Channel estimation method and implementing device for orthogonal frequency division multiplexing mobile communication system
Technical Field
The present invention relates to a channel estimation method in a wireless mobile communication system, and more particularly, to a channel estimation method and an implementation apparatus in an orthogonal frequency division multiplexing mobile communication system, and more particularly, to a channel estimation method and an implementation apparatus in combination with a prefix and a pilot.
Background
Orthogonal Frequency Division Multiplexing (OFDM), a multi-carrier transmission mode, greatly reduces the sensitivity of the system to the frequency selectivity of a multipath fading channel by converting a high-speed transmission data stream into a set of low-speed parallel transmission data streams, and the introduction of a cyclic prefix further enhances the capability of the system to resist inter-symbol interference (ISI). In addition, the characteristics of high bandwidth utilization rate, simple implementation and the like enable OFDM to be applied more and more widely in the field of wireless communication, and channel estimation, which is a key technology of an OFDM system, also becomes a hot point of research in the field of communication and receives more and more attention.
There are two types of training sequence insertion forms for OFDM systems, block pilot and comb pilot. Most of the existing channel estimation methods are based on some pilot arrangement (block pilot or comb pilot).
These two types of channel estimation methods are each characterized: the channel estimation method based on the block pilot frequency is not very sensitive to the frequency selectivity of the channel, but is only suitable for the wireless channel with slow fading; the channel estimation method based on the comb-shaped pilot frequency can be suitable for a wireless channel with fast fading, but is sensitive to the frequency selectivity of the channel. However, combining the advantages of the two channel estimations, the method for estimating the channel is not common.
For an 802.16 system based on the OFDMA standard, it has both block pilots (hereinafter referred to as prefixes) and comb pilots (hereinafter referred to directly as pilots). If only the prefix is used for channel estimation, because the prefix symbol is only sent once in one frame, when the channel is a fast fading channel, the change of the channel can not be accurately tracked only by using the prefix, thereby affecting the demodulation performance of the system. If the channel estimation is performed only by using the pilot frequency, when the frequency selectivity of the channel is significant, the accuracy of the pilot frequency-based channel estimation method is also reduced, and even a "floor effect" is generated, that is, the bit error rate of the system is not increased along with the increase of the signal-to-noise ratio.
Taking an 802.16 system based on the OFDMA standard as an example, in the FUSC mode, the pilot carrier spacing is 66kHz, and for a channel whose channel model is SUI-5 (modified university of stanford temporary channel model-5), the coherence bandwidth of the channel is 70 kHz; for a channel with a channel model SUI-6 (modified Stanford university temporary channel model-6), the coherence bandwidth of the channel is 38 kHz. The pilot carrier spacing is close to the coherence bandwidth of the SUI-5 channel, and is greater than the coherence bandwidth of the SUI-6 channel, which is very prone to floor effects. Therefore, it is necessary to combine the prefix and the pilot signal to estimate the wireless channel.
The publication No. CN1505293A, publication date 6/16 in 2004, "channel estimation method and estimator in orthogonal frequency division multiplexing system", proposes a method for channel estimation by comprehensively using prefixes and pilots, which can be applied to fast fading channels, and also can be applied to channels with significant channel frequency selectivity, including the case where the pilot carrier spacing is smaller than the correlation bandwidth.
If the channel type selection method is not proper, the channel type cannot be accurately judged, the performance of the method is drastically deteriorated, and the patent does not give a judgment method of the channel type; for the main propagation path selection, the patent uses prefix impulse response to select the main path, wherein the threshold is selected, and the channel estimation performance is sensitive to the selection of the threshold, and simulation shows that if the threshold is not selected properly, the channel estimation performance is reduced sharply.
Disclosure of Invention
One of the objectives of the present invention is to provide a channel estimation method for an ofdm mobile communication system by combining a prefix and a pilot, which has excellent performance and is easy to implement, and can be adapted to various channel types, such as a frequency selective channel, a fast fading channel, and the like.
The invention also provides a device for realizing the method.
