CN101257469B - Method for using system information to inhibit phase noise in orthogonal frequency division multiplexing system - Google Patents

Abstract

The invention relates to a method using the system information to restrain the phase noise in the orthogonal frequency division multiplexing system, belonging to the transmission of digital information technical field. The method includes carrying the symbol mapping process at the transmitting terminal after the input signals are processed with scrambling code, channel encoding to generate the data symbols; obtaining the frame head; inserting the system information sub carrier into the data symbols to obtain the frame; multiple connecting the frame head and the frame to obtain the signal frame through forming filter, digital analog conversion, radio frequency up conversion, then the signal frame is transmitted to the receiving end; receiving the signal frame at the receiving end to obtain the system information sub carrier in the signal frame and fetch several sub carrier between the system information sub carrier for phase noise estimation, solve the LS equation to get the frequency domain estimation of the phase noise, finally carry the phase correction in the time domain to get the signal frame after being compensated. The signal frame is carried with de-mapping after being compensated, and output after coding process.

Description

Utilize system information to suppress the method for phase noise in the ofdm system

Technical field

The present invention relates to digital information transmission technical field, (Orthogonal FrequencyDivision Multiplexing OFDM) utilizes system information to suppress the method for phase noise to particularly a kind of OFDM in the system.

Background technology

Subject matter in the current digital information transmission technical (as wireless transmitting system) is how in the process of transmission signals, effectively suppresses the influence of phase noise to signal.Be illustrated in figure 1 as the system diagram of radio-frequency front-end in the wireless transmitting system.At transmitting terminal, export to digital to analog conversion (D/A) through the digital signal of baseband modulator, signal is launched by antenna after passing through low pass filter (LPF), upconverter, radio-frequency filter successively.In wireless propagation environment, signal is subjected to various interference and The noise.At receiving terminal, signal obtains base-band analog signal after passing through radio-frequency filter, low-converter, baseband filter (BBF) successively, exports to band receiver of base after passing through analog-to-digital conversion (A/D) again.Above-mentioned structure of radio-frequency front end has been introduced radio-frequency front-end problem of dtmf distortion DTMFs such as carrier frequency shift, phase noise, IQ imbalance, and than other two kinds of interference, phase noise more is difficult to overcome owing to its randomness.

Phase noise is that phase place change at random, phase fluctuation and the frequency fluctuation that noise effect and the noise effect introduced by externally fed and control circuit cause because the oscillator of receiving terminal and transmitting terminal is heated produces.Phase noise can both be introduced at transmitting terminal transmitter and receiving terminal tuner etc.Adopt high performance oscillator can reduce phase noise, but its price is also more expensive.For the broadcast type system, because the receiver One's name is legion can not use too expensive oscillator.Receiving terminal is compared with transmitting terminal, and phase noise is generally bigger, for guaranteeing receptivity preferably, mainly needs to eliminate the phase noise of user's receiving terminal.

The method of eliminating phase noise mainly contains:

1. utilize CPE (common phase error in the phase noise) alignment technique of pilot tone.After separating mapping, scattered pilot and continuous pilot are influenced by phase noise, and its subcarrier constellation point no longer drops on the real axis of planisphere, and the mean value of each pilot frequency deviation is exactly that CPE estimates.The CPE removing method can only estimate the CPE part in the phase noise, and the dispersion of phase noise part can't compensate.

2. based on the phase noise compensation technology of DCT orthogonal basis.Regard phase noise the set of one group of dextrorotation ripple as, provide least square method (LS) estimated value of this group dextrorotation wave system number according to the pilot frequency information of DVB-T, because pilot data is bigger, computation complexity improved when this method was found the solution the LS equation.

