CN104660390A - Communication method of multiple carrier-code division multiple access system of CDMA with combination of ACO-OFDM - Google Patents

Abstract

The invention discloses a communication method of a multiple carrier-code division multiple access system of CDMA with combination of ACO-OFDM. The communication method comprises the following main steps: (1) correspondingly adding each user symbol after frequency amplification, and generating a frequency amplification signal including all user information; (2) mapping the frequency amplification signal into a signal meeting Hermitian symmetry, wherein only odd number carriers include information, and even number carriers are completely 0, so that IFFT output can be real signals; (3) performing scale change on the mapped signals, wherein the change factor determines the variance of the IFFT output signals, so that the purpose of limiting the power of emitted light is achieved; (4) limiting the amplitude of the real signals output from IFFT, thereby achieving the purpose that a practical light source works in a linear area; (5) only extracting a front half of signals of odd number carriers output from IFFT through a receiving end; (6) passing the extracted signals through an equalizer so as to compensate amplitude limiting attenuation and scale change. By adopting the communication method, requirements of a light wireless communication intensity modulation direct detection system in a multi-user environment can be met, and the light emission power can be limited according to system requirements.

Description

A kind of CDMA is in conjunction with the light MC-CDMA system communication means of ACO-OFDM

Technical field

The present invention relates to the light wireless communication system of code division multiple access and the combination of light multi-carrier modulation, especially relate to the light MC-CDMA system of code division multiple access and the combination of asymmetric amplitude limit light OFDM.

Background technology

Along with user is to the continuous increase of real-time broadband wireless data demand, it is rare that traditional radio frequency (Radio Frequency, RF) access technology also exists frequency spectrum resource, poor stability, the deficiencies such as serious electromagnetic interference and huge energy consumption.Light wireless communication (Optical wireless communication, OWC) can provide huge frequency spectrum resource, without the need to frequency spectrum license, realizes the advantages such as simple, be considered to the very promising complementary type access technology of RF without electromagnetic interference and system.

Code division multiple access (Code Division Multiple Access, CDMA) as 3G (Third Generation) Moblie (3rd-Generation, core technology 3G), have that strong, the anti-narrow-band noise performance of multiple access capabilities is good, the advantage such as the strong and good confidentiality of anti-multipath fading ability, but the spectral efficient that wireless technology exists requires and severe frequency Selective intensity two aspects, how CDMA technology tackles the key that these two challenges become multiple access technique of future generation design.OFDM (OrthogonalFrequency Division Multiplexing, OFDM) is a class multi-carrier modulation technology, in the intersymbol interference that anti-multipath effect causes, have its Inherent advantage, is very applicable to High Bit-rate Wireless Data Communication.MC-CDMA (Multi Carrier-CodeDivision Multiple Access, MC-CDMA) be the technology combined by CDMA and OFDM, there is the advantage of the aspects such as the intersymbol interference that anti-narrow-band noise is disturbed, multipath effect causes, and there is the feature of frequency diversity.

As far back as document [1] " Multi-carrier CDMA in indoor wireless radio networks " in 1993. (N.Yee, J.P.Linnarz, G.Fettweis.Proceedings IEEE International Symposium on Personal, Indoor andMobile Radio Commun. [C], 1993, pp.109-113) MC-CDMA technical scheme is just proposed, its general principle shown in Figure 1 is: at transmitting terminal, be provided with U user and send information simultaneously, the information symbol of wherein u user's input is a _uk (), k represents a kth mark space, a _u(k) with equiprobability value in ± 1, T _sfor the information symbol cycle.A _uk () is first input to a reproducer, it exports as the parallel a in N road _uk symbol that () copies; Then the length of N road parallel symbol and u user is the orthogonal spreading sequence C of N _ube multiplied according to corresponding position (Chip), the N long symbol sequence of parallel output carries out carrier modulation again, and the frequency interval of modulated sub-carriers is F/T _s, wherein F is integer, is therefore orthogonal between each subcarrier and contains frequency expansion sequence symbol; Finally parallel N road modulation signal is added, as the output signal s of u user _u(t).Usually be independently between every subcarriers signal, so itself and signal s _ut () is a complex signal.After transmission, signal r (t) received at receiving terminal is the mixed signal of U subscriber signal.Shown in Figure 2, signal r (t) exports N subcarriers signal after demodulator, each road signal is multiplied by one again and merges N road signal after the factor respectively through low pass filter (Low Pass Filter, LPF), the baseband signal exported is added, and then produces the judgment variables b of a kth symbol after sampling _u.In the system of reality, modulation /demodulation can use inverse fast fourier transform/fast Fourier transform (IFFT/FFT) algorithm to realize.

After MC-CDMA proposes, scholars have carried out large quantity research to its application in RF system.But in light wireless communication system, the document of research application MC-CDMA technology is less, mainly contains following correlative study:

(1) Luan Yingzi of domestic Xian Electronics Science and Technology University is at list of references [2] " MC-CDMA for optical wirelesscommunications " (Y.Luan, G.Jiang, and J.Li.Proc.SPIE 6021, Optical Transmission, Switching, and Subsystems III [C], 2005, pp.60210Z-60210Z-8) propose in light wireless communication system, apply MC-CDMA technology, devise and adopt decorrelation-Parallel Interference Cancellation (DEC-PIC) technology to carry out the system model of Multiuser Detection when up link, give the formula calculating Signal to Interference plus Noise Ratio.

(2) Thailand scholar Muhammad Zubair Farooqui, Poompat Saengudomlert is at list of references [3] " Transmitpower reduction through subcarrier selection for MC-CDMA-based indoor optical wirelesscommunications with IM/DD " (Eurasip Journal on Wireless Communications And Networking [J], 2013:138), research in the downlink, intensity modulated/direct-detection (Intensity Modulation Direct Detection, IM/DD) method of average transmitting power is reduced in system by sub-carrier selection.

(3) British scholar F.E.Alsaadi, J.M.Elmirghani is at list of references [4] " MC-CDMA Indoor OpticalWireless System " (2007IEEE Global Telecommunications Conference [C], 2007, pp.2455-2460), list of references [5] " Mobile MC-CDMA optical wireless system employing an adaptive multibeamtransmitter and diversity receivers in a real indoor environment " (IEEE 2008InternationalConference on Communications [C], 2008, pp.5196-5203), list of references [6] " Adaptive Mobile LineStrip Multibeam MC-CDMA Optical Wireless System Employing Imaging Detection in a RealIndoor Environment " (IEEE Journal on Selected Areas In Communications [J], 2009, vol.27, no.9, pp.1663-1675) in, research is under different indoor environments, different detection techniques is adopted to analyze the systematic function of light MC-CDMA.

