CN102130871A - Channel estimation method and device - Google Patents
Channel estimation method and device Download PDFInfo
- Publication number
- CN102130871A CN102130871A CN201010022872XA CN201010022872A CN102130871A CN 102130871 A CN102130871 A CN 102130871A CN 201010022872X A CN201010022872X A CN 201010022872XA CN 201010022872 A CN201010022872 A CN 201010022872A CN 102130871 A CN102130871 A CN 102130871A
- Authority
- CN
- China
- Prior art keywords
- channel
- time
- estimation
- interpolation
- filter
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Noise Elimination (AREA)
Abstract
The invention discloses a channel estimation method which comprises the following steps of: performing channel response estimation on a discrete pilot sub-carrier in an orthogonal frequency division multiplexing symbol output after fast Fourier transform to obtain a channel response value of the discrete pilot sub-carrier; performing interpolation estimation in time direction according to the channel response value of the discrete pilot sub-carrier to obtain a time direction channel response value; transforming the time direction channel response value from a frequency domain to a time domain through fast Fourier transform to estimate the maximum multi-path time delay of a channel; and selecting a filter set corresponding to the maximum multi-path time delay from a preset filter group according to the maximum multi-path time delay and performing interpolation estimation in frequency direction according to the channel response value in the time direction to obtain the response value of the entire channel. The invention also provides a channel estimation device. By adopting the channel estimation method and device, the performance requirement on channel estimation can be met, and low complexity of a system can be realized.
Description
Technical Field
The invention relates to a channel estimation method suitable for an orthogonal frequency division multiplexing system and a channel estimation device applied to a receiver.
Background
Orthogonal Frequency Division Multiplexing (OFDM) is a multi-carrier technique whose main principle is to divide a Channel into several Orthogonal Sub-channels (Sub-channels), modulate with one Sub-carrier (Sub-carrier) on each Sub-Channel, and transmit the Sub-carriers in parallel. Thus, although the overall channel is non-flat and frequency selective, each sub-channel is relatively flat, and narrow-band transmission is performed on each sub-channel, the signal bandwidth is smaller than the corresponding bandwidth of the channel, so that mutual interference between sub-channels can be greatly reduced; in addition, because the carriers of each sub-channel are orthogonal to each other in the OFDM system, their frequency spectrums can be overlapped with each other, which not only reduces the mutual interference between the sub-carriers, but also improves the frequency spectrum utilization ratio, so that the OFDM system is widely applied to communication systems such as Digital Video Broadcasting (DVB).
The performance of a wireless communication system is mainly limited by a wireless channel, and a propagation path between a transmitter and a receiver is very complicated, ranging from a simple line of sight to a propagation subject to various complicated topographic features, such as buildings, mountains, forests, and the like. Furthermore, the wireless channel is not fixed and predictable as a wired channel, but has a large randomness, resulting in distortion of the amplitude, phase and frequency of the received signal, which is difficult to analyze. In order to ensure that the communication system has good performance in a wireless channel environment, the time-varying multipath wireless channel must be estimated as accurately as possible, which is particularly difficult in high-speed mobile situations.
Since coherent demodulation has a performance improvement of 2.3dB-3dB compared to non-coherent demodulation, a coherent demodulation technique is generally employed at the receiving end. Coherent demodulation requires information of a channel, and soft decoding of the channel also requires information of the channel, so that good channel estimation is required, and the performance of channel estimation directly affects the demodulation and decoding results.