3. recover the phase noise compensation technology of pilot tone based on carrier wave.Use the protection frequency band to prevent of the interference of other subcarrier to the phase reference pilot tone; re-use a digital filter and finish the extraction of phase noise; the bandwidth of this filter has determined to estimate the bandwidth of phase noise, finishes the compensation of phase noise at last by Fourier transform (DFT).This method is carried out Noise Estimation owing to only used a carrier wave to recover pilot tone, if this pilot tone is in deep fade under the multipath situation, estimated performance can't guarantee.

(4.ICI disturbing between subcarrier) self elimination method.The ICI self elimination method is that it is the reindexing form not, but owing to introduce bulk redundancy, has reduced the availability of frequency spectrum not by the method for the another kind of non-data-aided compensation phase noise influence of any pilot data.

Summary of the invention

Method in order effectively to suppress the phase noise in the ofdm system, to the present invention proposes to utilize system information to suppress phase noise in a kind of ofdm system comprises:

Step 101: at transmitting terminal, the signal of input is handled, after chnnel coding handles, carried out sign map and handle through scrambler, generate data symbol; Obtain frame head; N system information subcarrier inserted in the described data symbol, and wherein N is a positive integer; Then the data symbol that has inserted N system information subcarrier is carried out inverse discrete Fourier transformer inverse-discrete, obtain frame; Frame head and frame multiple connection are become signal frame, will be transmitted into receiving terminal after this signal frame process shaping filter, digital to analog conversion, the radio frequency up-conversion process then;

Step 102: at receiving terminal received signal frame, obtain the frame in the signal frame, it is carried out discrete Fourier transform operations recover N system information subcarrier, take out the middle experimental process carrier wave of N system information subcarrier and carry out the phase noise estimation;

Step 103: utilize the system information subcarrier that receives to be the convolution relation of phase noise frequency domain response, use least square method (LS) to try to achieve the frequency domain response of phase noise with the insertion system information of making a start;

Step 104: discrete Fourier transform (DFT) is carried out in the response to described phase noise, tries to achieve the estimated value of phase noise, time domain to received signal frame carry out phasing, the signal frame that is compensated, signal frame to described compensation is separated mapping, and decoding is handled, then output.

In the present invention, the number N of described system information subcarrier is preferably greater than 8.

In step 102 of the present invention, preferably take out middle (N-2Bw) the individual subcarrier of N system information subcarrier and carry out the phase noise estimation, wherein, Bw is the phase noise estimated bandwidth.

In the present invention, least square method described in the described step 103 adopts weighted least-squares method, or total least square method.

Method among the present invention is calculated simply, and the dispersion part in can compensation phase noise has improved the availability of frequency spectrum, can guarantee estimation performance under the multipath situation, thereby can in time follow the tracks of the phse conversion of received signal, suppresses phase noise effectively.

Description of drawings

Fig. 1 is the system diagram of wireless transmitting system radio-frequency front-end in the prior art;

Fig. 2 utilizes system information to suppress the method flow diagram of phase noise in the ofdm system in the embodiment of the invention;

Fig. 3 is the transmitting terminal system diagram of the multi-carrier mode in the embodiment of the invention;

Fig. 4 is the frame head structure chart in the embodiment of the invention;

Fig. 5 is the system information position view in the embodiment of the invention;

Fig. 6 is that the embodiment of the invention utilizes system information to suppress the block diagram of phase noise at receiving terminal;

Fig. 7 is that the out of phase noise is chosen the phase noise figure that bandwidth obtains under Gaussian channel;

Fig. 8 is that the out of phase noise is chosen error sign ratio performance chart under the bandwidth under Gaussian channel;

Fig. 9 is the amplitude-frequency characteristic figure of emulation multipath channel;

Figure 10 is the error sign ratio performance chart of least square method and weighted least-squares method under the emulation multipath channel.

Embodiment

For making purpose of the present invention, content and advantage clearer, embodiment of the present invention is described further in detail below in conjunction with accompanying drawing.

The invention process provides the method for utilizing system information to suppress phase noise in a kind of ofdm system, by using this method, can accurately and timely follow the tracks of the phse conversion of received signal, effectively suppresses the phase noise in the ofdm system.