In light wireless communication system, usual employing IM/DD modulation system, namely be exactly incoherent light-emitting diode (the Light-Emitting Diode of Signal shock transmitting terminal, LED) intensity (but not amplitude), so just needs the signal of driving LED to be non-negative real number signal.But in list of references [2,3], when research is applied to the MC-CDMA system of light wireless communication, CDMA and traditional OFDM combines, so MC-CDMA system output signal is a complex signal, if when being directly used in driving LED, the phase information of signal will be lost.In list of references [4,5,6], MC-CDMA output signal is complex signal equally, proposes to transmit real part and the imaginary part of complex signal respectively at transmitting terminal two LED in literary composition.But in IM/DD system during receiving terminal direct-detection, the two ways of optical signals separated in synchronization of load real part and imaginary part information will to be difficult to realize in photoelectric detector (Photodiode, PD).In addition, two paths of signals remerges after transmission, the twice of the additive noise interference Jiang Shi mono-road signal transmission suffered by the complex signal of recovery.

On the other hand, at list of references [2, 3, 4, 5, 6] all suppose in that the illuminating source of transmitting terminal is desirable device, and do not consider that actual device input/output signal when normal work always has certain dynamic range, namely be exactly input signal in certain dynamic range time, can think that the light signal strength that light source sends of making a start is proportional to input signal, if input signal is beyond this scope, make a start light source input signal need first after amplitude limiting processing, just can be input to LED light source, otherwise the relation of the light signal that light source exports and input signal will present nonlinear relation, when input signal is excessive even, LED light source is burned, cannot work again time too small.

Finally, in the MC-CDMA system of list of references [2,3,4,5,6] receiving terminal light signal can along with send simultaneously information number of users increase and increase.But in real system, must consider the health problem of human eye and skin, moreover consider the needs of room lighting in visible light communication, the light source luminescent power of light wireless communication normally will be restricted.

Summary of the invention

The object of the invention is design one is applicable under downlink multiuser environment, light wireless multi-carrier code division multiple access (O-MC-CDMA) communication system of multipath effect impact can be overcome, this system can meet intensity modulated direct-detection (IM/DD) system to the needs of nonnegative real number signal, the drive singal making LED light source be operated in linear zone can be provided, and user can be met to the restriction sending average light power.

The present invention is for solving its technical problem, and the technical scheme adopted is:

CDMA, in conjunction with a light MC-CDMA system communication means of ACO-OFDM, comprises following steps:

Transmitting terminal:

Step 1: the modulation symbol spread spectrum of each user, is added the spread-spectrum signal of all users, generates the spread-spectrum signal that comprises all user profile;

Step 2: map operation is carried out to spread-spectrum signal, mapping output signals meets hermitian symmetry characteristic, is a real signal to ensure that IFFT converts output signal;

Step 3: to the signal dimensional variation after mapping, changed factor determines the variance of IFFT output signal, reaches the object of restriction LED luminous power further;

Step 4: carry out IFFT conversion, and to output signal amplitude limit, using the DC bias signal sum of sliced output signal and LED as the drive singal of LED, it can be made to be operated in linear zone;

Step 5: Cyclic Prefix is added to the signal after amplitude limit, adds DC bias signal after parallel-serial conversion and digital-to-analogue conversion, driving LED light source luminescent;

Receiving terminal:

Step 1: receiver sensed light signal, to signal of telecommunication discretization after opto-electronic conversion, serioparallel exchange, is input to FFT after deleting Cyclic Prefix;

Step 2: only extract the odd subcarriers signal of the first half in FFT output signal at receiving terminal, then to the signal equalization process of extracting, to compensate the signal attenuation and the dimensional variation factor that amplitude limit brings;

Step 3: the signal diversifying exported by equalizer receives, and demodulation exports the information bit of user.

By method as above, the present invention makes code division multiple access (CDMA) and asymmetric amplitude limit light OFDM (Asymmetrically Clipped Optical Orthogonal Frequency Division Multiplexing, ACO-OFDM) combine and makes system output nonnegative real number signal; Carry out amplitude limit to the signal that IFFT exports, limitation signal meets makes LED light source be operated in linear zone; Introduce dimensional variation factor pair IFFT input sub-carrier signal and carry out dimensional variation, IFFT can be limited and output signal variance, thus threshold system can be used to send average light power.

In real process, the concrete steps of light MC-CDMA system design are as follows:

Transmitting terminal:

Step 1: its orthogonal spreading sequence spread spectrum of modulation symbol of each user, X _ui () represents i-th modulation symbol of u user, the average power normalization of usual modulation symbol.The spread-spectrum signal of all users is added, generates the spread-spectrum signal X that comprises all user profile _ss(m), m=0,1,2 ... G-1, G represent the length of frequency expansion sequence;

Step 2: spread-spectrum signal is carried out mapping (Mapping), mapping output signals meets hermitian symmetry (Hermitian) characteristic, and wherein only odd subcarriers comprises information, and even subcarriers is all 0, and mapping output signals is:

$X_{\mathrm{mapping}}={\left[\begin{array}{cccccccccccc}0& X_{\mathrm{ss}}\left(0\right)& 0& X_{\mathrm{ss}}\left(1\right)& 0& ...& X_{\mathrm{ss}}(G-1)& 0& X_{\mathrm{ss}}^{*}(G-1)& 0& ...& X_{\mathrm{ss}}^{*}\left(0\right)\end{array}\right]}^T$

Wherein () ^*represent conjugate operation, () ^tthe transposition of representing matrix.

Step 3: to the signal dimensional variation after mapping, namely gives X _mappingbe multiplied by a dimensional variation factor-alpha and obtain X _scaled.This factor determines the variance (mean square deviation) of IFFT output signal, thus reaches the object of restriction LED utilizing emitted light power.Convert the gross energy of input and output discrete signals and the constant character of average power according to IFFT, can obtain the dimensional variation factor is:

$\mathrm{\α}={\mathrm{\σ}}_x\sqrt{\frac{N}{\underset{n=0}{\overset{N-1}{\mathrm{\Σ}}}{\left|X_{\mathrm{mapping}}\left(n\right)\right|}^2}}$

Wherein σ _xrepresent the mean square deviation of IFFT output signal, || represent the modular arithmetic of signal, N represents the dimension N=4G that IFFT converts.

