CN101202721A - Method of Wiener-Kolmogorov model interpolation based on time domain signal-to-noise ratio - Google Patents

Abstract

A wiener interpolation method based on time domain SNR estimation includes the steps of acquiring a channel frequency response of a pilot subcarrier; utilizing two cascade interpolation filters to acquire the channel frequency response of a data subcarrier; utilizing a stored interpolation result to carry through frequency domain equalization of single tap on the data received on the data subcarrier. The performance of the wiener interpolation method estimated by the real time flatness SNR of the invention is very close to the performance under a situation estimated by an ideal SNR and is extremely better than the system of the method by adopting line interpolation. Furthermore, by adopting a method for storing the parameters of an interpolation filter in advance, the matrix inversion with large calculating amount can be avoided, thereby greatly simplifying the realizing complexity of a receiver.

Description

Receive interpolation method based on the dimension of time domain signal-to-noise ratio (SNR) estimation

Technical field

The present invention relates to broadcast system, particularly receive interpolation method based on the dimension of time domain signal-to-noise ratio (SNR) estimation.

Background technology

DRM is unique universal non-patent digital radio broadcasting system of shortwave, medium wave and long wave amplitude modulation broadcasting frequency range.Under same coverage condition, the DRM transmitter power is than the low 6-9dB of traditional analog transmissions acc power, and digital broadcasting is lower than the same adjacent frequency protective rate of analog broadcasting, and the anti-multipath interference performance is strong, is convenient to move receive; Tonequality can reach the quality of CD or FM multiplex; Additional data and multimedia messages can be provided; Compare with DAB, its receiver price is easier to be accepted by the mass audience.Its appearance is the sign that the following frequency range broadcasting of 30MHz is revived, and has become international standard at present.

In the coherent demodulation ofdm system, in order to carry out equilibrium to the received signal, receiver must obtain the amplitude and the phase information of channel by channel estimating.But, broadcast channel not only suffers because the frequency selective fading that multipath transmisstion causes, and the time selective fading that brought by Doppler frequency shift or Doppler's expansion, for the quality of reception that guarantees receiver and the real-time of reception, require receiver that broadcast channel is carried out promptly and accurately channel estimating.According to the requirement of DRM standard, transmitter also sends pilot data simultaneously when sending useful data, and we just can adopt the channel estimation scheme based on pilot tone like this.At first, we extract the received signal of pilot frequency locations, utilize the local pilot tone of receiver storage, calculate the channel frequency response of pilot frequency locations with least-squares algorithm, then, we estimate the channel frequency response at data subcarrier place with interpolation filter, and are last, and we carry out equilibrium with the frequency-domain equalizer of single tap to the reception receipt.When sending signal employing high order modulation,,, need channel estimating more accurately in order to obtain better receiver performance such as 16QAM or 64QAM.

Though the proposition of this scheme based on DRM standard ETSI ES 201 980.V2.1.1, is equally applicable to other based on channel estimating pilot tone, the coherent demodulation orthogonal FDM communication system.

Because under two fading channel conditions, just existing by many frequency selective fadings that cause through propagation, have again because under the channel condition of the time selective fading that Doppler frequency shift or Doppler's expansion cause, the pilot frequency design of rhombus has better anti-fading characteristic than block pilot frequency design or dressing pilot frequency design, just adopted the rhombus pilot frequency design of this time-frequency two-dimensional in the DRM standard, this scheme can reduce the degree that is subjected to receiver decreased performance under the situation about having a strong impact on that channel causes in some pilot tone.Rhombus pilot frequency design distribution map as shown in Figure 1.

At different channel conditions, comprised four kinds of different robust modes in the DRM standard, concrete description sees Table 1:

Table 1: robust mode and corresponding channel condition

Robust mode	Channel condition
		Mode A	Gaussian channel, the slight rate channel that falls is applicable to the medium wave and the long wave channel on daytime.
Mode B	Time and frequency-selective channel are than the shortwave and the medium wave channel at night of long delay expansion.
		Pattern C	Time and frequency-selective channel, channel condition is relatively poor, the short wave channel of bigger Doppler's expansion.
Pattern D	Unusual Jian Zhuan pattern, but because pilot interval has too closely influenced message transmission rate.

Different robust modes all has different pilot intervals on time orientation and frequency direction, concrete at interval big or small as shown in table 2:

Table 2: pilot interval size

Robust mode	N T	N F
			A	5	20
B	3	6
			C	2	4
D	3	3

In table 2, N _TAnd N _FPilot interval on the difference express time direction and the pilot interval on the frequency direction.

