CN102130864A - Channel estimation method and device - Google Patents

Abstract

The invention discloses a channel estimation method and a channel estimation device. The method comprises the following steps of: performing channel estimation at a pilot frequency position by combining the latest frequency offset estimated value of a target user; performing single pilot frequency-based frequency offset estimation by combining the channel estimation at the pilot frequency position; performing channel estimation at a data position by combining a single pilot frequency-based frequency offset estimated result; and directly performing frequency offset compensation on the estimated data by combining frequency offset estimation. The method and the device can effectively reduce the channel estimation complexity, obviously improve the channel estimation accuracy and obviously improve the system link performance under a high-speed moving scene.

Description

A kind of channel estimation methods and device

Technical field

The present invention relates to the communications field, be specifically related to a kind of channel estimation methods and device.

Background technology

At present, (Long Term Evolution, LTE) project is subjected to extensive concern as the Long Term Evolution of a kind of 3G to the Long Term Evolution of 3GPP tissue.It is in the up employing single-carrier frequency division multiple access of wireless access side physical layer (Single Carrier-Frequency Division Multiplex Access, SC-FDMA) technology.

In the sender physical layer, design the pilot frequency sequence that the destination is known, this sequence and unknown message sequence are in time-multiplexed mode, be modulated at different OFDMs (Orthogonal FrequencyDivision Multiplex respectively, OFDM) on the symbol, its frame structure as shown in Figure 1.Receiver wants demodulation to obtain the unknown message sequence, just need carry out channel estimating; That is: according to the channel response of pilot frequency information estimating pilot frequency position, estimate the channel response at unknown message sequence location place more on this basis.Just can obtain the unknown message sequence thus by follow-up processing.

Channel estimating is a kind of important link of receiver performance, especially under high-speed mobile condition, if between the sending and receiving system not exclusively synchronously and exist when bigger, frequency deviation, then channel estimating performance will more seriously influence the reception handling property.

Existingly be applicable to that LTE system uplink channel estimation approach is few, when all not considering system in only method partially and frequency deviation to the influence of channel estimating, therefore can't guarantee the accuracy of channel estimating; And computational complexity is all bigger, therefore difficult the realization.

Summary of the invention

In view of this, main purpose of the present invention is to provide a kind of channel estimation methods and device, reduces the channel estimating complexity, improves channel estimation accuracy.

For achieving the above object, technical scheme of the present invention is achieved in that

A kind of channel estimation methods, this method comprises:

The content that will be applied to channel estimating is carried out frequency offset estimating based on single pilot tone; Carry out channel estimating in conjunction with frequency offset estimation result based on single pilot tone.

Carry out comprising based on the process of the described frequency offset estimating of single pilot tone:

Carrying out system frequency deviation at the pilot frequency locations place based on single frequency pilot sign estimates, and will be at estimating on two included frequency pilot signs of current subframe that gained system frequency deviation value and the frequency deviation value that calculates gained at a last subframe carry out the recurrence smoothing processing, obtain the frequency offset estimating value of current subframe.

Before carrying out described frequency offset estimating, further the pilot frequency sequence that receives is comprised the processing of least square (LS) estimation, filtering noise reduction, equilibrium.

Further carry out compensate of frequency deviation to carrying out the result that described channel estimating obtains.

The method of described compensate of frequency deviation is:

According to the frequency offset estimating value of the current subframe of gained before, to SC-FDMA modulating data compensating for frequency offset.

The described channel estimating of carrying out is carried out at the Data Position place, and its process comprises:

Eliminate the start-phase deviation of pilot frequency locations channel estimating, use linear interpolation data estimator position channel estimating, the start-phase reduction of Data Position place channel estimating, Data Position place antenna merges and channel equalization, does reverse discrete Fourier transform (DFT) (IDFT) and obtains the SC-FDMA modulation symbol.

A kind of channel estimating apparatus, this device comprise continuous frequency deviation estimating unit, Data Position place channel estimating unit; Wherein,

Described frequency deviation estimating unit is used for the content that will be applied to channel estimating is carried out frequency offset estimating based on single pilot tone;

Described Data Position place channel estimating unit is used for carrying out channel estimating in conjunction with the frequency offset estimation result of described frequency deviation estimating unit.