For the channel estimation method in the OFDM system, the channel estimation method can be roughly divided into two types, namely pilot (pilot) assisted channel estimation and blind channel estimation, and because the pilot-assisted channel estimation method has superior performance, the current channel estimation method with practical value generally uses known pilot information. In the OFDM system, a channel estimation method using pilot-assisted modulation is usually adopted, that is, pilot information is inserted into a transmitted data stream, the pilot information is extracted after a receiving end receives the data stream, channel response at a pilot position is obtained through calculation, and channel response at other positions is estimated by using an interpolation estimation method. However, in the existing OFDM system, the design of the channel estimator has two main problems: firstly, the selection of the pilot frequency information, because of the time-varying characteristic of the wireless channel, the receiver needs to track the channel continuously, so the pilot frequency information must be transmitted continuously; secondly, the optimal channel estimator structure is searched under the condition of determining the pilot frequency sending mode and the channel estimation criterion to meet the requirements of lower complexity and good pilot frequency tracking capability. In practical designs, the selection of the pilot information and the design of the optimal channel estimator are often interrelated, since the performance of the channel estimator is related to the way the pilot information is transmitted. The manner in which the pilot is transmitted determines the method and performance of the channel estimation technique. The two most important parameters that must be considered in pilot information selection are: the maximum doppler shift, which determines the minimum correlation time, and the maximum multipath delay, which determines the minimum correlation bandwidth.
In a digital video terrestrial broadcast (DVB-T/H) system, two pilot modes, namely a continuous pilot mode and a scattered pilot mode, can be used, as shown in fig. 1, the continuous pilot refers to pilot information existing at the same position of each frame, information transmitted on pilot subcarriers is the same for each frame, and specific values are controlled by a pseudo random sequence (PRBS), as shown by K in the figureminAnd KmaxA subcarrier. The continuous pilot frequency is dispersed at a certain regular interval in a plurality of fixed sub-carrier positions, and 45/89/177 continuous pilot frequency sub-carriers exist in the 2k/4k/8k mode respectively. The scattered pilot refers to pilot subcarriers scattered and distributed according to a certain rule, scattered pilot subcarriers of adjacent four OFDM symbols do not repeat, four OFDM symbols are taken as a period, and for the l (l ═ 0 to 67) th symbol in a frame, the position of the scattered pilot belongs to the subset { K ═ K-min+ 3X (lmod4) +12p | pint, 0, K ∈ [ K ] of pmin，Kmax]There are 131/262/524 scattered pilot subcarriers in the 2k/4k/8k mode, respectively. And p is an integer and takes all possible values greater than or equal to zero as long as k does not exceed the effective range. The continuous pilot frequency exists in the fixed sub-carrier position of each OFDM symbol and is mainly used for carrier frequency deviation estimation and tracking, and the scattered pilot frequency has a very regular position and is used for channel estimation of frequency domain interpolation. DVB-T-The pilot arrangement mode in the H system determines that channel estimation can be finished only in a frequency domain, and channel estimation can be carried out according to scattered pilot subcarriers. Channel estimation of the frequency domain pilot can be divided into two steps. Firstly, extracting the channel response value of a pilot frequency point from a received frequency domain signal, and obtaining the channel response value of the pilot frequency position because the transmitted pilot frequency is known, and secondly, obtaining channel estimation on other data subcarriers by interpolation estimation, and then completing the equalization of the signal.
The complexity and performance of channel estimation are a pair of contradictions, and how to design and implement a channel estimator which meets the system performance requirements and has low implementation complexity is a difficult problem. In conventional pilot point channel estimation, the channel estimation is generally based on Least Square (LS) channel estimation algorithm. LS estimation can estimate the channel response of the pilot frequency point only by a single-order equalizer, and the operation amount of the LS estimation only increases linearly along with the number of the pilot frequency points. However, the LS algorithm is greatly affected by white gaussian noise and inter-subcarrier interference, and has poor performance when the signal-to-noise ratio is low, limiting the application of the method. Therefore, a channel estimation algorithm based on Minimum Mean Square Error (MMSE) or linear Minimum Mean Square Error (linemse, LMMSE) is provided, and the method has a good effect of suppressing inter-subcarrier interference and white gaussian noise. However, MMSE or LMMSE estimation utilizes correlation of all N subcarriers, requires a matrix multiplication of N × N, the complexity depends on the number of system carriers, the complexity is high, and the algorithm complexity is exponentially multiplied as the number of operation points increases.
