniaturization and low energy consumption.

Summary of the invention

The present invention is directed to the prior art above shortcomings, a kind of modified model Jakes channel estimation methods based on the chaos random phase is proposed, for generation of the broad sense stationary signal in the multipath situation, and carry out channel estimating, be specially adapted to wireless communication system and visible light communication system, and its production method is simple, be easy to realize.

The present invention is achieved by the following technical solutions, the present invention includes following steps:

The first step at first gathers, detects or obtain for setting up the system parameters of modified model Jakes model from network, selected independent variable have (0,2 π] be uniformly distributed the discrete iteration chaos system x of characteristic in scope _k+1=f (x _k), $x_{k+1}=\left\{\begin{array}{cc}2\mathrm{\&pi;}\frac{x_k}{a},& (0<x_k<a)\\ 2\mathrm{\&pi;}\frac{2\mathrm{\&pi;}-{\mathrm{<figure-callout\; id="2"\; label="x\; k"\; filenames="HDA0000376273010000011.png"\; state="\{\{state\}\}">x</figure-callout>}}_k}{2\mathrm{\&pi;}-a},& (a\&le;x_k\&le;2\mathrm{\&pi;})\end{array}\right.k=\mathrm{0,1,2},\&CenterDot;\&CenterDot;\&CenterDot;,$ Wherein: a is the parameter of discrete iteration chaos system, 0<a<2 π; And produce corresponding chaotic signal sequence { x ₀, x ₁, x ₂...; And by the chaotic signal sequence { x produced ₀, x ₁, x ₂... the fixing sample time length in interval, the phase signal sequence { Ψ of generation modified model Jakes model system ₁, Ψ ₂..., Ψ _n=x _bn, n=1,2 ..., N ₀, wherein: b means the sampling interval number of times and is integer.

Described system parameters comprises: the multipath sum N of the multi-path system of setting up ₀, the reception signal strength range value E after normalization ₀, each multipath path angle of arrival α _n(n=1,2 ..., N ₀), carrier frequency value ω _c, maximum doppler frequency value ω _m, resulting actual measurement receives signal R (t).

Second step, generate the in-phase component signal T in communication system according to the phase signal sequence of modified model Jakes model system _cand orthogonal component signal T (t) _s(t) signal, wherein:

ω _n=ω _mcos α _n, C _nfor the path attenuation coefficient value of each multipath, n=1,2 ..., N ₀; And according to T _cand T (t) _s(t) signal produces the real-valued signal estimated value

and according to resulting real-valued signal estimated value and actual measurement reception signal R (t), utilize minimum mean square error criterion to be estimated the path attenuation coefficient value of each multipath, ${\left\{{\hat{C}}_n\right\}}_{n=\mathrm{1,2},\&CenterDot;\&CenterDot;\&CenterDot;,{\mathrm{<figure-callout\; id="0"\; label="N"\; filenames="HDA0000376273010000011.png,HDA0000376273010000012.png"\; state="\{\{state\}\}">N</figure-callout>}}_0}=\underset{{\left\{C_n\right\}}_{n=\mathrm{1,2},\&CenterDot;\&CenterDot;\&CenterDot;,N_0}}{\arg \min }{[R\left(t\right)-\hat{R}\left(t\right)]}^2.$

The 3rd step, utilize the path attenuation coefficient value of resulting each multipath be applied to, in the multipath receiver of radio communication and visible light communication, realize the optimization of Receiver Complexity, specifically refer to: in the multipath receiver of radio communication and visible light communication, utilize the path attenuation coefficient value of resulting each multipath

and with fixing path attenuation coefficient threshold value C _thresholdcompare, by statistics, be greater than path attenuation coefficient threshold value C _thresholddecay number of path N ₀', using its in multipath receiver for storage with calculate the maximum of the multipath component of received signal, or the number of propping up the memory of circuit-switched data for storing multipath.

Technique effect

Compared with prior art, the present invention is simple for generation of the system model of output signal, and implementation Process is convenient, has good feasibility and practicality.

The accompanying drawing explanation

Fig. 1 is that the envelope probability-distribution function of signal that embodiment produces is counted the curve chart in situation at different multipaths.

Fig. 2 is the auto-correlation function of signal that embodiment produces and the correlation curve figure of Clarke reference model.

Embodiment

Below embodiments of the invention are elaborated, the present embodiment is implemented take technical solution of the present invention under prerequisite, provided detailed execution mode and concrete operating process, but protection scope of the present invention is not limited to following embodiment.