Further comprise pilot frequency locations place channel estimating unit, be used for the pilot frequency sequence that receives is comprised LS estimation, filtering noise reduction, balanced processing, and the pilot frequency locations place channel estimation results that will handle gained sends to described frequency deviation estimating unit and Data Position place channel estimating unit.

Further comprise balanced unit, frequency offset compensation element; Wherein,

Described balanced unit is used for the reception data at balanced corresponding character position place, obtains the frequency domain data estimate vector of corresponding data position; Again each frequency domain data estimate vector is IDFT respectively and handles, obtain the modulation symbol estimated result, and it is sent into described frequency offset compensation element;

Described frequency offset compensation element is used for according to the frequency offset estimation result of the current subframe of receiving the modulation symbol estimated result of receiving being carried out compensate of frequency deviation, obtains the channel estimation results that needs;

Described frequency deviation estimating unit is further used for the frequency offset estimation result of current subframe is sent to described frequency offset compensation element.

Under the situation that is provided with pilot frequency locations place channel estimating unit, the annexation between described frequency deviation estimating unit and pilot frequency locations place channel estimating unit, the Data Position place channel estimating unit is:

The input of described frequency deviation estimating unit links to each other with the output of pilot frequency locations place channel estimating unit, and the output of frequency deviation estimating unit links to each other with the input of Data Position place channel estimating unit, the input of frequency offset compensation element respectively.

As seen, owing to can the nearest frequency offset estimating value of combining target user be used for pilot frequency locations place channel estimating, and carry out frequency offset estimating based on single pilot tone in conjunction with pilot frequency locations place channel estimating, and, directly data estimator is carried out compensate of frequency deviation in conjunction with frequency offset estimating more at last in conjunction with carrying out Data Position place channel estimating based on the frequency offset estimation result of single pilot tone; Make the inventive method and device can effectively reduce the channel estimating complexity, and significantly improve channel estimation accuracy.

Description of drawings

Fig. 1 is the LTE sub-frame of uplink structure chart of prior art;

Fig. 2 is the channel estimating flow chart of the embodiment of the invention;

Fig. 3 is a channel estimating general flow chart of the present invention;

Fig. 4 is the channel estimating apparatus figure of the embodiment of the invention;

Fig. 5 is the present invention and the link simulation performance schematic diagram of prior art on the EVA300 channel.

Embodiment

On the invention thinking, channel estimation methods provided by the present invention comprises: utilize the last frequency offset estimating value of targeted customer to carry out pilot frequency locations place channel estimating; And carry out frequency offset estimating based on single pilot tone in conjunction with pilot frequency locations place channel estimating; In conjunction with carrying out Data Position place channel estimating based on the frequency offset estimation result of single pilot tone; In conjunction with frequency offset estimation result data estimator is carried out compensate of frequency deviation based on single pilot tone.

Channel estimating apparatus provided by the present invention comprises continuous pilot frequency locations place channel estimating unit, frequency deviation estimating unit, Data Position place channel estimating unit and frequency offset compensation element; Wherein, described pilot frequency locations place channel estimating unit utilizes last estimation gained user's frequency deviation valuation to carry out pilot frequency locations place channel estimating; Described frequency deviation estimating unit is used for carrying out frequency offset estimating based on single pilot tone in conjunction with pilot frequency locations place channel estimating; Described Data Position place channel estimating unit is used for carrying out in conjunction with the frequency offset estimation result of described frequency deviation estimating unit the channel estimating at Data Position place; Described frequency offset compensation element is used in conjunction with the frequency offset estimation result of described frequency deviation estimating unit the data of estimating being carried out compensate of frequency deviation.

Referring to Fig. 2, Fig. 2 is the channel estimating flow chart of the embodiment of the invention, and this flow process may further comprise the steps:

The FFT of step 201:N point time domain sequences.

Particularly, can do the FFT conversion, obtain frequency domain receiving sequence y the N point time domain sequences that base band is received.

Step 202: the frequency domain user data separates.