Step 4: the IFFT computing of employing is:

$x\left(k\right)=\frac{1}{\sqrt{N}}\underset{n=0}{\overset{N-1}{\mathrm{\Σ}}}{X}_{\mathrm{scaled}}\left(n\right)\exp \left(\frac{j2\mathrm{\πnk}}{N}\right)$

Wherein k=0,1 ..., N-1.IFFT conversion output signal x (k) is ambipolar sequence of real numbers, and according to the characteristic of ACO-OFDM, sequence of real numbers has skew-symmetry (Antisymmetry), therefore can delete negative value signal and not drop-out.But when actual realization, sliced output signal x _clippedthe DC bias signal B of (k) and LED _dCsum is as the drive singal of LED, if the dynamic range of input signal corresponding when making LED be operated in linear zone is (i _min, i _max), the difference of DC bias signal size, will cause limitation signal x _clippedk the dynamic range of () is also different, so the amplitude limit thresholding ε of IFFT output time-domain signal x (k) _bollomand ε _topfor:

ε _bottom＝max{i _min-B _DC，0}

ε _top＝i _max-B _DC

Step 5: add Cyclic Prefix (Cyclic Prefix, CP) to the signal after amplitude limit, after parallel-serial conversion (P/S) and digital-to-analogue conversion (D/A), adds DC bias signal, driving LED light source luminescent.

Receiving terminal:

Step 1: suppose that receiving terminal is user u, receiver sensed light signal, to signal of telecommunication discretization (A/D) after opto-electronic conversion, serioparallel exchange (S/P), is input to FFT after deleting Cyclic Prefix;

Step 2: the hermitian symmetry corresponding to the IFFT input signal of making a start, the odd subcarriers signal of the first half in FFT output signal is only extracted at receiving terminal, then to the signal equalization process of extracting, to compensate the signal attenuation and the dimensional variation factor that amplitude limit brings;

Step 3: by equalizer output signal diversity reception, gain (gain) factor is multiplied by each road, then sues for peace to all signals, exports the judgment variables of a user u, and last demodulation exports the information bit of user u.

Compared with prior art, the invention has the beneficial effects as follows:

(1) be different from traditional MC-CDMA system, the present invention is combined by CDMA and ACO-OFDM, achieves O-MC-CDMA system.

Traditional MC-CDMA is that CDMA and OFDM combines, the signal normally complex signal that in system, IFFT conversion exports.And the present invention has carried out hermitian symmetry to IFFT input signal, and only odd subcarriers comprises information, and even subcarriers is all 0.After IFFT conversion, the signal of output is a real number signal, then adds DC bias signal after amplitude limit, and being input to the signal that LED carries out intensity modulated will be a non-negative real signal, meet the needs of intensity modulated direct-detection system.

(2) object of restriction LED luminous power is reached by introducing the dimensional variation factor.On the other hand, make the average signal-to-noise ratio of receiving end signal constant, be conducive to theory analysis.

In O-MC-CDMA system, carry out dimensional variation, be namely multiplied by an invariant to the signal of IFFT input, the average symbol power of the mean square deviation that this Summing Factor IFFT outputs signal, the dimension of IFFT and user is relevant.Therefore dimensional variation can limit the variance yields of IFFT output signal, and the variance of signal does not change with the change of number of users and modulation system of making a start, thus reaches the object of restriction LED luminous power.

(3) consider input signal dynamic range when actual LED works in linear zone, real number signal amplitude limit is exported to IFFT, can ensure that the drive singal of LED is in this dynamic range.

According to central-limit theorem, the more and Fourier of number of users change dimension greatly time, it is the normal distribution of zero that IFFT output signal obeys average.So by clipping operation, avoid the situation being in nonlinear operation district because LED input signal is too small or excessive, even the LED possibility that do not work or be burned.In addition, the physical realizability of system is stronger.

The present invention obtains the subsidy of state natural sciences fund (61461026,61271245) and Gansu Province's Natural Science Fund In The Light (1212RJZA019).

Accompanying drawing explanation

Fig. 1 is conventional multi-carrier code division multiple access system transmitting terminal schematic diagram;

Fig. 2 is conventional multi-carrier code division multiple access system receiving terminal schematic diagram;

Fig. 3 is light MC-CDMA system transmitting terminal schematic diagram of the present invention;

Fig. 4 is light MC-CDMA system receiving terminal schematic diagram of the present invention;

Fig. 5 is that IFFT outputs signal clipping operation schematic diagram;

Fig. 6 is sub-carrier number when being 512, scattering multipath channel frequency domain response factor map of magnitudes;

Fig. 7 is light MC-CDMA system bit error rate analysis chart under line of signt transmission channel;

Fig. 8 is light MC-CDMA system bit error rate analysis chart under scattering multipath channel;

In Fig. 7 and Fig. 8, abscissa represents signal to noise ratio E _b/ N ₀, unit is decibel (dB), and ordinate represents bit error rate (Bit ErrorRate, BER).The solid line of tape symbol " " represents error rate theoretical value during desirable amplitude limit, BER Simulation value during the desirable amplitude limit of the represented by dotted arrows of tape symbol " ■ ".The solid line of tape symbol " ◇ " represents error rate theoretical value during monolateral amplitude limit, BER Simulation value during the monolateral amplitude limit of the represented by dotted arrows of tape symbol " ◆ ", the solid line of tape symbol " △ " represents error rate theoretical value during bilateral amplitude limit, BER Simulation value during the bilateral amplitude limit of the represented by dotted arrows of tape symbol " ▲ ".

Embodiment

See accompanying drawing 3, the embodiment of light MC-CDMA system transmitting terminal is:

(1) suppose that U user communicates simultaneously, the binary bit sequence of each user's input is b _u, u=0,1,2 ... U-1.Every segment length is divided into be log the binary information sequence that each user inputs ₂the binary data section of M-bit, the order of modulation adopting Gray code to be mapped to distributed rectangular is on quadrature amplitude modulation (Quadrature Amplitude Modulation, the QAM) planisphere of M; Modulator exports the modulation symbol of U parallel user, uses X _ui () represents i-th modulation symbol of u user, the average power of usual modulation symbol is normalized to 1.