In order to achieve the purpose of the invention, the invention provides a method and a device for channel estimation by combining a prefix and a pilot frequency. Generating a pilot signal used for channel estimation by a system transmitter, wherein the pilot signal comprises a pilot insertion mode and two parts of content selected by a pilot value; the generated pilot signal is transformed into a time domain pilot signal through inverse Fourier transform; transmitting the time domain pilot signal on a wireless multi-path fading channel; and the system receiver performs Fourier transform on the time domain pilot signal to obtain a frequency domain pilot signal.
The invention is realized in such a way that:
a channel estimation method of an orthogonal frequency division multiplexing mobile communication system, comprising the processes of:
the first step, the impulse response rough estimation of the corresponding symbol channel is obtained according to the prefix, or the decision feedback information, or the impulse response information of the last OFDM symbol;
a second step of generating a construction matrix according to a local reference pilot signal, wherein the construction matrix is a conjugate transpose square matrix, the dimension of the construction matrix is the length of the rough estimation of the channel impulse response, and the element t (p) of the construction matrix is calculated according to the formula (1);
<math><mrow><mi>t</mi><mrow><mo>(</mo><mi>p</mi><mo>)</mo></mrow><mo>=</mo><mo>-</mo><mi>a</mi><munder><mi>Σ</mi><mrow><mi>k</mi><mo>⋐</mo><mi>P</mi></mrow></munder><msup><mrow><mo>|</mo><mi>X</mi><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><mo>|</mo></mrow><mn>2</mn></msup><msubsup><mi>W</mi><mi>N</mi><mrow><mo>-</mo><mi>kp</mi></mrow></msubsup><mo>,</mo><mi>P</mi><mo>=</mo><mn>0,1</mn><mo>,</mo><mo>.</mo><mo>.</mo><mo>.</mo><mo>,</mo><mi>L</mi><mo>-</mo><mn>1</mn><mo>-</mo><mo>-</mo><mo>-</mo><mrow><mo>(</mo><mn>1</mn><mo>)</mo></mrow></mrow></math>
wherein k represents an offset of the carrier; p represents a set of pilot carrier offsets; a is a convergence factor, and satisfies 0 < a < 1; x (k) denotes a local reference pilot signal;
is an inverse fourier transform factor;
a third step of calculating a cross-correlation vector from the local reference pilot signal and the received pilot signal, the length of the cross-correlation vector being equal to the dimension of the construction matrix, the elements r (p) of the cross-correlation vector being calculated according to equation (2);
<math><mrow><mi>r</mi><mrow><mo>(</mo><mi>p</mi><mo>)</mo></mrow><mo>=</mo><mi>a</mi><munder><mi>Σ</mi><mrow><mi>k</mi><mo>⋐</mo><mi>P</mi></mrow></munder><mi>Y</mi><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><msup><mi>X</mi><mo>*</mo></msup><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><msubsup><mi>W</mi><mi>N</mi><mrow><mo>-</mo><mi>kp</mi></mrow></msubsup><mo>,</mo><mi>P</mi><mo>=</mo><mn>0</mn><mo>,</mo><mo>.</mo><mo>.</mo><mo>.</mo><mo>,</mo><mi>L</mi><mo>-</mo><mn>1</mn><mo>-</mo><mo>-</mo><mo>-</mo><mrow><mo>(</mo><mn>2</mn><mo>)</mo></mrow></mrow></math>
wherein y (k) represents a received pilot signal; a is a convergence factor, and satisfies 0 < a < 1;
a fourth step of obtaining a fine estimation result of the channel impulse response by a recursive equation according to the construction matrix and the cross-correlation vector, wherein the recursive equation is as follows: hm,n+1=Hm,n+R-THm,nIn the formula, Hm,nRepresents the fine estimation result of the nth channel impulse response of the mth orthogonal frequency division multiplexing symbol in a certain frame, wherein n is 0, 1, 2, …;
a fifth step of judging whether the fine estimation result obtained according to the recursion equation is converged, if so, outputting the fine estimation result of the impulse response, and if not, continuing to calculate the recursion equation;
sixthly, performing fast Fourier transform of all the carrier numbers of the system on the fine estimation result of the output impulse response to obtain a channel estimation result of all the carrier numbers;
and a seventh step of selecting and outputting useful data carrier channel estimation results from the channel estimation results of all the carriers in the system.