Particularly, can be the user to frequency domain data and separate, and extracting objects user's reception pilot frequency sequence.As: according to the frequency resource that the frequency division multi-user occupies separately, directly separate these frequency divisions user, and extracting targeted customer's reception pilot tone and data symbol with the time-frequency position of making a start corresponding.

In actual applications, can the hypothetical target user and receiver between not exclusively synchronously, when having system, frequency deviation, satisfy following relation on frequency pilot sign between sending and receiving this moment and the data symbol respectively:

Wherein, matrix F is represented the FFT transformation matrix that N is ordered, DM represent M point discrete Fourier conversion (DFT) transformation matrix (N＞M), P represent and extract the corresponding elementary row transformation matrix of frequency resource of CU, and diag{} represents the diagonal matrix opened by vector { }, and The conjugation symmetry transformation of subscript () H representing matrix ().Y represents targeted customer's signal that the OFDM symbol that receives extracts again after FFT handles, n represents noise or disturb vector, subscript () (l) and () (i) to be used for expression vector () respectively be at pilot tone and Data Position.

After the completing steps 202, just can carry out channel estimating and frequency offset estimating at targeted customer's pilot frequency locations place, concrete operating process is made up of to 205 step 203.

Step 203: improved LS estimates and the filtering noise reduction.

Particularly, can be according to known user's pilot frequency sequence, to use improved LS algorithm to receiving pilot frequency sequence, and adopt simple filtering method, the level and smooth or FIR filtering as fixing point obtains the channel estimating at targeted customer's pilot frequency locations place; Concrete processing method can for:

Utilize two user's pilot frequency sequences of known frequency offsets size earlier

With

Match obtains the frequency deviation size

User's pilot frequency sequence (, then think if current subframe is first subframe

Otherwise be the last frequency deviation value of estimating)

That is:

$x_{\mathrm{\Δ}\hat{f}(n-1)}^{\left(l\right)}=\mathrm{\γ}\·(x_{{\mathrm{\β}}_1}^{\left(l\right)}-x_{{\mathrm{\β}}_2}^{\left(l\right)})\·(\mathrm{\Δ}\hat{f}(n-1)-{\mathrm{\β}}_1)+x_{{\mathrm{\β}}_1}^{\left(l\right)}---\left(3\right)$

Wherein, γ can be definite by numerical simulation, normally a definite constant value.

Afterwards, utilize the LS algorithm to carry out channel estimating:

${\hat{h}}_{\mathrm{ls}}^{\left(l\right)}=\mathrm{diag}{\left\{\left(x_{\mathrm{\Δ}\hat{f}(n-1)}^{\left(l\right)}\right)\right\}}^{-1}\·y^{\left(l\right)}---\left(4\right)$

Again estimated result is carried out the filtering noise reduction of frequency domain sequence.Exist when overcoming system the filtering performance that causes to worsen partially, inclined to one side when needing at first estimating system usually, and filter tap coefficients adjusted, i.e. the skew of each subcarrier that at first causes partially during estimating system:

${\mathrm{\θ}}^{\left(l\right)}=\frac{\mathrm{angle}\left(\underset{m=1}{\overset{M-\mathrm{\δ}}{\mathrm{\Σ}}}\right[{\hat{h}}_{\mathrm{ls}}^{\left(l\right)}\left(m\right)\·\mathrm{conj}\left({\hat{h}}_{\mathrm{ls}}^{\left(l\right)}(m+\mathrm{\δ})\right)\left]\right)}{\mathrm{\δ}}---\left(5\right)$

Then, utilize the subcarrier skew again, adjust the FIR filter (in actual applications, can be not limited to FIR filtering, and use other filtering mode commonly used) on p rank, establish its tap coefficient vector

Then filter is adjusted into:

Wherein, the Toeplitz matrix opened by vector () of Toep () expression.

At last, utilize adjusted filter right

Carry out filtering, obtain pilot frequency locations place channel estimating

That is:

${\hat{h}}^{\left(l\right)}=T*{\hat{h}}_{\mathrm{ls}}^{\left(l\right)}---\left(7\right)$

Step 204: pilot frequency locations place channel equalization.