(2) by the modulation symbol X of each user _ui () copies respectively, the length copying the length frequency expansion sequence used with spread spectrum is identical, if spreading sequence length is G; Then the parallel signal each user being copied output is multiplied with each of the frequency expansion sequence of this user accordingly, realizes the spread spectrum of user's modulation symbol; Finally the spread-spectrum signal of U user is added according to corresponding position, exports the spread-spectrum signal X comprising U user profile that length is G _ss(m), m=0,1,2 ... G-1.Spreading procedure can be equivalent to the product of a Spread Spectrum Matrix C and user symbol vector X (i), is expressed as follows:

$X_{\mathrm{ss}}=\mathrm{CX}\left(i\right)=\left[\begin{array}{ccc}{c}_0\left(0\right)c_1\left(0\right)& ...& c_{U-1}\left(0\right)\\ c_0\left(1\right)c_1\left(1\right)& ...& c_{U-1}\left(1\right)\\ .& .& .\\ .& .& .\\ .& .& .\\ c_0(G-1)c_1(G-1)& ...& c_{U-1}(G-1)\end{array}\right]\left[\begin{array}{c}{X}_0\left(i\right)\\ X_1\left(i\right)\\ .\\ .\\ .\\ X_{U-1}\left(i\right)\end{array}\right]$

Wherein c _um {-1,1} represents that the length of user u is the m position (Chip) of G frequency expansion sequence, spread-spectrum signal X to () ∈ _ssm () can be expressed as:

$\begin{array}{cc}{X}_{\mathrm{ss}}\left(m\right)=\underset{u=0}{\overset{U-1}{\mathrm{\Σ}}}{c}_u\left(m\right)X_u\left(i\right)& 0\≤m\≤G-1\end{array}$

(3) in order to meet the requirement of intensity modulated direct-detection system, it is real number signal that quadrature modulation is outputed signal, the input signal of IFFT should meet hermitian (Hermitian) symmetry, is namely exactly the conjugate lens picture that the latter half of input signal equals first half.ACO-OFDM only uses odd subcarriers load information, and even subcarriers is all 0.So need spread-spectrum signal X _ssm () carries out mapping (Mapping) operation.Mapping output signals is:

$X_{\mathrm{mapping}}={\left[\begin{array}{cccccccccccc}0& X_{\mathrm{ss}}\left(0\right)& 0& X_{\mathrm{ss}}\left(1\right)& 0& ...& X_{\mathrm{ss}}(G-1)& 0& X_{\mathrm{ss}}^{*}(G-1)& 0& ...& X_{\mathrm{ss}}^{*}\left(0\right)\end{array}\right]}^T$

Wherein () ^*represent conjugate operation, () ^tthe transposition of representing matrix.Therefore the dimension N=4G of IFFT conversion.

(4) in order to can the average power of restriction origin light signal, to mapping output signals X _mappingbe multiplied by a dimensional variation factor, this factor can determine the average power that IFFT outputs signal, and can determine further to send light signal average power.Signal X after dimensional variation _scaledbe input to IFFT, the IFFT computing that the present invention adopts is:

$\begin{array}{c}x\left(k\right)=\frac{1}{\sqrt{N}}\underset{n=0}{\overset{N-1}{\mathrm{\Σ}}}{X}_{\mathrm{scaled}}\left(n\right)\exp \left(\frac{j2\mathrm{\πnk}}{N}\right)\\ =\frac{1}{\sqrt{N}}\underset{n=0}{\overset{N-1}{\mathrm{\Σ}}}\mathrm{\α}{X}_{\mathrm{mapping}}\left(n\right)\exp \left(\frac{j2\mathrm{\πnk}}{N}\right)\end{array}$

Wherein k=0,1 ..., N-1.Convert the gross energy of input and output discrete signals and the constant character of average power according to IFFT, can obtain the dimensional variation factor is:

$\mathrm{\α}={\mathrm{\σ}}_x\sqrt{\frac{N}{\underset{n=0}{\overset{N-1}{\mathrm{\Σ}}}{\left|X_{\mathrm{mapping}}\left(n\right)\right|}^2}}$

Wherein σ _xrepresent the mean square deviation of IFFT output signal, || represent the modular arithmetic of signal.

(5) LED normally works usually, time particularly in indoor visible light communication, in order to meet the needs of room lighting, needs in LED drive circuit, add DC bias signal B _dC.Therefore the signal of load information and DC bias signal sum are as the drive singal x of LED _lEDt (), can be expressed as:

x _LED(t)＝x _clipped(t)+B _DC

If when LED is in linear work district, the input drive signal dynamic range of requirement is (i _min, i _max), the amplitude limit thresholding ε of IFFT output time-domain signal x (k) can be determined thus _bottomand ε _topfor:

ε _bottom＝max{i _min-B _DC，0}

ε _top＝i _max-B _DC

According to the character of ACO-OFDM modulation, IFFT output signal has skew-symmetry (Antisymmetry), namely

$\begin{array}{cc}x\left(k\right)=-x(k+\frac{N}{2})& 0<k<\frac{N}{2}\end{array}$

So as amplitude limit thresholding ε b _ottom=0, ε _topdesirable amplitude limit is called, now the amplitude limit not information that comprises of lossing signal during=+ ∞.And because the setting of DC bias signal and the restriction of LED physical characteristic in real system, x (k) signal usually carries out upper and lower bilateral amplitude limit possibly.

To the bilateral clipping operation of x (k) as shown in Figure 5, corresponding amplitude limit is expressed as function to real system:

$x_{\mathrm{clipped}}\left(k\right)=\left\{\begin{array}{cc}{\mathrm{\&epsiv;}}_{\mathrm{top}}& x\left(k\right)\&GreaterEqual;{\mathrm{\&epsiv;}}_{\mathrm{top}}\\ x\left(k\right)& {\mathrm{\&epsiv;}}_{\mathrm{bottom}}<x\left(k\right)<{\mathrm{\&epsiv;}}_{\mathrm{top}}\\ {\mathrm{\&epsiv;}}_{\mathrm{bottom}}& x\left(k\right)\&le;{\mathrm{\&epsiv;}}_{\mathrm{bottom}}\end{array}\right.$

According to central-limit theorem, when number of users is more and IFFT dimension is larger (usual N >=64), time-domain signal x (k) Gaussian distributed that IFFT exports.Non-linear amplitude limit is carried out to Gaussian Profile signal, is equivalent to signal and has been multiplied by a decay factor, add the random noise of and the incoherent non-gaussian distribution of signal, be expressed from the next:

x _clipped(k)＝ρx(k)+n _clip

E[x(k)n _clip]＝0

Wherein ρ represents decay factor, n _cliprepresent clipped noise, mathematic expectaion is asked in E [] expression.Decay factor can be calculated by following formula:

$\mathrm{\&rho;}=\frac{E\left[x{x}_{\mathrm{clipped}}\right]}{{\mathrm{\&sigma;}}_x^2}=\frac{1}{{\mathrm{\&sigma;}}_x^2}{\&Integral;}_{-\&infin;}^{\&infin;}\frac{{\mathrm{xx}}_{\mathrm{clipped}}}{{\mathrm{\&sigma;}}_x\sqrt{2\mathrm{\&pi;}}}(-\frac{x^2}{2{\mathrm{\&sigma;}}_x^2})\mathrm{dx}=Q\left({\mathrm{\&lambda;}}_{\mathrm{bottom}}\right)-Q\left({\mathrm{\&lambda;}}_{\mathrm{top}}\right)$

Wherein λ _bottomand λ _toprepresent normalized amplitude limit bound respectively, Q () represents the complimentary cumulative function of the normal distribution of zero mean unit variance, is expressed as follows respectively:

$\begin{array}{c}{\mathrm{\&lambda;}}_{\mathrm{bottom}}=\frac{{\mathrm{\&epsiv;}}_{\mathrm{bottom}}}{{\mathrm{\&sigma;}}_x}\\ {\mathrm{\&lambda;}}_{\mathrm{top}}=\frac{{\mathrm{\&epsiv;}}_{\mathrm{top}}}{{\mathrm{\&sigma;}}_x}\\ Q\left(u\right)=\frac{1}{\sqrt{2\mathrm{\&pi;}}}{\&Integral;}_u^{\&infin;}\exp \left(\frac{-{\mathrm{\&upsi;}}^2}{2}\right)\mathrm{d\&upsi;}\end{array}$

(6) in order to effectively eliminate the impact of light radio channel multi-path effect, to sliced output signal x _clippedadd cyclic prefix CP.The length of Cyclic Prefix should be more than or equal to the maximum delay of channel.CP length added by supposing is L, by x _clippedl the element below of signal copies to x _clippedbefore, forming a length is the parallel signal of L+N.This signal is input to analog-to-digital conversion module after parallel-serial conversion, and analog-to-digital conversion module adopts the raised cosine filter of roll-off factor β=1.Finally add DC bias signal B _dCform LED drive singal x _lEDthe intensity of (t) directly modulation LED, the average power of LED utilizing emitted light signal is:

E[x _LED(t)]＝σ _x(φ(λ _bottom)-φ(λ _top)+λ _topQ(λ _top)+λ _bottom(1-Q(λ _bottom)))+B _DC

Wherein φ () represents the probability-distribution function of standardized normal distribution

$\mathrm{\&phi;}\left(u\right)=\frac{1}{\sqrt{2\mathrm{\&pi;}}}\exp \left(\frac{-u^2}{2}\right)$

See accompanying drawing 4, the embodiment of light MC-CDMA system receiving terminal is:

(1) intensity modulated direct-detection system (IM/DD) can not transmit phase place and frequency information, and therefore light wireless system is a linear baseband transmission system.LED light signal is transmitted through indoor wireless channels, and when transmitting terminal and receiving terminal invariant position, light wireless channel can regard quasi-static channel opens as, and its channel impulse response h (t) is nonnegative value.Light signal is after multi-path transmission, light signal is converted to the signal of telecommunication by photoelectric detector (being made up of photodiode and transimpedance amplifier), and wherein the thermal noise of itself of the noise that causes of bias light and circuit can concentrate the white noise n thinking Gaussian Profile _aWGN, its bilateral power spectral density is N ₀/ 2.The signal of telecommunication so received is:

y(t)＝h(t)*x _LED(t)+n _AWGN(t)

Wherein * represents linear convolution operation.The impulse response of room light wireless multi-path channels is similar with radio-frequency (RF) communication system, also can be expressed as:

$h\left(t\right)=\underset{l=0}{\overset{P-1}{\mathrm{\&Sigma;}}}{h}_l\mathrm{\&delta;}(t-\mathrm{l\&Delta;\&tau;})$

Wherein P represents the total path number of multipath channel, h _lrepresent the l footpath fading channel factor, Δ τ represents the time delay of channel, and δ () is Dirac function.Therefore, what receive can be expressed as by noise jamming multipath signal y (t):

$y\left(t\right)=\underset{l=0}{\overset{P-1}{\mathrm{\&Sigma;}}}{h}_l{x}_{\mathrm{LED}}(t-\mathrm{l\&Delta;\&tau;})+n_{\mathrm{AWGN}}$

Supposing the system ideal synchronisation, y (t) deletes cyclic prefix CP after uniform sampling.Suppose that Cyclic Prefix L is more than or equal to the maximum delay expansion of wireless channel, eliminate the interference (ICI) between receiving terminal carrier wave and intersymbol interference (ISI), the N road parallel signal exported after serial to parallel conversion can be expressed as:

$\left[\begin{array}{c}y\left(0\right)\\ y\left(1\right)\\ .\\ .\\ .\\ y(N-1)\end{array}\right]=\left[\begin{array}{cccccccc}{h}_0& 0& ...& h_P& h_{P-1}& ...& h_2& h_1\\ h_1& h_0& ...& 0& h_P& ...& h_3& h_2\\ .& & & .& & & .& \\ .& & & .& & & .& \\ .& & & .& & & .& \\ h_P& h_{P-1}& ...& h_0& 0& ...& 0& 0\\ 0& h_P& ...& h_1& h_0& ...& 0& 0\\ .& & & .& & & .& \\ .& & & .& & & .& \\ .& & & .& & & .& \\ 0& 0& ...& h_{P-1}& h_{P-2}& ...& h_0& 0\\ 0& 0& ...& h_P& h_{P-1}& ...& h_1& h_0\end{array}\right]\left[\begin{array}{c}{x}_{\mathrm{LED}}\left(0\right)\\ x_{\mathrm{LED}}\left(1\right)\\ .\\ .\\ .\\ x_{\mathrm{LED}}(N-1)\end{array}\right]+\left[\begin{array}{c}{n}_{\mathrm{AWGN}}\left(0\right)\\ n_{\mathrm{AWGN}}\left(1\right)\\ .\\ .\\ .\\ n_{\mathrm{AWGN}}(N-1)\end{array}\right]$

For simplicity, above formula can be abbreviated as:

y＝hx _LED+n _AWGN

Wherein h is the circular matrix of a N × N, and only relevant to channel fading coefficient.