And the fifth step of outputting an impulse response fine estimation result by judging whether the mean square error of the results of two continuous recursive operations is less than a fixed constant or not, and if not, taking the result of the recursive operation as the input of the next recursive operation and continuing the recursive operation.
And the fifth step fixes recursion times, outputs a fine estimation result of impulse response after carrying out recursion calculation of corresponding times, and takes the fine estimation result as the impulse response information of the next orthogonal frequency division multiplexing symbol.
The length of the fine estimation of the channel impulse response is equal to the dimension of the construction matrix.
An apparatus for implementing a channel estimation method of an orthogonal frequency division multiplexing mobile communication system, comprising:
the channel impulse response rough estimator obtains the impulse response rough estimation of the channel according to the prefix, the decision feedback or the impulse response information of the last orthogonal frequency division multiplexing symbol;
a construction matrix generator, which generates a construction matrix according to the local reference pilot signal, wherein the construction matrix is a conjugate transpose square matrix, the dimension of the construction matrix is the length of the rough estimation of the channel impulse response, and the element t (p) of the construction matrix is calculated according to the formula (1);
<math><mrow><mi>t</mi><mrow><mo>(</mo><mi>p</mi><mo>)</mo></mrow><mo>=</mo><mo>-</mo><mi>a</mi><munder><mi>Σ</mi><mrow><mi>k</mi><mo>⋐</mo><mi>P</mi></mrow></munder><msup><mrow><mo>|</mo><mi>X</mi><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><mo>|</mo></mrow><mn>2</mn></msup><msubsup><mi>W</mi><mi>N</mi><mrow><mo>-</mo><mi>kp</mi></mrow></msubsup><mo>,</mo><mi>P</mi><mo>=</mo><mn>0,1</mn><mo>,</mo><mo>.</mo><mo>.</mo><mo>.</mo><mo>,</mo><mi>L</mi><mo>-</mo><mn>1</mn><mo>-</mo><mo>-</mo><mo>-</mo><mrow><mo>(</mo><mn>1</mn><mo>)</mo></mrow></mrow></math>
wherein k represents an offset of the carrier; p represents a set of pilot carrier offsets; a is a convergence factor, and satisfies 0 < a < 1; x (k) denotes a local reference pilot signal;is an inverse fourier transform factor;
a cross-correlation vector generator for generating a cross-correlation vector by calculation based on the local reference pilot signal and the received pilot signal, wherein the length of the cross-correlation vector is equal to the dimension of the construction matrix, and the element r (p) of the cross-correlation vector is calculated according to the formula (2);
<math><mrow><mi>r</mi><mrow><mo>(</mo><mi>p</mi><mo>)</mo></mrow><mo>=</mo><mi>a</mi><munder><mi>Σ</mi><mrow><mi>k</mi><mo>⋐</mo><mi>P</mi></mrow></munder><mi>Y</mi><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><msup><mi>X</mi><mo>*</mo></msup><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><msubsup><mi>W</mi><mi>N</mi><mrow><mo>-</mo><mi>kp</mi></mrow></msubsup><mo>,</mo><mi>P</mi><mo>=</mo><mn>0</mn><mo>,</mo><mo>.</mo><mo>.</mo><mo>.</mo><mo>,</mo><mi>L</mi><mo>-</mo><mn>1</mn><mo>-</mo><mo>-</mo><mo>-</mo><mrow><mo>(</mo><mn>2</mn><mo>)</mo></mrow></mrow></math>
wherein y (k) represents a received pilot signal; a is a convergence factor, and satisfies 0 < a < 1;
a data selector that selects either a coarse estimate of the channel impulse response or the result of a recursive equation as one input to the multiplier, the recursive equation being: hm,n+1=Hm,n+R-THm,nIn the formula, Hm,nRepresents the fine estimation result of the channel impulse response of the nth time of the mth orthogonal frequency division multiplexing symbol in a certain frame, wherein n is 0, 1, 2,…;
A multiplier for calculating and outputting the product of the construction matrix and the output of the data selector;
an adder for calculating and outputting the sum of outputs of the cross-correlation vector generator, the multiplier, and the data selector;
a termination condition decider for judging whether the recursion equation is converged, if so, terminating the recursion equation calculation and outputting channel impulse response, otherwise, continuing the recursion equation calculation;
a fast Fourier transformer for performing fast Fourier transform of the number of system carriers on the output of the termination condition decider;
and the channel estimation result selector selects useful data carrier channel estimation from the channel estimation results of all the carriers and outputs the useful data carrier channel estimation.
