CN101909035A - Method for recognizing subcarrier modulation modes of orthogonal frequency division multiplexing signal in wireless communication - Google Patents

Abstract

The invention relates to a method for recognizing subcarrier modulation modes of an orthogonal frequency division multiplexing signal in wireless communication, belonging to the field of wireless communication signal processing. The method comprises the following steps of: firstly; preprocessing a received radio-frequency signal by a receiving end of a wireless communication system; extracting N symbols from the received orthogonal frequency division multiplexing signal; respectively carrying out complementation, blind channel estimation and balance on the N symbols to acquire a modulation signal on a data subcarrier so that the modulation signal is matched with the planispheres of all the subcarrier modulation modes of the orthogonal frequency division multiplexing signal; and selecting the modulation mode with the maximum matching degree as a recognition result. An OFDM (Frequency Division Multiplexing) robust blind channel estimation algorithm when a virtual subcarrier exists is adopted in the method of the invention, which is suitable for a time-varying multi-path environment for balancing a non-ideal channel and has the advantages of simple flows, little amount of calculation, short observation time and time-varying channel environment suitability.

Description

The method of identification subcarrier modulation modes of orthogonal frequency division multiplexing signal in the radio communication

Technical field

The present invention relates to the method for identification subcarrier modulation modes of orthogonal frequency division multiplexing signal in a kind of radio communication, belong to the wireless communication signals process field.

Background technology

OFDM (hereinafter to be referred as OFDM) but technology because of characteristics such as its high spectrum utilization contrary frequency Selective Fading Channel and low complex degree equalizations, be widely used in the communication protocol of a plurality of formal commercializations, as digital broadcasting, IEEE 802.11 and IEEE 802.16 standards etc. in Europe, also be one of indispensable technology of the 4th third-generation mobile communication system.

The Modulation identification technology of signal has a wide range of applications in fields such as military surveillance, electronic countermeasures and information securities.Causing that in recent years in the cognition wireless electrical domain of extensive concern, Modulation identification technology can help cognitive user to find out the communications status of authorized user better, also begins to obtain people's attention.Therefore the Modulation identification technology of studying ofdm system is significant.

Present most Modulation Recognition supposes that all received signal passed through perfect channel equalization, even and the actual reception machine is estimated still can produce channel estimation errors by channel by training sequence.Modulation Recognition is applied to blind receiver usually, and this moment, channel estimation errors more be can not ignore.Document " Li Peng, Wang Fuping, Wang Zanji, Modulation Recognition of Communication Signal algorithm in the time-variant multipath channel, Tsing-Hua University's journal (natural science edition), 2007,47 (7): 1097-1100." by the cascade system that design is made up of segmentation blind equalization, parameter Estimation and Modulation Identification, solved single-carrier signal Modulation Identification problem in the time-variant multipath channel.Yet do not have research at present in the document at the ofdm signal subcarrier modulation modes identification problem under the time-variant multipath channel condition.

In the prior art, OFDM robust blind Channel Estimation algorithm when the virtual subnet carrier wave is arranged, for example based on the OFDM blind Channel Estimation algorithm of subspace " B.Su; P.P.Vaidyanathan.Subspace-based blind channel identificaionfor cyclic prefix systems using few received blocks[J] .IEEE Trans.Signal Processing; 2007,55 (10): 4979-4993. " adopt a spot of observation signal can obtain abundant Spatial Dimension by Search Space Smoothing.But when ofdm signal contained the virtual subnet carrier wave, space smoothing can't solve the scarce order problem of frequency-region signal matrix, so performance sharply descends.

Communications Conference Proceedigs, 1999,1:432-436. " in can't be operated in because of its observation time is long based on the Modulation Recognition of Higher Order Cumulants the time in the changing environment.The non-ideal communication channel equilibrium causes the subcarrier-modulated signal to produce the shake of phase place and amplitude, and the distortion of the planisphere of modulation signal when low signal-to-noise ratio is very serious.Therefore document " K.Woo; C.Kok; Clustering based distribution fitting algorithm for automatic modulation recognition[C] .IEEE ISCC; 2007:13-18. " in based on the clustering algorithm of planisphere and document " S.Barbarossa; A.Swami; B.Sadler.Classificaiton of digital constellations under unknown multipath propagationcondition[C] .Proc.SPIE, 2000,4045 (175): 175-186. " the symbol matching algorithm (Alphabet MatchedAlgorithm) based on planisphere in all can't operate as normal.