The using system information that being illustrated in figure 2 as the invention process provides is eliminated the method flow diagram of phase noise, specifically may further comprise the steps:

Step 101: at transmitting terminal, the signal of input is handled, after chnnel coding handles, carried out sign map and handle through scrambler, generate M data symbol, wherein M is a positive integer, as shown in Figure 3; Obtain frame head, protection interval before described frame head comprises, the PN sequence, the back protection is three parts at interval, as shown in Figure 4; N system information subcarrier inserted in the described data symbol, and wherein N is a positive integer, then it is carried out inverse discrete Fourier transformer inverse-discrete, obtains frame; Frame head and frame multiple connection are become signal frame, will be transmitted into receiving terminal after this signal frame process shape filtering, digital to analog conversion, the radio frequency up-conversion process then.

System information is the information that necessary demodulation sign indicating number is provided for each signal frame, comprises the code check, interlace mode information, frame information pattern of sign map mode, coding etc.As preferably, N＞8.As shown in Figure 5, work as N=36, during M=3744, in 3744 data symbols after 36 system information subcarriers insertion mappings, obtain the sequence of N+M=3780 symbol, be designated as Y[3780], the sequence Y[3780 of output] carry out inverse discrete Fourier transformer inverse-discrete and obtain time-domain signal, with this time-domain signal as frame.

Step 102: at receiving terminal received signal frame, obtain the frame in the signal frame, it is carried out discrete Fourier transform operations recover N system information subcarrier, take out the middle experimental process carrier wave of N system information subcarrier and carry out the phase noise estimation.

Because the pairing analog signal of sampled signal is a band-limited signal, therefore, receiving terminal will obtain the phase noise of entirely accurate, the pairing analog signal of sampled signal that not only requires to carry out the DFT operation is the band-limited signal of a bandwidth less than Nyquist (Nyquist) limit, and require the frequency spectrum of this analog signal to disperse, promptly can only comprise the component of subcarrier integer multiple frequency.And in the reality, though the phase noise bandwidth of oscillator can guarantee that within the specific limits its power spectrum does not generally disperse, but a continuous distribution.Therefore, by choosing the bandwidth B w of suitable phase noise, reduce the error that phase noise is estimated.Bw is big more, and the useful component that obtains is big more, but it is limited protecting bandwidth, and the subcarrier at edge is disturbed by data subcarrier and become unreliable; Owing to be subjected to the influence of factors such as interference of noise and channel estimating be undesirable, Bw is big more in addition, and the interference that is subjected to is also just big more.Generally speaking, Bw selects the length of 1～5 subcarrier.System information both sides each Bw subcarrier is as the frequency protection band of mid portion, and middle (N-2Bw) individual system information subcarrier promptly takes out middle (N-2Bw) individual system information subcarrier and carries out the phase noise estimation as the pilot tone of phase estimation.

The frame signal that receiving end enters before the DFT can be expressed as:

y(n)＝(h(n)*x(n))e ^J

⁽ⁿ⁾+w(n)

Wherein, " * " represents linear convolution,

(n) be illustrated in the phase noise of n sample point, its bandwidth is a Bw subcarrier spacing.Thereby the output that receiving end is carried out after the discrete Fourier transform (DFT) (DFT) can be expressed as Y (k):

$Y\left(K\right)=\mathrm{DFT}\left(r\left(n\right)\right)$

$=X\left(k\right)H\left(k\right)Q\left(0\right)+\underset{r\≠k}{\overset{k+\mathrm{Bw}}{\underset{r=k-\mathrm{Bw}}{\mathrm{\Σ}}}}X\left(r\right)H\left(r\right)Q(k-r)+W\left(k\right)$

Wherein Y (k), X (k), H (k), Q (k), W (k) are respectively y (n), x (n), h (n), e ^J

⁽ⁿ⁾, w (n) discrete Fourier transform (DFT).