Particularly, can be according to the channel estimating at pilot frequency locations place

Do channel equalization to receiving pilot frequency sequence; Concrete processing method can for:

Utilize the channel estimating at pilot frequency locations place

To receiving pilot data y ^(l)Carry out equilibrium, comparatively simple balanced implementation algorithm as the formula (8):

${\hat{y}}^{\left(l\right)}={\left(\mathrm{diag}\right\{{\hat{h}}^{\left(l\right)}\left\}\right)}^{-1}\·y^{\left(l\right)}---\left(8\right)$

Step 205: based on the frequency offset estimating of single pilot tone.

Particularly, can be according to the data of being exported after the channel equalization, carrying out system frequency deviation at the pilot frequency locations place based on single frequency pilot sign estimates, and will be at estimating on two included frequency pilot signs of current subframe that gained system frequency deviation value and the frequency deviation value that calculates gained at a last subframe carry out the recurrence smoothing processing, obtain the frequency offset estimating value of current subframe.Concrete processing method can for:

At first, generate two reference signals that are used for frequency offset estimating, they are respectively the pilot frequency sequences (normalization frequency deviation size is respectively ± β (0＜β≤0.5)) of two known dimensions frequency deviations, are designated as vectorial r _{_ β} ^(l)And r _β ^(l), promptly produce to be calibrated in the following way and the reference signal:

$r_c^{\left(l\right)}=\mathrm{diag}\left\{{\left(x^{\left(l\right)}\right)}^{*}\right\}\·{\hat{y}}^{\left(l\right)}$

$r_{-\mathrm{\β}}^{\left(l\right)}=\mathrm{diag}\left\{{[B(-\mathrm{\β})\·x^{\left(l\right)}]}^{*}\right\}\·{\hat{y}}^{\left(l\right)}---\left(9\right)$

$r_{\mathrm{\β}}^{\left(l\right)}=\mathrm{diag}\left\{{[B\left(\mathrm{\β}\right)\·x^{\left(l\right)}]}^{*}\right\}\·{\hat{y}}^{\left(l\right)}$

Secondly, all carry out the same grouping to be calibrated with the reference signal, suppose to be divided into the Q group, then k element of their q group is designated as r respectively _c ^{(l, q)}(k), r _{_ β} ^{(l, q)}(k) and r _β ^{(l, q)}(k); Wherein, q=1 ..., Q.

Then the frequency offset estimating size is:

$\mathrm{\Δ}{\hat{f}}^{\left(l\right)}=a\·\frac{\underset{q=1}{\overset{Q}{\mathrm{\Σ}}}({\left|\underset{k=0}{\overset{\mathrm{\δ}-1}{\mathrm{\Σ}}}{r}_{\mathrm{\β}}^{(l,q)}\left(k\right)\right|}^2-{\left|\underset{k=0}{\overset{\mathrm{\δ}-1}{\mathrm{\Σ}}}{r}_{-\mathrm{\β}}^{(l,q)}\left(k\right)\right|}^2)}{\underset{q=1}{\overset{Q}{\mathrm{\Σ}}}\left({\left|\underset{k=0}{\overset{\mathrm{\δ}-1}{\mathrm{\Σ}}}{\tilde{r}}_c^{(l,q)}\left(k\right)\right|}^2\right)}---\left(10\right)$

In the following formula, parameter a is for adjusting the factor, and a constant factor normally in system can be determined by emulation.

Afterwards, two pilot frequency locations places of a subframe can be estimated that the system frequency deviation of gained does on average, obtain the frequency deviation value that this subframe is estimated at the beginning of current, that is:

$\mathrm{\Δ}\hat{f}=(\mathrm{\Δ}{\hat{f}}^{\left(4\right)}+\mathrm{\Δ}{\hat{f}}^{\left(11\right)})/2---\left(11\right)$

At last, with the first frequency deviation value of estimating and a last subframe gained frequency deviation of current subframe

Carry out the recurrence smoothing processing, the frequency offset estimating of current n subframe is promptly arranged

For:

$\mathrm{\Δ}\hat{f}\left(n\right)=\left[\begin{array}{cc}\mathrm{\Δ}\hat{f}& n=1\\ (1-\mathrm{\μ})\·\mathrm{\Δ}\hat{f}(n-1)+\mathrm{\μ}\·\mathrm{\Δ}\hat{f}& n>1\end{array}\right.---\left(12\right)$

Wherein, μ is a smoothing factor, can get the constant between (0～1) usually.