Summary of the invention

The objective of the invention is to propose the method for identification subcarrier modulation modes of orthogonal frequency division multiplexing signal in a kind of radio communication, a kind of blind receiver system that is made of blind Channel Estimation and the cascade of blind Modulation Identification module is provided, under the situation that is implemented in the virtual subnet carrier wave ofdm signal being carried out blind Channel Estimation, and under time-variant multipath channel, the ofdm signal subcarrier modulation modes is discerned.

The method of identification subcarrier modulation modes of orthogonal frequency division multiplexing signal in the radio communication that the present invention proposes may further comprise the steps:

(1) receiving terminal in the wireless communication system carries out preliminary treatment to the radiofrequency signal that receives, and its step is as follows:

(1-1) to the radiofrequency signal that receives through down-conversion, obtain receiving orthogonal frequency-division multiplex singal r (t):

$r\left(t\right)=\underset{l=1}{\overset{L}{\mathrm{\Σ}}}{h}_l\left(t\right)s(t-{\mathrm{\τ}}_l)e^{j2\mathrm{\π}{f}_{\mathrm{dl}}t}+n\left(t\right),$

In the following formula, the orthogonal frequency-division multiplex singal that s (t) sends for transmitting terminal, the subcarrier number of orthogonal frequency-division multiplex singal is N _Sub, the data subcarrier position is P _Data, the virtual subnet carrier position is P _Null, symbol lengths is L _Ofdm, circulating prefix-length is L _Cp, L is the number of path of wireless channel, τ _lBe the time delay of l paths, h _l(t) be the Rayleigh fading factor of l paths, f _DlBe the Doppler frequency shift of l paths, n (t) is the additive white Gaussian noise of wireless channel;

(1-2) the Cyclic Prefix section r of i symbol of extraction from above-mentioned reception orthogonal frequency-division multiplex singal r (t) _CP(i) and data segment r _M(i), i=1,2 ... N;

(1-3) repeating step (1-2) extracts N symbol from receive orthogonal frequency-division multiplex singal r (t), be expressed as:

r＝[r _CP(1)?r _M(1)?r _CP(2)?r _M(2)…r _CP(N)?r _M(N)]；

(2) receiving terminal in the wireless communication system respectively to an above-mentioned N symbol compensate, blind Channel Estimation and equilibrium, detailed process is as follows:

(2-1) receiving terminal at wireless communication system generates an orthogonal frequency-division multiplex singal w who is made of N symbol, makes the number of the subcarrier of this orthogonal frequency-division multiplex singal w The data subcarrier position

With the virtual subnet carrier position

Respectively with above-mentioned reception orthogonal frequency-division multiplex singal r (t) in subcarrier number N _Sub, data subcarrier position P _DataWith virtual subnet carrier position P _NullIdentical;

(2-2) on the frequency domain of the individual symbol of i ' of the orthogonal frequency-division multiplex singal w that generates, generate a signal w _f, the signal w that is generating _fIn the virtual subnet carrier wave on load white Gaussian noise, the average that makes white Gaussian noise is 0, variance is σ _w ²Make the signal w of generation _fIn data subcarrier on the signal energy that loads be 0;

(2-3) signal w to generating _fObtain time-domain signal w by inverse Fourier transform _t, w _t=Ww _f, wherein W is the inverse Fourier transform matrix, makes time-domain signal w _tSymbol lengths With the symbol lengths L among the above-mentioned reception orthogonal frequency-division multiplex singal r (t) _OfdmIdentical;

(2-4) with the above-mentioned time-domain signal w that obtains _tData segment w as the individual symbol of i ' in the above-mentioned N orthogonal frequency-division multiplex singal _M(i '), w _M(i ')=w _t, and make the Cyclic Prefix section w of the individual symbol of i ' _CP(i ') is data segment w _M(i ') end Individual data, the circulating prefix-length of the individual symbol of i '