First three plants the application that robust mode can satisfy most of DRM broadcasting, for Mode A, because short protection at interval and narrower make it not be suitable for shortwave broadcasting at subcarrier spacing.Have only pattern D to be applicable to that channel model 6 in the standard, this channel model not only have very long time delay expansion, also have very big Doppler's expansion, it is a kind of approximate simulation to the sky wave propagation of region of the equator.

As everyone knows, optimum channel estimator under the minimum mean square error criterion is two-dimentional Weiner filter, but two-dimentional Weiner filter is not easy to realize in practical engineering application very much, but when channel is the steady irrelevant scatter channel of broad sense, the one-dimensional filtering device of two cascades is a kind of good selection schemes, frequency direction interpolation after can first time orientation interpolation, also can first frequency direction interpolation after the time orientation interpolation.We have provided the block diagram (frequency direction behind the first time orientation) of the one dimension interpolation filter of two cascades in Fig. 2.

For the ofdm system of assisting based on pilot tone, common channel estimation methods is, at first try to achieve the channel frequency response of pilot frequency locations, adopt two cascade one-dimensional filtering devices that the channel frequency response of data position is estimated then, for example: at first carry out simple linear interpolation at time orientation, and then in the enterprising line linearity interpolation of frequency direction.

Although it is optimal interpolation device under the minimum mean square error criterion that dimension is received interpolation device, it still has following weak point: the statistical property of transmission channel, the accuracy that signal noise ratio is estimated, and the dimension that is used for interpolation receives the tap number of interpolation device, all can influence the performance of receiver; Secondly, because the tap coefficient of real-time renewal interpolation device need be used the sizable matrix inversion operation of amount of calculation, this also can increase the complexity that receiver is realized greatly.

Summary of the invention

The purpose of this invention is to provide a kind of dimension and receive interpolation method based on the time domain signal-to-noise ratio (SNR) estimation.

For achieving the above object, a kind of dimension based on the time domain signal-to-noise ratio (SNR) estimation is received interpolation method, comprises step:

Obtain the channel frequency response at pilot sub-carrier place;

Utilize the interpolation filter of two cascades to obtain the channel frequency response at data subcarrier place;

The interpolation result of utilization storage carries out single tap to the data on the data subcarrier that receives frequency domain equalization.

The performance that adopts the dimension of level and smooth signal-to-noise ratio (SNR) estimation in real time of the present invention to receive under the situation of the performance of interpolation method and desirable signal-to-noise ratio (SNR) estimation is very approaching, be better than the systematic function that adopts linear interpolation method far away, and owing to adopted the method for storing the interpolation filter coefficient in advance, can avoid the very big matrix inversion operation of operand, thereby simplify the implementation complexity of receiver greatly.

Description of drawings

Fig. 1 is a rhombus pilot frequency design distribution map;

Fig. 2 is that two cascades are with dimension interpolation filter block diagram;

Fig. 3 is a DRM receiver simplified block diagram;

Fig. 4 is a pre-set factory bank of filters scheme;

Fig. 5 is the bit error rate characteristic of channel 3;

Fig. 6 is the bit error rate characteristic of channel 4.

Embodiment

Here we come our channel estimation method is described with the DRM system as an example, and the simplified block diagram of receiver as shown in Figure 3.

The present invention mainly introduces the content of channel estimating, exactly the part that marks with chain-dotted line in Fig. 3.

Optimum channel estimator under the minimum mean square error criterion is two-dimentional Weiner filter, but two-dimentional Weiner filter is not easy to realize in practical engineering application very much, and when channel was the steady irrelevant scatter channel of broad sense, the one dimension interpolation device of two cascades was a kind of good selection schemes.And the interpolation device of these two cascades can adopt different interpolation devices respectively, such as linear interpolation device, cubic spline interpolation device, receive interpolation device based on the interpolation device of DFT and dimension.The order of interpolation also has two kinds: can be earlier at the enterprising row interpolation of time orientation, and then on frequency direction interpolation; Also interpolation on frequency direction earlier, and then on time orientation interpolation.

Here we mainly discuss on frequency direction and to adopt dimension to receive interpolation device to carry out the method for channel estimating, ask dimension to receive need be in the interpolation device tap coefficient to the estimated value of signal to noise ratio, we have adopted a kind of method of new time domain signal-to-noise ratio (SNR) estimation, and will adopt the bit error rate of the receiver of this method to compare with the receiver bit error rate under desirable signal-to-noise ratio (SNR) estimation (receiver is known the size of signal power and the noise power accurately) situation, simulation result shows, this method is estimated snr exactly, and the performance of system is very near the systematic function under the desirable state of signal-to-noise.