(2) signal y is input to FFT module, and time-domain signal is converted into frequency-region signal, and the signal Y of output is:

Y＝F(hx _LED+n _AWGN)

＝Fh(αρF ^HX _mapping+F ^HN _clip+B _DC)+N _AWGN

Wherein () ^hthe conjugate transpose operation of representing matrix, F represents the unit Discrete Fourier transform of N × N,

$F=\frac{1}{\sqrt{N}}\left[\begin{array}{ccccc}1& 1& 1& ...& 1\\ 1& w& w^2& ...& w^{N-1}\\ 1& w^2& w^4& ...& w^{2(N-1)}\\ & .& & & .\\ & .& & & .\\ & .& & & .\\ 1& w^{N-1}& w^{2(N-1)}& ...& w^{(N-1)(N-1)}\end{array}\right]$

Wherein n _clipand N _aWGNrepresent clipped noise and the additive Gaussian noise of frequency domain respectively.DC bias signal B _dCafter Fourier transform, become the DC component in Y, it does not affect the demodulation of information.The random clipped noise n of non-gaussian distribution _clipafter FFT, can think the noise becoming Gaussian Profile, its noise variance can be expressed as:

$\begin{array}{c}{\mathrm{\&sigma;}}_{\mathrm{clip}}^2={\mathrm{\&sigma;}}_x^2(\mathrm{\&rho;}(1+{\mathrm{\&lambda;}}_{\mathrm{bottom}}^2)-2{\mathrm{\&rho;}}^2-{\mathrm{\&lambda;}}_{\mathrm{bottom}}(\mathrm{\&phi;}\left({\mathrm{\&lambda;}}_{\mathrm{bottom}}\right)-\mathrm{\&phi;}\left({\mathrm{\&lambda;}}_{\mathrm{top}}\right))\\ -\mathrm{\&phi;}\left({\mathrm{\&lambda;}}_{\mathrm{top}}\right)({\mathrm{\&lambda;}}_{\mathrm{top}}-{\mathrm{\&lambda;}}_{\mathrm{bottom}})+Q\left({\mathrm{\&lambda;}}_{\mathrm{top}}\right){({\mathrm{\&lambda;}}_{\mathrm{top}}-{\mathrm{\&lambda;}}_{\mathrm{bottom}})}^2)\end{array}$

Make H=FhF ^h, because the Fourier transform of circular matrix is diagonal matrix, so H matrix can be expressed as:

Definition H matrix the elements in a main diagonal H (k) is the channel frequency domain response factor.Therefore, from channel on the impact sending frequency-region signal, the frequency domain response factor being multiplied by a channel to each subcarrier of frequency-region signal of making a start is equivalent to.

(3) transmitting terminal is input to the signal X of IFFT _mappinghave hermitian symmetry, and only have odd subcarriers to comprise information, even subcarriers is 0 entirely.The first half odd subcarriers signal also only extracting the frequency-region signal that FFT exports that receiving terminal is corresponding.At transmitting terminal, dimensional variation and clipping operation are carried out to signal, therefore with the distortion of equalizer compensation to signal, namely obtained signal Y divided by invariant α ρ _eq(m) be:

$Y_{\mathrm{eq}}\left(m\right)=H(2m+1)\underset{u=0}{\overset{U-1}{\mathrm{\&Sigma;}}}{c}_u\left(m\right)X_u\left(i\right)+\frac{{\mathrm{\&sigma;}}_{\mathrm{clip}}}{\mathrm{\&alpha;\&rho;}}H(2m+1)N_{\mathrm{CN}}(2m+1)+\frac{{\mathrm{\&sigma;}}_{\mathrm{AWGN}}}{\mathrm{\&alpha;\&rho;}}{N}_{\mathrm{CN}}(2m+1)$

Wherein H (2m+1) represents 2m+1 element on H matrix leading diagonal, m=0,1 ..., G-1, N _cN(2m+1) be average be 0, variance is the white Gaussian noise of 1, σ _clip and σ _aWGNrepresent the standard deviation of amplitude limit Gaussian noise and additive Gaussian noise respectively.

(4) to Y _eqm () signal diversifying receives, collect the signal energy being dispersed in frequency domain.If reception user is r, each Y _eqm () is multiplied by one and is merged factor d _rsue for peace after (m), when receiving terminal adopts different merge algorithms, merge coefficient d _rm () is different.A judgment variables υ is obtained after merging _r, can be expressed as:

$\begin{array}{c}{\mathrm{\&upsi;}}_r=\underset{m=0}{\overset{G-1}{\mathrm{\&Sigma;}}}\underset{u=0}{\overset{U-1}{\mathrm{\&Sigma;}}}{X}_u\left(i\right)c_u\left(m\right)d_r\left(m\right)H(2m+1)+\underset{m=0}{\overset{G-1}{\mathrm{\&Sigma;}}}\frac{{\mathrm{\&sigma;}}_{\mathrm{clip}}}{\mathrm{\&alpha;\&rho;}}{d}_r\left(m\right)H(2m+1)N_{\mathrm{CN}}(2m+1)\\ +\underset{m=0}{\overset{G-1}{\mathrm{\&Sigma;}}}\frac{{\mathrm{\&sigma;}}_{\mathrm{AWGN}}}{\mathrm{\&alpha;\&rho;}}{d}_r\left(m\right)N_{\mathrm{CN}}(2m+1)\\ =\underset{m=0}{\overset{G-1}{\mathrm{\&Sigma;}}}{X}_r\left(i\right)c_r\left(m\right){\text{d}}_r\left(m\right)H(2m+1)+\underset{m=0}{\overset{G-1}{\mathrm{\&Sigma;}}}\underset{u=0,u\&NotEqual;r}{\overset{U-1}{\mathrm{\&Sigma;}}}{X}_u\left(i\right)c_u\left(m\right)d_r\left(m\right)H(2m+1)\\ +\underset{m=0}{\overset{G-1}{\mathrm{\&Sigma;}}}\frac{{\mathrm{\&sigma;}}_{\mathrm{clip}}}{\mathrm{\&alpha;\&rho;}}{d}_r\left(m\right)H(2m+1)N_{\mathrm{CN}}(2m+1)+\underset{m=0}{\overset{G-1}{\mathrm{\&Sigma;}}}\frac{{\mathrm{\&sigma;}}_{\mathrm{AWGN}}}{\mathrm{\&alpha;\&rho;}}{d}_r\left(m\right)N_{\mathrm{CN}}(2m+1)\end{array}$

Wherein Section 1 represents the decision signal desired by user r, and Section 2 represents multi-user interference (MUI) signal of other user to users r, and Section 3 represents the Gaussian reflectivity mirrors that amplitude limit brings, and Section 4 represents channel white Gaussian noise.