With the circulating prefix-length L among the above-mentioned reception orthogonal frequency-division multiplex singal r (t) _CpIdentical;

(2-5) repeating step (2-2)-(2-4),, generate N symbol, be expressed as:

w＝[w _CP(1)?w _M(1)?w _CP(2)?w _M(2)…w _CP(N)?w _M(N)]，i′＝1，2，…N；

(2-6) that N the symbol w and the above-mentioned N that the receives symbol that generate is superimposed, be compensated back signal: v=r+w

(2-7) signal v after the above-mentioned compensation is carried out blind Channel Estimation and equilibrium, extract the modulation signal u on the data subcarrier, u=[u (1) u (2) ... u (N _Sym)], N wherein _SymBe modulation signal quantity;

(3) planisphere of above-mentioned modulation signal u and all subcarrier modulation modes of orthogonal frequency division multiplexing signal is mated, select the modulation system of matching degree maximum, matching process is as follows:

(3-1) n the modulation signal u (n) among the above-mentioned modulation signal u carried out conventional demodulation with the k kind modulation system in all subcarrier modulation modes of orthogonal frequency division multiplexing signal, obtain signal z (n) after n the demodulation, n=1,2 ... N _Sym

(3-2) calculate the Euclidean distance square value of signal z (n) and above-mentioned n modulation signal u (n) after above-mentioned n the demodulation: P (n, k)=‖ z (n)-u (n) ‖ ²

(3-3) repeating step (3-1) and (3-2) is respectively to all N of above-mentioned steps (2-7) _SymIndividual modulation signal carries out demodulation, obtains N _SymSignal z=[z (1) z (2) after the individual demodulation ... z (N _Sym)], calculate N respectively _SymThe Euclidean distance square value of signal and above-mentioned modulation signal u obtains N after the individual demodulation _SymIndividual Euclidean distance square value, the calculation cost functional value:

Similitude between the standard signal of modulation signal u and k kind modulation system is cost function value J ₁(k), cost function value is more little, and modulation signal is similar more to standard signal;

(3-4) according to the coordinate figure of signal z after the above-mentioned demodulation in k kind modulation system planisphere, add up number of signals N after the demodulation at d constellation point place in the k kind modulation system planisphere _{K, d}

(3-5) repeating step (3-4) obtains number of signals [N after the demodulation at each constellation point place in the k kind modulation system planisphere _{K, 1}N _{K, 2}... N _{K, D}], wherein D is the quantity of constellation point in the k kind modulation system planisphere;

(3-6) according to number of signals [N after the demodulation at each constellation point place in the above-mentioned k kind modulation system planisphere _{K, 1}N _{K, 2}N _{K, D}], the calculation cost functional value:

$J_2\left(k\right)=\frac{1}{D}\underset{d=1}{\overset{D}{\mathrm{\Σ}}}{(N_{k,d}-\frac{1}{D}\underset{i=d^{\′}}{\overset{D}{\mathrm{\Σ}}}{N}_{{k,d}^{\′}})}^2$

The be evenly distributed degree of modulation signal u on k kind modulation system planisphere is cost function value J ₂(k), cost function value J ₂(k) more little, the distribution of modulation signal on planisphere is even more;

(3-7) with the above-mentioned cost function value J that obtains ₁(k) and J ₂(k) superpose, obtain cost function value: J (k)=α ₁J ₁(k)+α ₂J ₂(k)

α wherein ₁And α ₂Be setting threshold, α ₁And α ₂Span be [0～1], the matching degree of modulation signal u and k kind modulation system planisphere is cost function value J (k), cost function value J (k) is more little, matching degree is high more;

(3-8) repeating step (3-1)-(3-7) travels through all subcarrier modulation modes of orthogonal frequency division multiplexing signal, obtains cost function value respectively, and wherein the corresponding modulation system of minimum cost functional value is the Modulation Mode Recognition result of modulation signal u.

The method of identification subcarrier modulation modes of orthogonal frequency division multiplexing signal during the present invention proposes and line is communicated by letter, its advantage is:

1, used OFDM robust blind Channel Estimation algorithm when having the virtual subnet carrier wave in the inventive method, solved to a great extent when there is the virtual subnet carrier wave in ofdm signal and cause the problem that existing blind Channel Estimation algorithm can't be worked under the short observation time because of signal matrix lacks order.