After the completing steps 205, just can carry out channel estimating at the Data Position place, concrete operating process is made up of to 208 step 206.

Step 206: the start-phase deviation of eliminating the pilot frequency locations channel estimating.

Particularly, can be according to the frequency offset estimating value of current subframe, calculate two pilot frequency locations places respectively because the corresponding symbol start-phase deviation that frequency deviation is introduced, remove the channel estimation results at two pilot frequency locations places of equalization step 203 gained more respectively with two symbol start-phase deviations that obtain, finally obtain the channel estimation results of targeted customer at the pilot frequency locations place.Concrete processing method can for:

Utilization is to current subframe frequency offset estimating

Go to eliminate pilot frequency locations place channel estimating On the start-phase deviation, obtain

That is:

${\tilde{h}}^{\left(l\right)}=e^{-\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="H2010100345102E00011.png"\; state="\{\{state\}\}">j</figure-callout>}2\mathrm{\&pi;}\&CenterDot;\mathrm{\&Delta;}\hat{f}\left(n\right)\&CenterDot;t^{\left(l\right)}}\&CenterDot;{\hat{h}}^{\left(l\right)}---\left(13\right)$

Wherein, t ^(l)Be illustrated in the initial sampling point sequence number of the 1st symbol in the current subframe under the systematic sampling frequency, only relevant with the CP form usually with system bandwidth, and under the situation that system bandwidth and CP form have been determined, will be known constant.

Step 207: use linear interpolation data estimator position channel estimating.

Particularly, can use linear interpolation algorithm, carry out the channel estimating at Data Position place according to the channel estimation results at two pilot frequency locations places.Concrete processing method can for:

Utilize the estimation at pilot frequency locations place

The estimation of calculated data position

That is:

${\tilde{h}}^{\left(i\right)}=(i-4)\&CenterDot;\left(\frac{{\tilde{h}}^{\left(11\right)}-{\tilde{h}}^{\left(4\right)}}{7}\right)+{\tilde{h}}^{\left(4\right)},i\&Element;\{\mathrm{1,2},...7\}---\left(14\right)$

${\tilde{h}}^{\left(i\right)}=(i-11)\&CenterDot;\left(\frac{{\tilde{h}}^{\left(11\right)}-{\tilde{h}}^{\left(4\right)}}{7}\right)+{\tilde{h}}^{\left(11\right)},i\&Element;\{8,9,...14\}---\left(15\right)$

Step 208: the start-phase reduction of Data Position place channel estimating.

Particularly, can be according to the frequency offset estimating value of current subframe, the estimation at the Data Position place of recovering step 207 gained The start-phase of each data symbol, obtain the channel estimation results of final data position.Concrete processing method can for:

Frequency offset estimating value according to current subframe

Data Position place to the interpolation gained estimates

Carry out the start-phase reduction, can obtain Data Position place channel estimating

${\hat{h}}^{\left(i\right)}=e^{-\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="H2010100345102E00011.png"\; state="\{\{state\}\}">j</figure-callout>}2\mathrm{\&pi;}\&CenterDot;\mathrm{\&Delta;}\hat{f}\left(n\right)\&CenterDot;t^{\left(i\right)}}\&CenterDot;{\hat{h}}^{\left(i\right)}---\left(16\right)$

Step 209: Data Position place antenna merges and channel equalization.

Particularly, can be according to the channel estimation results at Data Position place, and the signal at the Data Position place that receives, utilize antenna to merge and equalization algorithm, obtain the OFDM modulating data of frequency domain.Concrete processing method can for:

Channel estimating according to the Data Position place

And the signal y at the Data Position place that receives ⁽ⁱ⁾, utilize antenna to merge and the equilibrium treatment algorithm, obtain the frequency domain data estimate vector at Data Position place

Step 210: do reverse discrete Fourier transform (DFT) (IDFT) and obtain the SC-FDMA modulation symbol.