Because the frequency domain response factor of channel is different, the impact that frequency-region signal different sub carrier is subject to channel is different, and therefore signal is after channel, and the orthogonality of different user is destroyed.Orthogonality is recovered to merge the impact that (Orthogonality Restoring Combining, ORC) algorithm can eliminate channel, recovers the orthogonality between different user.The merging factor of this algorithm is:

$d_r\left(m\right)=\frac{c_r\left(m\right)H^{*}(2m+1)}{{\left|H(2m+1)\right|}^2}$

Corresponding judgment variables is:

${\mathrm{\&upsi;}}_r=\underset{m=0}{\overset{G-1}{\mathrm{\&Sigma;}}}{X}_r\left(i\right)+\frac{{\mathrm{\&sigma;}}_{\mathrm{clip}}}{\mathrm{\&alpha;\&rho;}}\underset{m=0}{\overset{G-1}{\mathrm{\&Sigma;}}}{c}_r\left(m\right)N_{\mathrm{CN}}(2m+1)+\frac{{\mathrm{\&sigma;}}_{\mathrm{AWGN}}}{\mathrm{\&alpha;\&rho;}}\underset{m=0}{\overset{G-1}{\mathrm{\&Sigma;}}}\frac{c_r\left(m\right)H^{*}(2m+1)}{{\left|H(2m+1)\right|}^2}{N}_{\mathrm{CN}}(2m+1)$

Orthogonality recovery merge algorithm eliminates multi-user interference item obviously, further can according to v _revery bit signal to noise ratio E of demodulator input signal can be calculated _b/ N ₀.The signal to noise ratio that orthogonality recovers merge algorithm modulator input end signal is:

${\left(\frac{E_b}{N_0}\right)}_{\mathrm{ORC}}=\frac{{\left(\mathrm{\&alpha;\&rho;G}\right)}^2}{{\log }_2\left(M\right)(G{\mathrm{\&sigma;}}_{\mathrm{clip}}^2+\underset{m=0}{\overset{G-1}{\mathrm{\&Sigma;}}}\frac{{\mathrm{\&sigma;}}_{\mathrm{AWGN}}^2}{{\left|H(2m+1)\right|}^2})}$

(5) judgment variables υ _rbe input to demodulator, demodulation exports the binary information bit b ' of user r _r.According to every bit signal to noise ratio E _b/ N ₀the theoretical error rate of system can be obtained.The error rate theoretical formula of M rank quadrature amplitude modulation is:

$\mathrm{BER}=\frac{\sqrt{M}-1}{\sqrt{M}{\log }_2\left(\sqrt{M}\right)}\mathrm{erfc}\left(\sqrt{\frac{3{\log }_2\left(M\right)}{2(M-1)}\frac{E_b}{N_0}}\right)$

Wherein erfc () represents Gauss error function.

Emulation experiment

Compared with theoretical analysis result by emulation experiment, the correctness of verification system design and feasibility.

Simulation parameter is arranged: adopt 4QAM modulation, modulation symbol average power is normalized to 1.Adopt orthogonal GOLD sequence as frequency expansion sequence, sequence length is G=128, chooses m sequence preferably to being g ₀(x)=1+x ⁴+ x ⁷, g ₁(x)=1+x+x ²+ x ³+ x ⁷.The dimension of number of users U=8, IFFT/FFT is the mean square deviation that N=512, IFFT conversion outputs signal is σ _x=0.5.Light source of making a start chooses LUW W5SM white light LEDs (model: LUWW5SM-KXKY-6P7Q-Z, series: Golden manufacturer: OSRAM Opto Semiconductors Inc.), and hypothesis is when input signal is at (i _min=0.1, i _max=1) during interval, LED is operated in the range of linearity.As direct current biasing B _dC< i _mintime, be called deficient bias conditions, now signal is all subject to amplitude limit on upper and lower both sides, therefore is called bilateral amplitude limit, as B is chosen in emulation _dC=0.04, so clip level is ε _bottom=0.06, ε _top=0.96.Work as i _min≤ B _dC< i _maxtime, only may there is top amplitude limit in signal, be called monolateral amplitude limit, as B is chosen in emulation _dC=0.5, so clip level is ε _bottom=0, ε _top=0.5.Emulation also have chosen a kind of ideal situation and is called desirable amplitude limit, and namely clip level is: ε _bottom=0, ε _top=+∞.

Light wireless communication channel model: room light wireless channel is often divided into line-of-sight transmission (Line-of-Sight, LOS) and non-line-of-sight (being also called scattering) transmission (Non-Line-of-Sight, NLOS) two kinds, the channel model that emulation adopts is:

(1) line-of-sight transmission

Under sighting condition, wireless signal is unobstructedly at linear transmission between light signal and receiving terminal detector of making a start.Without loss of generality, suppose that channel impulse response is:

h(t)＝δ(t)

(2) scattering transmission

Scatter channel adopts roof reflector (Ceiling-Bounce) model, this channel model for only consider infinitely-great bright Bert (Lambertian) reflecting surface, transmitting terminal and a receiving terminal coordination and be all in the mid range in house, light has good simulation at the communication environment of roof generation individual reflection.Channel impulse response is:

$h_l(t,a)=\frac{6a^6}{{(t+a)}^7}u\left(t\right)$

Wherein u (t) representation unit step function, parameter a and root mean square postpone expansion D _rmsit is relevant,

$D_{\mathrm{rms}}=\frac{a}{12}\sqrt{\frac{13}{11}}$

Usual D _rmsspan representative value in indoor is between 2 nanoseconds to 20 nanoseconds.If D _rms=8 nanoseconds, Δ τ=0.75 nanosecond, the number of path P=15 of scatter channel, circulating prefix-length L=16.To h _l(t, a) carries out uniform sampling and obtains channel coefficients, be then normalized channel coefficients, makes it meet:

||h(t)|| ²＝1

Wherein || || represent 2 norms.The channel frequency domain response factor can be obtained by channel coefficients, shown in Fig. 6 as N=512 channel frequency response factor map of magnitudes.