2, used ofdm signal subcarrier-modulated recognizer when channel estimation errors is arranged in the inventive method, be applicable to the non-ideal communication channel equilibrium the time become in the multi-path environment, and existing clustering algorithm and symbol matching algorithm based on planisphere identification all can't be worked under this actual environment.

3, the implementation complexity of the inventive method is low, robust blind channel estimation method in the inventive method, its core procedure is the artificial random signal that generates of stack on the frequency domain of received signal, the core procedure of the subcarrier-modulated recognition methods in the inventive method is that signal is carried out conventional demodulation, compute euclidian distances and number of computations variance, the weights of using in the said method can preestablish, so the present invention has the simple and little characteristics of amount of calculation of flow process;

4, the inventive method is applicable to the time varying channel environment, is example with the ofdm signal of IEEE 802.11a standard, and the observation time that robust blind Channel Estimation algorithm needs is 20 OFDM symbols.The quantity of the modulation signal sampled point that Modulation Recognition needs is 500, approximates 10 OFDM symbols.Therefore total observation time is 30 OFDM symbols.And existing based on sef-adapting filter the blind Channel Estimation algorithm and need 100 OFDM symbols at least based on the Modulation Recognition of high-order statistic.Therefore the present invention has the advantage that the time varying channel environment is lacked and be applicable to observation time.

Description of drawings

Fig. 1 is the FB(flow block) of the inventive method.

Embodiment

The method of identification subcarrier modulation modes of orthogonal frequency division multiplexing signal in the radio communication that the present invention proposes, its FB(flow block) may further comprise the steps as shown in Figure 1:

(1) receiving terminal in the wireless communication system carries out preliminary treatment to the radiofrequency signal that receives, and its step is as follows:

(1-1) to the radiofrequency signal that receives through down-conversion, obtain receiving orthogonal frequency-division multiplex singal r (t):

$r\left(t\right)=\underset{l=1}{\overset{L}{\mathrm{\Σ}}}{h}_l\left(t\right)s(t-{\mathrm{\τ}}_l)e^{j2\mathrm{\π}{f}_{\mathrm{dl}}t}+n\left(t\right),$

In the following formula, the orthogonal frequency-division multiplex singal that s (t) sends for transmitting terminal, the subcarrier number of orthogonal frequency-division multiplex singal is N _Sub, the data subcarrier position is P _Data, the virtual subnet carrier position is P _Null, symbol lengths is L _Ofdm, circulating prefix-length is L _Cp, L is the number of path of wireless channel, τ _lBe the time delay of l paths, h _l(t) be the Rayleigh fading factor of l paths, f _DlBe the Doppler frequency shift of l paths, n (t) is an additive white Gaussian noise;

The Rayleigh fading h in l footpath _l(t) satisfy independent same distribution, can be expressed as

Its amplitude a _l(t) Rayleigh distributed, phase place

Obey the even distribution of [0～2 π]; The average of additive white Gaussian noise n (t) is 0, and variance is σ _n ²Signal, Rayleigh fading and Gaussian noise are separate.

Received signal to noise ratio is defined as:

$\mathrm{\ξ}=\frac{P_r-P_n}{P_n}=\frac{E\left[{\left|r\left(t\right)\right|}^2\right]-E\left[{\left|n\left(t\right)\right|}^2\right]}{E\left[{\left|n\left(t\right)\right|}^2\right]}=\frac{E\left[{\left|y\left(t\right)\right|}^2\right]}{E\left[{\left|n\left(t\right)\right|}^2\right]}$

P wherein _rBe received signal power, P _nBe noise power.