Step 103: utilize receiving system information subcarrier to be the convolution relation of phase noise frequency domain response, use least square method (LS) to try to achieve the frequency domain response of phase noise with the insertion system information of making a start.

For the system information T (p) among the X (k) (wherein ,-1890＜k≤1890) (wherein-N/2＜p≤N/2) is V (p) through signal behind the channel, and V (p)=H (p) T (p).Then have:

VQ＝Y

Wherein:

$Q=\left(\begin{array}{c}Q\left(\mathrm{Bw}\right)\\ Q\left(\mathrm{Bw}\right)\\ \·\\ \·\\ \·\\ Q\left(0\right)\\ \·\\ \·\\ \·\\ Q(-\mathrm{Bw}+1)\\ Q(-\mathrm{Bw})\end{array}\right),$

$Y=\left(\begin{array}{c}Y(-N/2+1)\\ Y(-N/2+2)\\ \·\\ \·\\ \·\\ Y\left(0\right)\\ \·\\ \·\\ \·\\ Y(N/2-1)\\ Y(N/2)\end{array}\right)$

VQ=Y is an overdetermined equation, and promptly the equation number is more than the unknown number number.Adopt the least square rule to find the solution this problem, problem is expressed as:

$Q_{\mathrm{LS}}={\underset{Q}{\arg \min }\left|\right|Y-\mathrm{VQ}\left|\right|}^2$

$={\left(V^{H}V\right)}^{-1}{V}^{H}Y$

Under the multipath channel of radio communication, produce frequency selective fading, the different sub carrier gain is different.Because interference of noise, the signal to noise ratio of different sub carrier will be different, therefore the confidence level of different sub carrier is also inequality, adopt weighted least-squares method (WLS) further to reduce error, the weight that subcarrier gains is big is big, the weight that subcarrier gains is little is little, and weighting matrix is defined as a diagonal matrix:

W＝diag(|H(-N/2+Bw+1)|，|H(-N/2+Bw+2)|，…，|H(N/2-Bw)|)

Then weighted least-squares separate into:

Q _WLS＝(V ^HWV) ^-1V ^HWY

Total least square method is at V ^HThe least-squares estimation that V does not adopt the singular value decomposition technology to carry out under the full rank situation, list of references [V.Huffel, J.Vandewalle, The Total Least Squares Problems:Computation Aspects andAnalysis, SIAM, 1991].

Step 104: the response of described phase noise is estimated to carry out discrete Fourier transform (DFT), tries to achieve the estimated value of phase noise, time domain to received signal frame carry out phasing, the signal frame that is compensated, signal frame to described compensation is separated mapping, and decoding is handled, then output.

After estimating the Q value, carry out inverse fourier transform and obtain q (n), 0≤n＜3780, then the phase place of q (n) is exactly the phase noise of estimating, compensates for the input signal time domain, and compensation back signal is:

r′(n)＝r(n)·e ^{-J·angle(q(n))}。

By the signal frame r ' that obtains after the compensation (n), replace not through the signal frame r of overcompensation (n), thereby suppressed the influence of phase noise to signal, to the signal frame r ' after the compensation (n), separate mapping again, processing such as decoding are exported then.

Comprehensive step 102,103,104, Fig. 6 has shown the block diagram of the embodiment of the invention at receiving terminal elimination phase noise.

More obvious for the beneficial effect that makes the embodiment of the invention, below the method based on system information elimination phase noise that the embodiment of the invention proposed has been carried out Computer Simulation, can from emulated data, contrast the effect of the embodiment of the invention;

Main simulation parameter is as shown in table 1.