And the termination condition decision device judges whether the mean square error of the results of two continuous recursive calculations is smaller than a fixed constant or not, if so, outputs an impulse response fine estimation result, otherwise, takes the result of the recursive calculation as the input of the next recursive calculation and continues the recursive calculation.
The termination condition decider fixes recursion times, outputs a fine estimation result of impulse response after carrying out recursion calculation of corresponding times, and takes the fine estimation result as the impulse response information of the next orthogonal frequency division multiplexing symbol.
Compared with the existing channel estimation method, the method is suitable for fast fading channels and slow fading channels, including the condition that the density of pilot frequency is less than the related bandwidth, does not need to judge the type of the channel and the main transmission path selection of the channel, and has the characteristics of small calculated amount, small required storage space, no need of threshold selection, easy realization and the like.
Drawings
FIG. 1 is a schematic diagram of a receiver architecture incorporating the present invention;
fig. 2 is a schematic diagram of an internal implementation of the channel estimator proposed by the present invention;
fig. 3 is a flow chart corresponding to the internal implementation schematic diagram of the channel estimator proposed by the present invention.
Detailed Description
The invention is described in detail below with reference to the attached drawing figures:
in order to achieve the first object, the method provided by the invention mainly comprises the following steps:
1. obtaining the impulse response rough estimation H of the channel of the mth symbol according to the prefix or the decision feedback information or the impulse response information of the last OFDM symbolm,0。
2. A construction matrix T is generated according to equation (1), the dimension of T being equal to the length of the coarse estimate of the impulse response. The construction matrix T is a matrix satisfying T ═ THSquare matrix of relationships, here THRepresents the conjugate transpose of the construction matrix T, which, when the dimension of the construction matrix T is L, is of the form:
wherein,
<math><mrow><mi>t</mi><mrow><mo>(</mo><mi>p</mi><mo>)</mo></mrow><mo>=</mo><mo>-</mo><mi>a</mi><munder><mi>Σ</mi><mrow><mi>k</mi><mo>⋐</mo><mi>P</mi></mrow></munder><msup><mrow><mo>|</mo><mi>X</mi><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><mo>|</mo></mrow><mn>2</mn></msup><msubsup><mi>W</mi><mi>N</mi><mrow><mo>-</mo><mi>kp</mi></mrow></msubsup><mo>,</mo><mi>P</mi><mo>=</mo><mn>0,1</mn><mo>,</mo><mo>.</mo><mo>.</mo><mo>.</mo><mo>,</mo><mi>L</mi><mo>-</mo><mn>1</mn><mo>-</mo><mo>-</mo><mo>-</mo><mrow><mo>(</mo><mn>2</mn><mo>)</mo></mrow></mrow></math>
k represents an offset of the carrier;
p represents a set of pilot carrier offsets;
a is a convergence factor, and satisfies 0 < a < 1;
x (k) denotes a local reference pilot signal;
is an inverse fourier transform factor.