Embodiment 1

The present embodiment is to generate the multipath signal sequence of a modified model Jakes model based on the chaos random phase with the Matlab simulation software, verifies the degree of agreement of its statistical property and Clarke statistical property that reference model requires, and carries out corresponding channel estimating.

The specific implementation process of this method is as follows:

1) at first select for setting up the system parameters of modified model Jakes model, i.e. the multipath of system sum N ₀, receive signal strength range value E ₀, each multipath signal path attenuation coefficient value C _n(n=1,2 ..., N ₀), the path angle of arrival α of each multipath signal _n(n=1,2 ..., N ₀), carrier frequency value ω _cwith maximum doppler frequency value ω _m.

In the present embodiment, N ₀value be several different situations such as 3,5,8,25,50, wherein suppose N ₀the=50th, a kind of multipath sum levels off to infinite situation; E ₀=1; C _n(n=1,2 ..., N ₀) adopt the rand () function carried in the Matlab simulation software produce (0,1] interval equally distributed stochastic variable; α _n(n=1,2 ..., N ₀) adopt the rand () function carried in the Matlab simulation software produce (0,2 π] interval equally distributed stochastic variable; ω _c=10 ⁶hertz; ω _m=100 hertz.

2) next selected independent variable have (0,2 π] the discrete iteration chaos system x that there is the characteristic of being uniformly distributed in scope _k+1=f (x _k), (k=0,1,2 ...), and produce required chaotic signal sequence { x ₀, x ₁, x ₂....

Described discrete iteration chaos system, specifically: $x_{k+1}=\left\{\begin{array}{cc}2\mathrm{\&pi;}\frac{x_k}{a},& (0<x_k<a)\\ 2\mathrm{\&pi;}\frac{2\mathrm{\&pi;}-{\mathrm{<figure-callout\; id="2"\; label="x\; k"\; filenames="HDA0000376273010000011.png"\; state="\{\{state\}\}">x</figure-callout>}}_k}{2\mathrm{\&pi;}-a},& (a\&le;x_k\&le;2\mathrm{\&pi;})\end{array}\right.(k=\mathrm{0,1,2},\&CenterDot;\&CenterDot;\&CenterDot;),$ Wherein: the value of parameter a is 4.8729, and the initial value of discrete iteration chaos system is x ₀=0.2802.

According to produced chaotic signal sequence { x ₀, x ₁, x ₂... the fixing sample time length in one, interval, the discrete chaos random phasic signal sequence { Ψ of generation modified model Jakes simulator simulation model system ₁, Ψ ₂....

Described sampling process, specifically: Ψ _n=x _bn, (n=1,2 ..., N ₀), wherein: the value of parameter b is 20.

3) generate T according to following formula _cand T (t) _s(t) signal: $T_s\left(t\right)=\frac{\sqrt{2}}{\sqrt{{\mathrm{<figure-callout\; id="0"\; label="N"\; filenames="HDA0000376273010000011.png,HDA0000376273010000012.png"\; state="\{\{state\}\}">N</figure-callout>}}_0}}\underset{n=1}{\overset{N_0}{\mathrm{\&Sigma;}}}{C}_n\&CenterDot;\sin ({\mathrm{\&omega;}}_{n}t+{\mathrm{\&Psi;}}_n),$ ω _n=ω _mcos α _n(n=1,2 ..., N ₀), here

it is one group of real-valued sequence to be solved.

4) produce the real-valued signal estimated value according to following formula $\hat{R}\left(t\right):\hat{R}\left(t\right)=E_0{T}_c\left(t\right)\cos {\mathrm{\&omega;}}_{c}t-E_0{T}_s\left(t\right)\sin {\mathrm{\&omega;}}_{c}t.$

5) finally according to resulting real-valued signal estimated value

and actual measurement reception signal R (t), and utilize minimum mean square error criterion to be estimated the path attenuation coefficient value of each multipath, specifically:

can obtain resulting estimated value

be equation group $\frac{\&PartialD;{[R\left(t\right)-\hat{R}\left(t\right)]}^2}{\&PartialD;C_n}=0(n=\mathrm{1,2},\&CenterDot;\&CenterDot;\&CenterDot;,N_0)$ Solution.

As shown in Figure 1, for the present embodiment at N ₀the envelope probability-distribution function of the lower signal that produces of=3,5,25 situation is counted the curve chart in situation at different multipaths, in figure, has also drawn the Rayleigh fading signal as a reference.