And, consider that traditional dimension receives interpolation device and need the tap coefficient of real-time update interpolation device, this needs the real-time matrix inversion operation of carrying out, it is very complicated to cause hardware to be realized, we have proposed a kind of scheme, the tap coefficient of the some groups of interpolation devices that calculated in advance is gone out stores in advance, the signal-noise ratio estimation method that proposes according to us is estimated signal to noise ratio then, judge to adopt which group interpolation device coefficient to carry out interpolation operation according to the snr value of estimating, thereby avoided the huge matrix inversion operation of operand, reduced the complexity that receiver is realized greatly.

1) calculating of interpolation filter coefficient

In the middle of channel estimating, for the performance that obtains at receiving terminal, can be according to time orientation of estimating or the channel correlation function on the frequency direction, estimated value in conjunction with the signal to noise ratio that adopts distinct methods to obtain, real-time update dimension is received the tap coefficient of interpolation device, then according to these coefficients, and the channel frequency response of the pilot sub-carrier position that obtains in advance, obtain the channel frequency response of data subcarrier position by interpolation, can be described below with formula.

$\hat{H}=R_{H\hat{R}}{R}_{\hat{P}\hat{P}}^{-1}\hat{P}=W_{\mathrm{opt}}\hat{P}$

$R_{\hat{P}\hat{P}}=R_{\mathrm{PP}}+\frac{1}{\mathrm{SNR}}$

In the following formula,

The channel frequency response of expression data subcarrier position and the covariance matrix of the channel frequency response of the pilot sub-carrier position that has noise that estimates;

It is the auto-covariance matrix of the channel frequency response at the pilot frequency locations place that has noise that estimates; $W_{\mathrm{opt}}=R_{H\hat{P}}{R}_{\hat{P}\hat{P}}^{-1},$ It is by the best wiener filter coefficients vector that calculates, and can obtain the channel frequency response of data subcarrier position by interpolation according to it.Owing to there is the computing of matrix inversion, this method needs very big operand.

Because we have adopted the one dimension interpolation device of two cascades to carry out channel estimating, and we are at first in the enterprising line linearity interpolation of time orientation, on frequency direction, tie up then and receive interpolation, so here we only need auto-covariance and cross covariance matrix on the calculated rate direction.

Receive interpolation if on time orientation, tie up, just need auto-covariance and cross covariance matrix on direction computing time.

Consider that channel model does not match and can cause loss on the system letter energy, we need consider more abominable transmission channel conditions, normally equally distributed retarding power spectrum, and the supposition maximum delay expands to τ _MaxSituation.Under this channel condition, in order to determine the tap coefficient of Weiner filter, we only need two parameters, are exactly the maximum delay extended by tau of channel _MaxThe signal of the maximum of hoping the run time of with system and the ratio of noise power.

The retarding power spectrum of supposing channel is obeyed evenly and is distributed, and its extension length equals the protection length at interval of the ofdm system under this pattern, and we can obtain the correlation function on the frequency direction under this condition:

$r\left(\mathrm{\Δk}\right)=\sin c\left(\mathrm{\Δk}\frac{G}{N}\right)$

In following formula, Δ is represented subcarrier spacing, and k represents the position of subcarrier, and G represents to protect length at interval, and N represents the length of OFDM symbol useful part.

2) signal-to-noise ratio (SNR) estimation

Estimated snr whether real-time and accurately, to the receiver performance important influence, below we introduce a kind of method of time domain signal-to-noise ratio (SNR) estimation.

At first the channel frequency response to the pilot sub-carrier position carries out inverse Fourier transform, obtains the estimated value of time domain channel impulse response, distinguishes signal calculated energy and noise energy then, thereby obtains the estimated value of current sign signal to noise ratio.

$h_{n,l}=\frac{1}{M}\underset{k=0}{\overset{M-1}{\mathrm{\&Sigma;}}}{\hat{H}}_{\mathrm{ls}}^{k,\mathrm{<figure-callout\; id="2"\; label="l\; e\; j"\; filenames="A20061016588900161.png,A20061016588900162.png"\; state="\{\{state\}\}">l</figure-callout>}}{e}^j{2\mathrm{\&pi;kn}/M}^{},0\&le;n\&le;M-1$

Following formula is represented IFFT, h _{N, l}Be the estimated value of channel impulse response,

The channel frequency response of expression pilot sub-carrier position.