(1-2) initial data of ofdm signal is modulated at frequency domain, launches after process IFFT transforms to time domain.The frequency domain data of supposing i OFDM symbol is u _f(i), the transformation matrix of IFFT is W, and the data segment of time domain transmission signal is so: s _M(i)=Wu _f(i); Signal adds Cyclic Prefix (hereinafter to be referred as CP) back to be transmitted in multipath channel, and Cyclic Prefix makes the data segment of time-domain signal and channel carry out circular convolution, is equivalent to the phase multiplication at frequency domain, and establishing i the OFDM symbol data section that receives is r _M(i), the frequency domain value of channel is diagonal matrix H _f, the time domain noise is n (i), so: r _M(i)=WH _fu _f(i)+and n (i), wherein comprised the modulation signal on the data subcarrier among the column vector u (i), also comprised the signal on the virtual subnet carrier wave, its energy is zero:

From above-mentioned reception orthogonal frequency-division multiplex singal r (t), extract the Cyclic Prefix section r of i symbol _CP(i) and data segment r _M(i);

(1-3) repeating step (1-2), i=1,2 ... N, the Cyclic Prefix section and the data segment of N symbol of extraction from receive orthogonal frequency-division multiplex singal r (t) are expressed as:

r＝[r _CP(1)?r _M(1)?r _CP(2)?r _M(2)…r _CP(N)?r _M(N)]

(2) receiving terminal in the wireless communication system to an above-mentioned N symbol compensate, blind Channel Estimation and equilibrium, detailed process is as follows:

(2-1) receiving terminal at wireless communication system generates an orthogonal frequency-division multiplex singal w who is made of N symbol, makes the number of the subcarrier of this orthogonal frequency-division multiplex singal w

The data subcarrier position

With the virtual subnet carrier position

Respectively with above-mentioned reception orthogonal frequency-division multiplex singal r (t) in subcarrier number N _Sub, data subcarrier position P _DataWith virtual subnet carrier position P _NullIdentical;

(2-2) on the frequency domain of the individual symbol of i ' of the orthogonal frequency-division multiplex singal w that generates, generate a signal w _f, the signal w that is generating _fIn the virtual subnet carrier wave on load white Gaussian noise, the average that makes white Gaussian noise is 0, variance is σ _w ², recommended parameter is: σ _w ²=0.13 (P _r-P _n), P wherein _rBe received signal power, P _nBe noise power; Make the signal w of generation _fIn data subcarrier on the signal energy that loads be 0;

(2-3) make the signal w of generation _fObtain time-domain signal w by inverse Fourier transform _t, w _t=Ww _f, wherein W is the inverse Fourier transform matrix, makes time-domain signal w _tSymbol lengths

With the symbol lengths L among the above-mentioned reception orthogonal frequency-division multiplex singal r (t) _OfdmIdentical;

(2-4) with the above-mentioned time-domain signal w that obtains _tData segment w as the individual symbol of i ' in the above-mentioned N orthogonal frequency-division multiplex singal _M(i '), w _M(i ')=w _t, and make the Cyclic Prefix section w of the individual symbol of i ' _CP(i ') is data segment w _M(i ') end

Individual data, the circulating prefix-length of the individual symbol of i '

With the circulating prefix-length L among the above-mentioned reception orthogonal frequency-division multiplex singal r (t) _CpIdentical;

(2-5) repeating step (2-2)-(2-4), i '=1,2 ... N generates N symbol, is expressed as:

w＝[w _CP(1)?w _M(1)?w _CP(2)?w _M(2)…w _CP(N)?w _M(N)]

(2-6) that N the symbol w and the above-mentioned N that the receives symbol that generate is superimposed, be compensated back signal: v=r+w

(2-7) signal v after the above-mentioned compensation is carried out ofdm signal blind Channel Estimation and channel equalization based on the subspace, referring to document " B.Su; P.P.Vaidyanathan.Subspace-based blind channel identificaion for cyclicprefix systems using few received blocks[J] .IEEE Trans.Signal Processing; 2007,55 (10): 4979-4993. "; Extract the modulation signal u on the data subcarrier, u=[u (1) u (2) ... u (N _Sym)], N wherein _SymBe modulation signal quantity;

(3) planisphere of above-mentioned modulation signal u and all subcarrier modulation modes of orthogonal frequency division multiplexing signal is mated, common modulation system is: binary phase shift keying (hereinafter to be referred as BPSK), Quadrature Phase Shift Keying (hereinafter to be referred as QPSK), comprise the quadrature amplitude modulation (hereinafter to be referred as 16QAM) of 16 kinds of symbols and comprise these four kinds of the quadrature amplitude modulation (hereinafter to be referred as 64QAM) of 64 kinds of symbols, the coupling back obtains four cost function J (k), the matching degree of expression modulation signal and these four kinds of modulation systems, k=1 wherein, 2,3,4 represent four kinds of modulation systems.Choose the pairing modulation system of minimum cost function as recognition result.