As shown in Figure 2, for the present embodiment at N ₀the auto-correlation function of the lower signal that produces of the situation of=8,50 (inf) and the correlation curve figure of Clarke reference model.From Fig. 1 and Fig. 2, the signal envelope probability-distribution function that the present embodiment produces can be regarded a wide-sense stationary process substantially as, and its correlation function and time and (t ₁+ t ₂) substantially irrelevant.Simultaneously, selected discrete iteration chaos system have (0,2 π] equally distributed characteristic in scope.Therefore, the signal that the method produces can meet the requirement of Clarke modeling statistics characteristic.

Simultaneously, the variance of definition channel estimation errors is

calculate and average by 100 times, relatively the present embodiment and prior art are (yellow easypro, Yan Yunbao, Li Yu, " research and the improvement of wireless channel Jakes model. " Chuxiong Normal University's journal, vol.24, no.12, pp.42-44, Dec.2009) channel estimation error variance value, can obtain: prior art (yellow relaxing, Yan Yunbao, Li Yu, " research and the improvement of wireless channel Jakes model. " Chuxiong Normal University's journal, vol.24, no.12, pp.42-44, the mean value of channel estimation error variance Dec.2009) is 0.1352, and the mean value of the resulting channel estimation error variance of the present embodiment is 0.0947.Therefore, the channel estimation errors of the method is less than art methods, has certain superiority.

Claims (3)

1. the modified model Jakes channel estimation methods based on the chaos random phase, is characterized in that, comprises the following steps:

The first step at first gathers, detects or obtain for setting up the system parameters of modified model Jakes model from network, selected independent variable have (0, π, 2 models] be uniformly distributed the discrete iteration chaos system x of characteristic in enclosing _k+1=f (x _k), $x_{k+1}=\left\{\begin{array}{cc}2\mathrm{\&pi;}\frac{x_k}{a},& (0<x_k<a)\\ 2\mathrm{\&pi;}\frac{2\mathrm{\&pi;}-x_k}{2\mathrm{\&pi;}-a},& (a\&le;x_k\&le;2\mathrm{\&pi;})\end{array}\right.k=\mathrm{0,1,2},\&CenterDot;\&CenterDot;\&CenterDot;,$ Wherein: a is the parameter of discrete iteration chaos system, 0<a<2 π; And produce corresponding chaotic signal sequence { x ₀, x ₁, x ₂...; And by the chaotic signal sequence { x produced ₀, x ₁, x ₂... the fixing sample time length in interval, the phase signal sequence { Ψ of generation modified model Jakes model system ₁, Ψ ₂..., Ψ _n=x _bn, n=1,2 ..., N ₀, wherein: b means the sampling interval number of times and is integer;

Second step, generate the in-phase component signal T in communication system according to the phase signal sequence of modified model Jakes model system _cand orthogonal component signal T (t) _s(t) signal, wherein:

ω _n=ω _mcos α _n, C _nfor the path attenuation coefficient value of each multipath, n=1,2 ..., N ₀; And according to T _cand T (t) _s(t) signal produces the real-valued signal estimated value

and according to resulting real-valued signal estimated value

and actual measurement reception signal R (t), utilize minimum mean square error criterion to be estimated the path attenuation coefficient value of each multipath, ${\left\{{\hat{C}}_n\right\}}_{n=\mathrm{1,2},\&CenterDot;\&CenterDot;\&CenterDot;,N_0}=\underset{{\left\{C_n\right\}}_{n=\mathrm{1,2},\&CenterDot;\&CenterDot;\&CenterDot;\&CenterDot;{,N}_0}}{\arg \min }{[R\left(t\right)-\hat{R}\left(t\right)]}^2;$

The 3rd step, utilize the path attenuation coefficient value of resulting each multipath

be applied to, in the multipath receiver of radio communication and visible light communication, realize the optimization of Receiver Complexity.

2. method according to claim 1, is characterized in that, described system parameters comprises: the multipath sum N of the multi-path system of setting up ₀, the reception signal strength range value E after normalization ₀, each multipath path angle of arrival α _n(n=1,2 ..., N ₀), carrier frequency value ω _c, maximum doppler frequency value ω _m, resulting actual measurement receives signal R (t).

3. method according to claim 1, is characterized in that, described the 3rd step specifically refers to: in the multipath receiver of radio communication and visible light communication, utilize the path attenuation coefficient value of resulting each multipath

and with fixing path attenuation coefficient threshold value C _thresholdcompare, by statistics, be greater than path attenuation coefficient threshold value C _thresholddecay number of path N ₀', using its in multipath receiver for storage with calculate the maximum of the multipath component of received signal, or the number of propping up the memory of circuit-switched data for storing multipath.

Images