Generally; the length of channel impulse response is less than protection length at interval; channel impulse response sample value energy in the protection at interval and the energy that can be considered as useful signal; and the sample value energy of the channel impulse response beyond the protection at interval and can be considered as the energy of noise and interference, so can be by following formula estimated snr:

$p_l=\frac{1}{G}\underset{n=0}{\overset{G-1}{\mathrm{\&Sigma;}}}{\left|h_{n,l}\right|}^2$

${\mathrm{\&sigma;}}_l=\frac{1}{M-G+1}\underset{n=G}{\overset{M}{\mathrm{\&Sigma;}}}{\left|h_{n,l}\right|}^2$

${\mathrm{SNR}}_l=10\lg \left(\frac{p_l}{{\mathrm{\&sigma;}}_l}\right)$

In the following formula, p _lThe useful signal energy that expression estimates, σ _lThe noise that expression estimates and the energy of interference, l is-symbol index, G are represented to protect length at interval, and M is the number of pilot sub-carrier.

Under the fast fading channel condition, the estimated value of the signal to noise ratio that obtains with said method can be affected, thereby further influence is to the accuracy of interpolation filter coefficient estimation.In order to overcome this problem, we can overcome this influence with iir filter.

${\tilde{p}}_l={\mathrm{\&lambda;}}_p{\tilde{p}}_{l-1}+(1-{\mathrm{\&lambda;}}_p)p_l$

${\widetilde{\mathrm{\&sigma;}}}_l={\mathrm{\&lambda;}}_{\mathrm{\&sigma;}}{\widetilde{\mathrm{\&sigma;}}}_{l-1}+(1-{\mathrm{\&lambda;}}_{\mathrm{\&sigma;}}){\mathrm{\&sigma;}}_l$

${{\mathrm{SNR}}_l}^{\&prime;}=10\lg \left(\frac{{\tilde{p}}_l}{{\widetilde{\mathrm{\&sigma;}}}_l}\right)$

${\mathrm{SNR}}^{\&prime;\&prime;}=\frac{1}{K}\underset{k=0}{\overset{K-1}{\mathrm{\&Sigma;}}}{{\mathrm{SNR}}_k}^{\&prime;}$

Because the energy of noise is along with the time slowly changes, so we select λ _σ=0.5, for the variation of the signal energy under the real-time tracking fast fading channel, we select λ _p=0.2.

3) pre-set factory filter

For fear of the very big matrix inversion operation of operand, to reduce the complexity that receiver is realized, we adopt the scheme of packing coefficient bank of filters in advance.The tap coefficient of the some groups of interpolation devices that this scheme will be in advance calculates according to different snr values is stored in receiving terminal in advance, in the middle of the actual reception process, according to 2) in method calculate signal to noise ratio, select corresponding Weiner filter to finish interpolation operation according to the snr value of estimating then, thereby avoided the huge matrix inversion operation of operand, reduced the complexity that receiver is realized.

Below we are described with our simulation example, its block diagram is as shown in Figure 4.

In Fig. 4, interpolation filter Filter[12dB], Filter[18dB], Filter[23dB] have different tap coefficients, they are to be 12dB in signal to noise ratio, 18dB, calculate respectively under the situation of 23dB, these coefficients only need once to calculate, and are kept in the receiver then.In the middle of the actual reception process, by introduce above in real time level and smooth after signal-noise ratio estimation method, the signal to noise ratio that receiver can obtain estimating.When the signal to noise ratio that estimates during less than 15dB, receiver adopts interpolation device Filter[12dB] finish estimation to the channel frequency response at data subcarrier place; When the signal to noise ratio that estimates during greater than 20dB, receiver adopts interpolation device Filter[23dB] finish estimation to the channel frequency response at data subcarrier place; When the signal to noise ratio that estimates more than or equal to 15dB during smaller or equal to 20dB, receiver adopts interpolation device Filter[18dB] finish estimation to the channel frequency response at data subcarrier place.

In the middle of real system, as the case may be, can select the filter of different numbers, the number of filter is many more, needs the tap coefficient of storage many more, and the hardware resource that needs is also just many more, but channel estimation value is more accurate, and the performance of receiver is better.When reality realizes, the consideration of between resource consumption and systematic function, compromising according to actual needs.

By this method, receiver only needs to judge according to the snr value that estimation obtains, and selects to be applicable to the tap coefficients value of the interpolation filter under this signal to noise ratio, carries out interpolation arithmetic then, thereby can effectively avoid matrix inversion operation, simplify the implementation complexity of receiver greatly.