Particularly, can carry out the IDFT processing and be converted to time domain data, obtain the SC-FDMA modulating data with frequency domain data with step 209 gained.Concrete processing method can for:

To each Data Position place

Be IDFT respectively to obtain the estimation of SC-FDMA modulation symbol

Step 211: the SC-FDMA modulation symbol is done compensate of frequency deviation.

Particularly, can be according to the frequency offset estimating value of the current subframe of gained before, to SC-FDMA modulating data compensating for frequency offset.Concrete processing method can for:

Utilize the current subframe frequency offset estimating of gained Estimation to the SC-FDMA modulation symbol

Carry out compensate of frequency deviation, can obtain final data and estimate

${\hat{s}}^{\left(i\right)}=\left\{\begin{array}{cc}{\tilde{s}}^{\left(i\right)}& \left|\mathrm{\&Delta;}\hat{f}\left(n\right)\right|<0.02\\ \mathrm{diag}\left\{\left[\begin{array}{ccc}{e}^{-\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="H2010100345102E00011.png"\; state="\{\{state\}\}">j</figure-callout>}2\mathrm{\&pi;}\frac{-M/2}{M}\mathrm{\&Delta;}\hat{f}\left(n\right)}& ...& e^{-\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="H2010100345102E00011.png"\; state="\{\{state\}\}">j</figure-callout>}2\mathrm{\&pi;}\frac{M/2-1}{M}}\end{array},\mathrm{\&Delta;}\hat{f}\left(n\right)\right]\right\}\&CenterDot;{\tilde{s}}^{\left(i\right)}& \left|\mathrm{\&Delta;}\hat{f}\left(n\right)\right|\&GreaterEqual;0.02\end{array}\right.---\left(17\right)$

Need to prove, under the actual application environment, when system works in more abominable following time of channel condition, the performance of system will be subjected to relatively seriously influencing, even " floor effect " appears, promptly adopt the various measures that improve signal quality all can not improve the performance of link.In addition, when existing unavoidably between actual sending and receiving system, imperfect zeitgeber frequently, this will make and exist certain time frequency deviation between receive-transmit system, and the existence of this factor also can further worsen the performance of system, when serious even cause system physical layer chain rupture to occur.Flow process shown in Figure 2 compared with prior art not only is applicable to the channel estimation process of user under the high-speed mobile scene, and can also suppress system bigger under the high-speed mobile condition effectively the time, frequency deviation is to the system link Effect on Performance.And, by with channel estimating with combine based on the frequency offset estimating algorithm of single pilot tone, make systematicness can be to bigger system time the, frequency deviation performance sane, and by introducing simple compensate of frequency deviation algorithm, further weakened the influence of the asynchronous factor of system, thereby the system that makes can overcome " floor effect ", and the realization of whole flow process is simple, flexible, can be the lifting that cost exchanges performance for low operand and complexity.

Therefore, flow process shown in Figure 2 can effectively reduce the channel estimating complexity, and significantly improves channel estimation accuracy.

Flow process shown in Figure 2 can be simplified as shown in Figure 3.Referring to Fig. 3, Fig. 3 is a channel estimating general flow chart of the present invention, and this flow process may further comprise the steps:

Step 310: carry out pilot frequency locations place channel estimating in conjunction with frequency offset estimating.

Step 320: carry out frequency offset estimating based on single pilot tone in conjunction with pilot frequency locations place channel estimating.

Step 330: in conjunction with carrying out Data Position place channel estimating based on the frequency offset estimation result of single pilot tone.

Step 340: in conjunction with the compensate of frequency deviation that carries out data estimator based on the frequency offset estimation result of single pilot tone.