3. The cross-correlation vector R is calculated, the length of R being equal to the dimension L of the matrix T, the cross-correlation vector R being of the form:
R=[r(0),......r(L-1)]T; (3)
wherein,
<math><mrow><mi>r</mi><mrow><mo>(</mo><mi>p</mi><mo>)</mo></mrow><mo>=</mo><mi>a</mi><munder><mi>Σ</mi><mrow><mi>k</mi><mo>⋐</mo><mi>P</mi></mrow></munder><mi>Y</mi><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><msup><mi>X</mi><mo>*</mo></msup><mrow><mo>(</mo><mi>k</mi><mo>)</mo></mrow><msubsup><mi>W</mi><mi>N</mi><mrow><mo>-</mo><mi>kp</mi></mrow></msubsup><mo>,</mo><mi>P</mi><mo>=</mo><mn>0</mn><mo>,</mo><mo>.</mo><mo>.</mo><mo>.</mo><mo>,</mo><mi>L</mi><mo>-</mo><mn>1</mn><mo>-</mo><mo>-</mo><mo>-</mo><mrow><mo>(</mo><mn>4</mn><mo>)</mo></mrow></mrow></math>
y (k) denotes a received pilot signal;
a has the same meaning as above.
4. And obtaining a fine estimation result of the channel impulse response according to the formula (5).
Hm,n+1=Hm,n+R-THm,n (5)
Wherein,
Hm,ndenotes the fine estimation result of the nth channel impulse response of the mth OFDM symbol in a certain frame, where n is 0, 1, 2, …. When n is 0, Hm,0Is a coarse estimate of the channel impulse response, H1,0=H0,0,H0,0Is the channel impulse response estimate of the prefix. Hm,nIs a vector of length equal toThe dimensions of the matrix T are constructed.
5. To Hm,n+1And performing N-point FFT to obtain the channel estimation results of all N carriers, wherein N refers to the number of all the carriers of the system.
6. And selecting useful data carrier channel estimation results from the channel estimation results of the N carriers and outputting the useful data carrier channel estimation results.
The channel estimation device corresponding to the above steps comprises the following modules:
1. and a cross-correlation vector generator for generating a cross-correlation vector R represented by equation (3).
2. And the construction matrix generator is used for generating a construction matrix T according to the channel impulse response coarse estimator, wherein the dimension of the matrix T is L.
3. And the channel impulse response rough estimator calculates the impulse response rough estimation of the channel according to the prefix, the decision feedback or the impulse response information of the previous OFDM symbol.
4. A data selector operative to select either the coarse estimate of the channel impulse response or the result of recursive equation (5) as one input to the multiplier.
5. And the multiplier is used for calculating and outputting the product of the construction matrix T and the output of the data selector.
6. And an adder for calculating and outputting the sum of the outputs of the cross-correlation vector generator, the multiplier, and the data selector.
7. A termination condition determiner for determining whether the recursive equation (5) has converged, terminating the recursive calculation if it has converged, and outputting a channel impulse response Hm,n+1Otherwise, the recursive computation is continued.
8. FFT converter for outputting H to termination condition deciderm,n+1An N-point FFT is performed, N referring to the number of all carriers of the system.
9. And a channel estimation result selector for selecting useful data carrier channel estimation from the channel estimation results of the N carriers and outputting the useful data carrier channel estimation.
In the receiver in fig. 1, a front-end processing module 101 receives a radio frequency signal through a receiving antenna, then performs down-conversion, analog-to-digital conversion, cyclic prefix removal and the like on the received signal, and finally, data is output to an FFT module 102.
The FFT block 102 performs a fast fourier transform on the input data so that the data is transformed from the time domain to the frequency domain. The FFT module 102 outputs data from the data carrier to the channel compensation module 104 and data from the pilot carrier to the channel estimation module 103.
The channel compensation module 104 receives the user data output by the FFT module 102 and the data carrier channel estimation output by the channel estimation module 103, and performs amplitude and phase compensation on the user data using the channel estimation of the data carrier.
Finally, the data is output to the demodulation and decoding module 105 for subsequent processing.