The main simulation parameter of table 1

Symbol rate	7.56M symbol/second
		The subcarrier planisphere	64QAM
The OFDM sub-carrier number	3780
		Subcarrier spacing	2KHz
Training sequence length	420
		System information length	36

The production method of phase noise adopts the phase noise emulation mode of IEEE 802.11g standard recommendation, and its step is as follows:

(1) produces independent identically distributed gaussian random sequence;

(2) Butterworth low pass filter of design, its three dB bandwidth is Bw _3dB

(3) random sequence that produces in (1) is carried out filtering with the filter that designs in (2), output is desired phase noise, and its variance is σ ²

Simulation parameter is set to Bw _3dB=1kHz, σ ²=3 °.

Connecing receiving terminal under Gaussian channel adopts the phase noise that different estimated bandwidths obtains to estimate as shown in Figure 7 respectively, connecing receiving terminal adopts error sign ratio performance curve that different estimated bandwidths obtains as shown in Figure 8 respectively, as can be seen from the figure error sign ratio is minimum during Bw=3, compare with no phase noise compensation, have the signal to noise ratio (snr) about 2.5dB to improve.

Multipath channel models adopts the channel model shown in the table 2, this multipath channel is Chinese pbs film TV management (the StateAdministration of Radio Film and Television of general bureau, SARFT) test channel that proposes in the Digital Television test report, its amplitude-frequency characteristic as shown in Figure 9.Select w=3 subcarrier spacing of phase noise bandwidth B, adopt least square method and weighted least-squares method to obtain its error sign ratio performance curve as shown in figure 10 respectively, as can be seen from the figure least square method and weighted least-squares method performance are approaching, compare with no phase noise compensation, the SNR that has an appointment about 1.5dB improves.

The impulse response of table 2 simulated channel

Tap	Postpone (μ s)	Power (dB)
			1	0	0
2	0.2	-10
			3	1.9	-14
4	3.9	-18
			5	8.2	-20
6	15	-20

Above specific embodiments of the invention are had been described in detail, but the present invention is not restricted to the foregoing description.Within the spirit and principle of claim of the present invention, any modification of being done, be equal to and replace and improvement etc., all should be included within protection scope of the present invention.

Claims (4)

1. utilize system information to suppress the method for phase noise in an ofdm system, it is characterized in that described method comprises:

Step 101: at transmitting terminal, the signal of input is handled, after chnnel coding handles, carried out sign map and handle through scrambler, generate data symbol; Obtain frame head; N system information subcarrier inserted in the described data symbol, and wherein N is a positive integer; Then the data symbol that has inserted N system information subcarrier is carried out inverse discrete Fourier transformer inverse-discrete, obtain frame; Frame head and frame multiple connection are become signal frame, will be transmitted into receiving terminal after this signal frame process shaping filter, digital to analog conversion, the radio frequency up-conversion process then;

Step 102: at receiving terminal received signal frame, obtain the frame in the signal frame, it is carried out discrete Fourier transform operations recover N system information subcarrier, take out the middle experimental process carrier wave of N system information subcarrier and carry out the phase noise estimation;

Step 103: utilize the system information subcarrier that receives to be the convolution relation of phase noise frequency domain response, use least square method to try to achieve the frequency domain response of phase noise with the insertion system information of making a start;

Step 104: the frequency domain response to described phase noise carries out discrete Fourier transform (DFT), tries to achieve the estimated value of phase noise, time domain to received signal frame carry out phasing, the signal frame that is compensated, signal frame to described compensation is separated mapping, and decoding is handled, then output.

2. utilize system information to suppress the method for phase noise in the ofdm system according to claim 1, it is characterized in that the number N of described system information subcarrier is greater than 8.

3. utilize system information to suppress the method for phase noise in the ofdm system according to claim 1, it is characterized in that, in described step 102, (N-2Bw) individual subcarrier carries out the phase noise estimation in the middle of taking out N system information subcarrier, wherein, Bw is the phase noise estimated bandwidth.

4. utilize system information to suppress the method for phase noise in the ofdm system according to claim 1, it is characterized in that least square method comprises weighted least-squares method described in the described step 103, or total least square method.

Images