Embodiment

Actual DRM broadcast channel is the frequency selectivity that existing multipath causes, two fading channels of the time selectivity that Doppler expands or frequency displacement causes are arranged again.Wherein multipath transmisstion is mainly because the ionospheric reflection of differing heights causes that the expansion of the maximum delay of channel can reach several milliseconds, and Doppler's expansion and frequency displacement cause mainly due to the spectral characteristic of ionospheric reflection and receiver mobile.With the mid latitudes is example, and the maximum of time delay expansion can reach 6ms, and Doppler's expansion then can be up to 5Hz.Generally, the representative value that time delay expansion and Doppler expand is 2ms and 1Hz, and this is the parameter value of our channel model 4 used just.

In line with the principle of proceeding from the reality, we have investigated the situation of channel model 3 and channel model 4 in the DRM standard, and wherein channel model 3 is aimed at the USConsortium model of intermediate frequency and high frequency, and channel model 4 is aimed at the standard CC IR model of high frequency.The simulation parameter that we use provides in table 3, and the concrete parameter of channel model 3 and channel model 4 provides in table 3 and table 4 respectively:

Table 3: simulation parameter

Robustness Mode	B
		Band Width	10K
Coding Scheme	64QAM
		Code Rate	0.6
Channel Model	3and			4

Table 4: channel 3 parameters

	Path 1	Path 2	Path 3	Path4
					Delay	0	0.7ms	1.5ms	2.2ms
Path Gain	1	0.7	0.5	0.25
					Doppler Shift	0.1Hz	0.2Hz	0.5Hz	1.0Hz
Doppler Spread	0.1Hz	0.5Hz	1.0Hz	2.0Hz

Table 5: channel 4 parameters

	Path 1	Path 2
			Delay	0	2ms
Path Gain	1	1
			Doppler Shift	0	0
Doppler Spread	1Hz	1Hz

Simulation result to channel model 3 and channel model 4 is distinguished as shown in Figure 5 and Figure 6.From simulation result as can be seen, the performance that adopts this dimension of level and smooth signal-to-noise ratio (SNR) estimation in real time to receive under the situation of the performance of interpolation method and desirable signal-to-noise ratio (SNR) estimation (receiving terminal deterministic signal power and noise power) is very approaching, be better than the systematic function that adopts linear interpolation method far away, and owing to adopted the method for storing the interpolation filter coefficient in advance, can avoid the very big matrix inversion operation of operand, thereby simplify the implementation complexity of receiver greatly.

Claims (10)

1. the dimension based on the time domain signal-to-noise ratio (SNR) estimation is received interpolation method, comprises step:

Obtain the channel frequency response at pilot sub-carrier place;

Utilize the interpolation filter of two cascades to obtain the channel frequency response at data subcarrier place;

The interpolation result of utilization storage carries out single tap to the data on the data subcarrier that receives frequency domain equalization.

2. method according to claim 1, the interpolation filter that it is characterized in that described two cascades be respectively on the time orientation with frequency direction on interpolation filter.

3. method according to claim 1 is characterized in that the interpolation order of described two interpolation filters is carried out earlier on time orientation, carry out on frequency direction then.

4. method according to claim 1 is characterized in that the interpolation order of described two interpolation filters is carried out earlier on frequency direction, carry out at time orientation then.

5. method according to claim 1 is characterized in that described interpolation filter is an one of the following: linear interpolation device, cubic spline interpolation device, receive interpolation device based on the interpolation device of DFT and dimension.

6. method according to claim 5 is characterized in that if use dimension to receive interpolation device, then before carrying out interpolation operation, must obtain to tie up and receive the tap coefficient of interpolation device.

7. method according to claim 6 is characterized in that obtaining to tie up that to receive the tap coefficient of interpolation device be according to the signal to noise ratio that estimates its coefficient to be carried out real-time calculating to upgrade.

8. method according to claim 6, it is characterized in that obtaining to tie up that to receive the tap coefficient of interpolation device be according to certain channel model, calculate in advance dimension under the different signal to noise ratios receive interpolation device tap coefficient and be stored in receiving terminal, according to estimating that the signal to noise ratio that obtains just can choose the interpolation device of corresponding coefficient and finish interpolation.

9. method according to claim 8 is characterized in that described estimated snr comprises step:

Channel frequency response sequence to the pilot sub-carrier place is carried out inverse Fourier transform, obtains the estimated sequence of time domain channel shock response;

Distinguish signal calculated energy and noise energy, obtain the estimated value of current sign signal to noise ratio.

10. method according to claim 9 is characterized in that adopting different iir filters to carry out filtering to estimating the signal power and the noise power that obtain.

Images