In order to realize Fig. 2, flow process shown in Figure 3 smoothly, can carry out setting as shown in Figure 4.Referring to Fig. 4, Fig. 4 is the channel estimating apparatus figure of the embodiment of the invention, and this device comprises continuous FFT unit, user's separative element, pilot frequency locations place channel estimating unit, Data Position place channel estimating unit, balanced unit, frequency offset compensation element; Also comprise frequency deviation estimating unit, the one end links to each other with pilot frequency locations place channel estimating unit, and the other end links to each other with Data Position place channel estimating unit, frequency offset compensation element respectively; As: the input of frequency deviation estimating unit links to each other with the output of pilot frequency locations place channel estimating unit, and the output of frequency deviation estimating unit then links to each other with the input of Data Position place channel estimating unit, the input of frequency offset compensation element respectively.

When practical application, the FFT unit is converted to frequency domain data with the time-domain signal of receiving, and sends to user's separative element, is extracted targeted customer's reception pilot frequency sequence y by user's separative element ^(l), send into pilot frequency locations place channel estimating unit again, utilize two known frequency offsets sizes (normalization frequency deviation β by pilot frequency locations place channel estimating unit ₁, β ₂≤ 0.5) user's emission pilot frequency sequence

With

And the last frequency deviation valuation of estimating

Direct match obtains frequency deviation and is

User's pilot tone

The LS algorithm of application enhancements obtains pilot frequency locations place channel estimating initial value again; Afterwards, through the filtering noise reduction, and output pilot frequency locations place channel estimating

Give frequency deviation estimating unit and Data Position place channel estimating unit based on single pilot tone.

Based on the frequency deviation estimating unit of single pilot tone at first according to the channel estimating at two the pilot frequency locations places of targeted customer on a subframe

Receive pilot frequency sequence y to two respectively ^(l)Carry out channel equalization separately, to two pilot signals after the equilibrium

Carry out frequency offset estimating independently of one another, so just can in a subframe, obtain two frequency offset estimating values of targeted customer based on single pilot tone.At this moment, if current subframe is first subframe, so the frequency offset estimating that on average obtains current subframe is done in these two frequency deviation valuations

Otherwise, these two frequency deviation valuations are carried out the recurrence smoothing processing with the frequency offset estimating value of a last subframe, obtain the frequency offset estimating of current subframe

Afterwards, frequency deviation estimating unit will

Send into Data Position place channel estimating unit and frequency offset compensation element.

Data Position place channel estimating unit is at first according to frequency offset estimating Calculate the common phase deviation at two pilot frequency locations places, and go the balanced channel estimating of corresponding position separately with the common phase deviation that obtains To obtain Then; Utilize interpolation algorithm, according to two groups of estimations

Obtain the channel estimating at Data Position place At last, utilize frequency offset estimating

Recovering each Data Position place estimates

The common phase deviation, obtain the channel estimating at Data Position place

And send to balanced unit.

Balanced unit can utilize to be received

The reception data y at balanced corresponding character position place ⁽ⁱ⁾Thereby, obtain the frequency domain data estimate vector of corresponding data position

Again each frequency domain data estimate vector is IDFT respectively and handles, estimate thereby obtain modulation symbol

And it is sent into frequency offset compensation element; By the frequency offset estimating of frequency offset compensation element according to current subframe

To the modulation sign estimation Carry out compensate of frequency deviation, can obtain the SC-FDMA modulation symbol and estimate

Referring to Fig. 5, Fig. 5 is the present invention and the link simulation performance schematic diagram of prior art on the EVA300 channel.Abscissa among Fig. 5 is represented the link simulation signal to noise ratio snr (dB) set, and ordinate is illustrated under the corresponding signal-to-noise ratio settings, the link performance index frame error rate (FER) of the present invention and prior art.As we can see from the figure, existing channel estimating the performance level land occurs very soon along with the raising of SNR, i.e. " floor effect "; And the technology of the present invention can exist not exclusively synchronously in system, when particularly having bigger system, under the situation of frequency deviation, can overcome " floor effect " preferably, compare with the link performance under the ideal communication channel of equal channel condition (not having system frequency deviation) is estimated, its performance is also more approaching, and the normal performance working point of the link that it reflects is near the actual signal to noise ratio working point of actual reception machine system.And, because therefore the moderate complexity that the algorithm of the technology of the present invention is realized has the excellent engineering meaning.