Next mating with modulation signal u and QPSK modulation system (k=2) is example, describes the computational process of cost function value J (2) in detail.The computational process all fours of other modulation system cost function value.Matching process is as follows:

(3-1) n the modulation signal u (n) among the above-mentioned modulation signal u carried out conventional demodulation with modulation system QPSK, obtain signal z (n) after n the demodulation; For example u (n)=0.51+j0.43 is with the conventional demodulation of QPSK modulation system, and the result is Z (n) be on the QPSK standard planisphere with the constellation point of u (n) Euclidean distance minimum.

(3-2) the Euclidean distance square value of signal z (n) and above-mentioned n modulation signal u (n) after above-mentioned n the demodulation of calculating: $P(n,2)={\left|\right|u\left(n\right)-z\left(n\right)\left|\right|}^2={\left|\right|0.51+j0.43-\sqrt{2}/2-j\sqrt{2}/2\left|\right|}^2=0.465;$

(3-3) repeating step (3-1) and (3-2) is respectively to all N of above-mentioned steps (2-7) _SymIndividual modulation signal carries out demodulation, obtains N _SymSignal z=[z (1) z (2) after the individual demodulation ... z (N _Sym)], calculate N respectively _SymThe Euclidean distance square value of signal and above-mentioned modulation signal u obtains N after the individual demodulation _SymIndividual Euclidean distance square value, the calculation cost functional value:

Promptly as the matching degree of modulation signal u and k kind modulation system planisphere, cost function value is more little, and matching degree is big more;

N for example _Sym=500 o'clock, respectively 500 modulation signals are finished the work of (3-2) after, the calculation cost function:

$J_1\left(2\right)=\frac{1}{500}\underset{n=1}{\overset{500}{\mathrm{\Σ}}}\{1-e^{-P(n,2)}\}=0.46$

Wherein, cost function value 0.46 is for giving an example.

(3-4) according to the coordinate figure of signal z after the above-mentioned demodulation in k kind modulation system planisphere, add up number of signals N after the demodulation at d constellation point place in the k kind modulation system planisphere _{K, d}

(3-5) repeating step (3-4) obtains number of signals [N after the demodulation at each constellation point place in the k kind modulation system planisphere _{K, 1}N _{K, 2}N _{K, D}], wherein D is the quantity of constellation point in the k kind modulation system planisphere; Satisfy equation

For example the standard constellation point of QPSK modulation system has 4, i.e. D=4, and 500 sampled points the results are shown in Table 1 after through the conventional demodulation of QPSK so;

Table 1 cost function calculation for example

(3-6) according to number of signals [N after the demodulation at each constellation point place in the above-mentioned k kind modulation system planisphere _{K, 1}N _{K, 2}N _{K, D}], the calculation cost functional value:

$J_2\left(k\right)=\frac{1}{D}\underset{d=1}{\overset{D}{\mathrm{\Σ}}}{(N_{k,d}-\frac{1}{D}\underset{d^{\′}=1}{\overset{D}{\mathrm{\Σ}}}{N}_{{k,d}^{\′}})}^2$

For example, calculate average each constellation point number of signals on every side according to table 1

Cost function J ₂(k) be:

The original bit that transmits as convolutional encoding, scrambler with after interweaving, presents approximate random through source encoding.They will be modulated on all constellation point with equal probabilities, so 16QAM or 64QAM received signal be evenly distributed on planisphere, and bpsk signal concentrates near the real number axis.This becomes a key character of Modulation Identification.So cost function value J ₂(k) big more, represent that the number of signals around each constellation point is inhomogeneous more, modulation signal is that the likelihood ratio of BPSK is higher; Cost function value J ₂(k) more little, the distribution of modulation signal on planisphere is even more;

(3-7) with the above-mentioned cost function value J that obtains ₁(k) and J ₂(k) superpose, obtain cost function value:

J(k)＝α ₁J ₁(k)+α ₂J ₂(k)

α wherein ₁And α ₂Be setting threshold, α ₁And α ₂Value as shown in table 2, the matching degree of modulation signal u and k kind modulation system planisphere is cost function value J (k), cost function value J (k) is more little, matching degree is high more;

Weights α during table 2 various constellations figure k ₁And α ₂Value

k	1	2	3	4
					α 1	1.0	1.0	1.0	1.0
α 2	0	1.5×10 -5	1.5×10 -6	6×10 -7

Final cost function is:

J(2)＝α ₁(2)J ₁(2)+α ₂(2)J ₂(2)

＝1.0×0.46+1.5×10 ^-6×4242

＝0.466

So just calculated the cost function value after modulation signal mates with the QPSK planisphere;

(3-8) repeating step (3-1)-(3-7) travels through all subcarrier modulation modes of orthogonal frequency division multiplexing signal, obtains matching degree respectively, and wherein the corresponding modulation system of the highest matching degree is the Modulation Mode Recognition result of modulation signal u.

Claims (1)

1. the method for identification subcarrier modulation modes of orthogonal frequency division multiplexing signal in the radio communication, this method may further comprise the steps:

(1) receiving terminal in the wireless communication system carries out preliminary treatment to the radiofrequency signal that receives, and its step is as follows:

(1-1) to the radiofrequency signal that receives through down-conversion, obtain receiving orthogonal frequency-division multiplex singal r (t):

$r\left(t\right)=\underset{l=1}{\overset{L}{\mathrm{\Σ}}}{h}_l\left(t\right)s(t-{\mathrm{\τ}}_l)e^{j2\mathrm{\π}{f}_{\mathrm{dl}}t}+n\left(t\right),$

In the following formula, the orthogonal frequency-division multiplex singal that s (t) sends for transmitting terminal, the subcarrier number of orthogonal frequency-division multiplex singal is N _Sub, the data subcarrier position is P _Data, the virtual subnet carrier position is P _Null, symbol lengths is L _Ofdm, circulating prefix-length is L _Cp, L is the number of path of wireless channel, τ _lBe the time delay of l paths, h _l(t) be the Rayleigh fading factor of l paths, f _DlBe the Doppler frequency shift of l paths, n (t) is the additive white Gaussian noise of wireless channel;

(1-2) the Cyclic Prefix section r of i symbol of extraction from above-mentioned reception orthogonal frequency-division multiplex singal r (t) _CP(i) and data segment r _M(i), i=1,2 ... N;

(1-3) repeating step (1-2) extracts N symbol from receive orthogonal frequency-division multiplex singal r (t), be expressed as: r=[r _CP(1) r _M(1) r _CP(2) r _M(2) ... r _CP(N) r _M(N)];

(2) receiving terminal in the wireless communication system respectively to an above-mentioned N symbol compensate, blind Channel Estimation and equilibrium, detailed process is as follows:

(2-1) receiving terminal at wireless communication system generates an orthogonal frequency-division multiplex singal w who is made of N symbol, makes the number of the subcarrier of this orthogonal frequency-division multiplex singal w

The data subcarrier position

With the virtual subnet carrier position

Respectively with above-mentioned reception orthogonal frequency-division multiplex singal r (t) in subcarrier number N _Sub, data subcarrier position P _DataWith virtual subnet carrier position P _NullIdentical;

(2-2) on the frequency domain of the individual symbol of i ' of the orthogonal frequency-division multiplex singal w that generates, generate a signal w _f, the signal w that is generating _fIn the virtual subnet carrier wave on load white Gaussian noise, the average that makes white Gaussian noise is 0, variance is σ _w ²Make the signal w of generation _fIn data subcarrier on the signal energy that loads be 0;

(2-3) signal w to generating _fObtain time-domain signal w by inverse Fourier transform _t, w _t=Ww _f, wherein W is the inverse Fourier transform matrix, makes time-domain signal w _tSymbol lengths

With the symbol lengths L among the above-mentioned reception orthogonal frequency-division multiplex singal r (t) _OfdmIdentical;