Simulation result:

Fig. 7 and system bit error rate (Bit Error Rate, the BER) performance under Figure 8 shows that line of signt transmission channel and scattering multipath channel, abscissa represents signal to noise ratio E _b/ N ₀, unit is decibel (dB), and ordinate represents bit error rate.The solid line of tape symbol " " represents error rate theoretical value during desirable amplitude limit, BER Simulation value during the desirable amplitude limit of the represented by dotted arrows of tape symbol " ■ ".The solid line of tape symbol " ◇ " represents error rate theoretical value during monolateral amplitude limit, BER Simulation value during the monolateral amplitude limit of the represented by dotted arrows of tape symbol " ◆ ", the solid line of tape symbol " △ " represents error rate theoretical value during bilateral amplitude limit, BER Simulation value during the bilateral amplitude limit of the represented by dotted arrows of tape symbol " ▲ "

As can be seen from Figure 7, error rate of system simulation result and theory analysis value are identical, demonstrate the correctness of present system design, specific implementation process analysis and optimum configurations.During desirable amplitude limit, clipped noise is zero, and the information of transmission is not by the impact of amplitude limit, and therefore system error performance is best.Monolateral amplitude limit and bilateral amplitude limit all introduce clipped noise, and system error performance declines.Such as when signal to noise ratio is at 12 decibels, the desirable clipping situation error rate can reach 3.43 × 10 ^-5, and during bilateral and monolateral clipping situation, the error rate is respectively 2.95 × 10 ^-4with 1.5 × 10 ^-4.

As can be seen from accompanying drawing 8, error rate of system simulation result and theory analysis value are also consistent.Scattering multipath channel bit error rate performance is obviously poor than line of signt transmission channel.Such as when desirable amplitude limit, the error rate will reach 10 ^-4, the signal to noise ratio (E needed under scatter channel _b/ N ₀) more than line-of-sight transmission channel about 19.5 decibels.Simultaneously because the effect of amplitude limit, system error performance is deteriorated.

More than the specific embodiment of the present invention and simulating, verifying.It should be pointed out that those of ordinary skill in the art can clearly understand, present system design for above embodiment and emulation only for illustration of with the correctness of verification method and feasibility, and be not limited to the inventive method.Although effectively can illustrate and describe the present invention by embodiment, there is many changes and do not depart from spirit of the present invention in the present invention.Without departing from the spirit and substance of the case in the method for the present invention, those skilled in the art are when making various corresponding change or distortion according to the inventive method, but these change accordingly or are out of shape the protection range all belonging to the inventive method and require.

Claims (5)

1. CDMA is in conjunction with a light MC-CDMA system communication means of ACO-OFDM, it is characterized in that: specifically implement according to following steps:

Transmitting terminal:

Step 1: the modulation symbol spread spectrum of each user, is added the spread-spectrum signal of all users, generates the spread-spectrum signal that comprises all user profile;

Step 2: map operation is carried out to spread-spectrum signal, mapping output signals meets hermitian symmetry characteristic, is a real signal to ensure that IFFT converts output signal;

Step 3: to the signal dimensional variation after mapping, changed factor determines the variance of IFFT output signal, reaches the object of restriction LED luminous power further;

Step 4: carry out IFFT conversion, and to output signal amplitude limit, using the DC bias signal sum of sliced output signal and LED as the drive singal of LED, it can be made to be operated in linear zone;

Step 5: Cyclic Prefix is added to the signal after amplitude limit, adds DC bias signal after parallel-serial conversion and digital-to-analogue conversion, driving LED light source luminescent;

Receiving terminal:

Step 1: receiver sensed light signal, to signal of telecommunication discretization after opto-electronic conversion, serioparallel exchange, is input to FFT after deleting Cyclic Prefix;

Step 2: only extract the odd subcarriers signal of the first half in FFT output signal at receiving terminal, then to the signal equalization process of extracting, to compensate the signal attenuation and the dimensional variation factor that amplitude limit brings;

Step 3: the signal diversifying exported by equalizer receives, and demodulation exports the information bit of user.

2. a kind of CDMA according to claim 1 is in conjunction with the light MC-CDMA system communication means of ACO-OFDM, it is characterized in that: transmitting terminal step 2 pair spread-spectrum signal map operation, spread-spectrum signal X _ssm () can be expressed as:

Wherein c _um () represents that the length of user u is the m position of G frequency expansion sequence, U represents total number of users, X _ui () represents i-th modulation symbol of u user, the average power of modulation symbol is 1; Mapping output signals meets hermitian symmetry characteristic, and wherein only odd subcarriers comprises information, and even subcarriers is all 0, and mapping output signals is:

Wherein () ^*represent conjugate operation, () ^tthe transposition of representing matrix.

3. a kind of CDMA according to claim 1 is in conjunction with the light MC-CDMA system communication means of ACO-OFDM, it is characterized in that: transmitting terminal step 3 is to the signal dimensional variation after mapping, convert the gross energy of input and output discrete signals and the constant character of average power according to IFFT, the dimensional variation factor is:

Wherein σ _xrepresent the mean square deviation of IFFT output signal, || represent the modular arithmetic of signal, N represents the dimension N=4G that IFFT converts.

4. a kind of CDMA according to claim 1 is in conjunction with the light MC-CDMA system communication means of ACO-OFDM, it is characterized in that: transmitting terminal step 4 is carried out to signal limiter after IFFT, and IFFT computing is:

Wherein k=0,1 ... N-1; IFFT conversion output signal x (k) is ambipolar sequence of real numbers, sliced output signal x _clippedthe DC bias signal B of (k) and LED _dCsum is as the drive singal of LED, if the dynamic range of input signal corresponding when making LED be operated in linear zone is (i _min, i _max), the difference of DC bias signal size, will cause limitation signal x _clippedk the dynamic range of () is also different, the amplitude limit thresholding ε of IFFT output time-domain signal _bollomand ε _lopfor:

ε _bottom＝max{i _min-B _DC，0}

ε _top＝i _max-B _DC。

5. a kind of CDMA according to claim 1 and 2 is in conjunction with the light MC-CDMA system communication means of ACO-OFDM, it is characterized in that: the hermitian symmetry and the only odd subcarriers load information that correspond to the IFFT input signal of transmitting terminal in receiving terminal step 2, so only extract the first half odd subcarriers signal in FFT output signal at receiving terminal, just can demodulate all transmission informations, then equilibrium treatment is done to the signal extracted, to compensate the signal attenuation and the dimensional variation factor that amplitude limit brings, namely divided by amplitude limit decay factor and the change of scale factor.