Fig. 2 is a schematic diagram of an internal implementation of the channel estimator proposed by the present invention.
The channel impulse response rough estimator 201 obtains the impulse response rough estimation of the channel according to the prefix, the decision feedback or the impulse response information of the previous OFDM symbol.
The construction matrix generator 202 generates a construction matrix T of a corresponding dimension according to equation (1) based on the dimension of the matrix determined by the channel impulse response coarse estimator.
The cross-correlation vector generator 203 generates a cross-correlation vector R shown in equation (3) based on the local reference pilot signal and the pilot signal at the receiving end.
The data selector 204 is used to select the initial value of each symbol for each recursion, and if it is the first symbol, the first initial value H1,0Equal to prefix impulse response H0,0Else, each time the initial value Hm,0Are all the lastImpulse response H of OFDM symbolm-1,n。
The multiplier 205 calculates and outputs the product of the outputs of the construction matrix generator and the data selector.
The adder 206 calculates the sum of the outputs of the cross-correlation vector generator, the multiplier, and the data selector, and outputs the sum.
The termination condition decider 207 is for deciding the end of the recursive computation.
The fast fourier transformer 208 is a channel estimate H for the output of the termination condition deciderm,n+1And taking N-point Fourier transform, wherein N refers to the number of all carriers of the system.
The channel estimation result selector 209 selects and outputs channel estimation of a useful data carrier from the channel estimation results of the N carriers.
Fig. 3 corresponds to a flow chart of a channel estimation implementation.
Step 301 calculates the rough estimation of the impulse response of the channel according to the information such as the prefix, the decision feedback or the impulse response of the previous OFDM symbol, and the rough estimation is equivalent to the initial value H calculated in each recursionm,0If it is the first OFDM symbol of a frame, the initial value H of the recursive computation1,0Selected prefix impulse response estimate H0,0Otherwise, selecting the impulse response estimation H of the last symbolm-1,n. And determining the dimension of the matrix T according to the length of the channel impulse response rough estimation.
Step 302 determines the dimension of matrix T according to the pilot frequency, prefix or channel impulse response information of the previous OFDM symbol, and generates a constructed matrix T according to equation (1). Taking the OFDMA system based on the 802.16 protocol as an example, the length of the rough estimation of the channel impulse response can be obtained from the prefix information, and the dimension of the construction matrix T is equal to this length. The present invention does not limit the method of determining the dimension of the matrix T using the pilot, the prefix, or the channel impulse response information of the last OFDM symbol. Simulation shows that the method is insensitive to the accuracy of the channel impulse response length.
Step 303 generates a cross-correlation vector R according to the cross-correlation R (p) between the local reference pilot signal and the pilot signal at the receiving end, as shown in equation (3), where the length of the vector R is equal to the dimension of the matrix T.
Step 304 computes recursive equation (5) from the construction matrix T and the cross-correlation vector R.
Step 305 determines whether the recursive computation is finished according to the result of each recursive computation, and the termination condition may be determined in two ways:
one is to determine whether the average difference of the results of two consecutive recursion operations is less than a fixed constant, such as 0.01. Is, an impulse response H is outputm,n+1Step 306 is to get the last recursion result Hm,n+1As input H for the next recursionm,nThe recursive computation is continued.
The other is to fix the recursion number n, recursively calculate equation (5) n times, and then output the impulse response Hm,n+1。
Step 307 is Hm,n+1As the initial value H for the recursive calculation of the next OFDM symbolm+1,0。
Step 308 is the impulse response H to the channelm,n+1An N-point FFT is performed, resulting in channel estimation for all N carriers, where N refers to the number of all carriers in the system.
Step 309 is to select useful data carrier channel estimates from the channel estimates of N carriers and output them, where N refers to the number of all carriers in the system.
The foregoing is a description of one embodiment of the present invention and those skilled in the art will appreciate that various modifications and changes may be made to the embodiment of the present invention without departing from the spirit and scope of the invention as defined in the appended claims.