In sum as seen, because the inventive method and device can be in advance the content that will be applied to channel estimating carried out frequency offset estimating based on single pilot tone, and then in conjunction with carrying out channel estimating based on the frequency offset estimation result of single pilot tone; Therefore, the inventive method and device all can effectively reduce the channel estimating complexity, and obviously improve channel estimation accuracy.

The above is preferred embodiment of the present invention only, is not to be used to limit protection scope of the present invention, all any modifications of being done within the spirit and principles in the present invention, is equal to and replaces and improvement etc., all should be included within protection scope of the present invention.

Claims (10)

1. a channel estimation methods is characterized in that, this method comprises:

The content that will be applied to channel estimating is carried out frequency offset estimating based on single pilot tone; Carry out channel estimating in conjunction with frequency offset estimation result based on single pilot tone.

2. method according to claim 1 is characterized in that, carries out comprising based on the process of the described frequency offset estimating of single pilot tone:

Carrying out system frequency deviation at the pilot frequency locations place based on single frequency pilot sign estimates, and will be at estimating on two included frequency pilot signs of current subframe that gained system frequency deviation value and the frequency deviation value that calculates gained at a last subframe carry out the recurrence smoothing processing, obtain the frequency offset estimating value of current subframe.

3. method according to claim 2 is characterized in that, before carrying out described frequency offset estimating, further the pilot frequency sequence that receives is comprised the processing of least square LS estimation, filtering noise reduction, equilibrium.

4. according to each described method of claim 1 to 3, it is characterized in that, further carry out compensate of frequency deviation carrying out the result that described channel estimating obtains.

5. method according to claim 4 is characterized in that, the method for described compensate of frequency deviation is:

According to the frequency offset estimating value of the current subframe of gained before, to SC-FDMA modulating data compensating for frequency offset.

6. method according to claim 4 is characterized in that, the described channel estimating of carrying out is carried out at the Data Position place, and its process comprises:

Eliminate the start-phase deviation of pilot frequency locations channel estimating, use linear interpolation data estimator position channel estimating, the start-phase reduction of Data Position place channel estimating, Data Position place antenna merges and channel equalization, is reverse discrete Fourier transform (DFT) IDFT and obtains the SC-FDMA modulation symbol.

7. a channel estimating apparatus is characterized in that, this device comprises continuous frequency deviation estimating unit, Data Position place channel estimating unit; Wherein,

Described frequency deviation estimating unit is used for the content that will be applied to channel estimating is carried out frequency offset estimating based on single pilot tone;

Described Data Position place channel estimating unit is used for carrying out channel estimating in conjunction with the frequency offset estimation result of described frequency deviation estimating unit.

8. device according to claim 7, it is characterized in that, further comprise pilot frequency locations place channel estimating unit, be used for the pilot frequency sequence that receives is comprised LS estimation, filtering noise reduction, balanced processing, and the pilot frequency locations place channel estimation results that will handle gained sends to described frequency deviation estimating unit and Data Position place channel estimating unit.

9. according to claim 7 or 8 described devices, it is characterized in that, further comprise balanced unit, frequency offset compensation element; Wherein,

Described balanced unit is used for the reception data at balanced corresponding character position place, obtains the frequency domain data estimate vector of corresponding data position; Again each frequency domain data estimate vector is IDFT respectively and handles, obtain the modulation symbol estimated result, and it is sent into described frequency offset compensation element;

Described frequency offset compensation element is used for according to the frequency offset estimation result of the current subframe of receiving the modulation symbol estimated result of receiving being carried out compensate of frequency deviation, obtains the channel estimation results that needs;

Described frequency deviation estimating unit is further used for the frequency offset estimation result of current subframe is sent to described frequency offset compensation element.

10. device according to claim 9 is characterized in that, under the situation that is provided with pilot frequency locations place channel estimating unit, the annexation between described frequency deviation estimating unit and pilot frequency locations place channel estimating unit, the Data Position place channel estimating unit is:

The input of described frequency deviation estimating unit links to each other with the output of pilot frequency locations place channel estimating unit, and the output of frequency deviation estimating unit links to each other with the input of Data Position place channel estimating unit, the input of frequency offset compensation element respectively.

Images