(2-4) with the above-mentioned time-domain signal w that obtains _tData segment w as the individual symbol of i ' in the above-mentioned N orthogonal frequency-division multiplex singal _M(i '), w _M(i ')=w _t, and make the Cyclic Prefix section w of the individual symbol of i ' _CP(i ') is data segment w _M(i ') end

Individual data, the circulating prefix-length of the individual symbol of i '

With the circulating prefix-length L among the above-mentioned reception orthogonal frequency-division multiplex singal r (t) _CpIdentical;

(2-5) repeating step (2-2)-(2-4),, generate N symbol, be expressed as:

w＝[w _CP(1)?w _M(1)?w _CP(2)?w _M(2)…w _CP(N)?w _M(N)]，i′＝1，2，…N；

(2-6) that N the symbol w and the above-mentioned N that the receives symbol that generate is superimposed, be compensated back signal: v=r+w

(2-7) signal v after the above-mentioned compensation is carried out blind Channel Estimation and equilibrium, extract the modulation signal u on the data subcarrier, u=[u (1) u (2) ... u (N _Sym)], N wherein _SymBe modulation signal quantity;

(3) planisphere of above-mentioned modulation signal u and all subcarrier modulation modes of orthogonal frequency division multiplexing signal is mated, select the modulation system of matching degree maximum, matching process is as follows:

(3-1) n the modulation signal u (n) among the above-mentioned modulation signal u carried out conventional demodulation with the k kind modulation system in all subcarrier modulation modes of orthogonal frequency division multiplexing signal, obtain signal z (n) after n the demodulation, n=1,2 ... N _Sym

(3-2) calculate the Euclidean distance square value of signal z (n) and above-mentioned n modulation signal u (n) after above-mentioned n the demodulation: P (n, k)=‖ z (n)-u (n) ‖ ²

(3-3) repeating step (3-1) and (3-2) is respectively to all N of above-mentioned steps (2-7) _SymIndividual modulation signal carries out demodulation, obtains N _SymSignal z=[z (1) z (2) after the individual demodulation ... z (N _Sym)], calculate N respectively _SymThe Euclidean distance square value of signal and above-mentioned modulation signal u obtains N after the individual demodulation _SymIndividual Euclidean distance square value, the calculation cost functional value:

Similitude between the standard signal of modulation signal u and k kind modulation system is cost function value J ₁(k), cost function value is more little, and modulation signal is similar more to standard signal;

(3-4) according to the coordinate figure of signal z after the above-mentioned demodulation in k kind modulation system planisphere, add up number of signals N after the demodulation at d constellation point place in the k kind modulation system planisphere _{K, d}

(3-5) repeating step (3-4) obtains number of signals [N after the demodulation at each constellation point place in the k kind modulation system planisphere _{K, 1}N _{K, 2}N _{K, D}], wherein D is the quantity of constellation point in the k kind modulation system planisphere;

(3-6) according to number of signals [N after the demodulation at each constellation point place in the above-mentioned k kind modulation system planisphere _{K, 1}N _{K, 2}N _{K, D}], the calculation cost functional value:

$J_2\left(k\right)=\frac{1}{D}\underset{d=1}{\overset{D}{\mathrm{\Σ}}}{(N_{k,d}-\frac{1}{D}\underset{i=d^{\′}}{\overset{D}{\mathrm{\Σ}}}{N}_{{k,d}^{\′}})}^2$

The be evenly distributed degree of modulation signal u on k kind modulation system planisphere is cost function value J ₂(k), cost function value J ₂(k) more little, the distribution of modulation signal on planisphere is even more;

(3-7) with the above-mentioned cost function value J that obtains ₁(k) and J ₂(k) superpose, obtain cost function value:

J(k)＝α ₁J ₁(k)+α ₂J ₂(k)

α wherein ₁And α ₂Be setting threshold, α ₁And α ₂Span be [0～1], the matching degree of modulation signal u and k kind modulation system planisphere is cost function value J (k), cost function value J (k) is more little, matching degree is high more;

(3-8) repeating step (3-1)-(3-7) travels through all subcarrier modulation modes of orthogonal frequency division multiplexing signal, obtains cost function value respectively, and wherein the corresponding modulation system of minimum cost functional value is the Modulation Mode Recognition result of modulation signal u.