In addition, after the channel response value of the pilot frequency position is obtained, the channel response value of the data subcarrier position can be obtained through time-frequency two-dimensional interpolation filtering. The channel estimation technology based on pilot frequency interpolation is different in complexity, realizability and achievable performance, wherein the best performance is two-dimensional wiener filtering, the channel estimation is performed by using the wiener filter, the performance of the channel estimation technology is related to the order of the wiener filter, the performance is more excellent when the order is larger, but the operation complexity is higher, so the compromise is needed between the performance and the operation complexity. Because the Two-Dimensional wiener filter needs a great operation complexity, a corresponding substitution method is proposed, for example, in the reference of "An Adaptive Two-Dimensional Channel estimation for Wireless OFDM with Application to Mobile DVB-T" (Frieder Sanzi. JoachimSpeidel, IEEE on Broadcasting, VOL.46, NO.2, JUNE 2000), the adopted substitution method is to use the cascade connection of Two one-Dimensional wiener filters to substitute the original Two-Dimensional wiener filter, and the operation complexity is greatly reduced on the premise of not reducing the performance. However, with a wiener filter (whether one-dimensional or two-dimensional), whose filter coefficients are obtained from the autocorrelation function and the cross-correlation function of the channel, the coefficients of the wiener filter are theoretically updated in real time by tracking the change of the channel, but actually calculating the coefficients of the wiener filter in real time is very complicated and the complexity is still relatively high.
Disclosure of Invention
The invention provides a channel estimation method and a channel estimation device applied to a receiver, which can reduce the complexity of a system.
The embodiment of the invention provides a channel estimation method, which comprises the following steps: performing channel response estimation on the scattered pilot frequency sub-carriers in the OFDM symbols output after the fast Fourier transform to obtain channel response values of the scattered pilot frequency sub-carriers; according to the channel response value of the scattered pilot frequency subcarrier, interpolation estimation is carried out in the time direction to obtain a time direction channel response value; transforming the time direction channel response value from the frequency domain to the time domain through inverse fast Fourier transform, and estimating the maximum multipath time delay of the channel; selecting a filter group corresponding to the maximum multipath time delay from a preset filter group according to the maximum multipath time delay; and performing interpolation estimation in the frequency direction by adopting the selected filter group according to the channel response value in the time direction to obtain the response value of the whole channel.
Optionally, the interpolation estimation in the time direction is to perform interpolation by using two adjacent scattered pilot subcarriers in the time direction, and a plurality of OFDM symbols before and after the OFDM symbol are used to complete the interpolation of one OFDM symbol.
Optionally, when the OFDM symbols included in the OFDM interpolation have different timings, removing/applying timing adjustment is performed.
Optionally, the estimating of the maximum multipath delay of the channel specifically includes: performing multiple interpolation and inverse fast Fourier transform on the time direction channel response value to obtain channel impact response in a time domain, and then calculating the average energy of the channel impact response in the inverse fast Fourier transform length; sliding backwards by the length of a guard interval by taking the initial position of the OFDM symbol determined by the fine synchronization of the time sequence as a starting point, and detecting and recording the last point of the sample value energy which is greater than the average energy in the channel impulse response in the interval, wherein the last point is the last path of the multi-path time delay; and subtracting the recorded time corresponding to the last path and the starting position of the OFDM symbol to obtain the maximum multi-path time delay. The multiple interpolation is triple interpolation.
Optionally, after obtaining the maximum multipath delay, selecting a corresponding filter group in the filter group according to the ratio of the maximum multipath delay to the guard interval. The filter group has four filter banks, and each filter bank has three filters respectively.
Another embodiment of the present invention provides a channel estimation apparatus applied to a receiver, the channel estimation apparatus including: the pilot frequency estimation unit is used for carrying out channel response estimation on the scattered pilot frequency sub-carriers in the OFDM symbols output after the fast Fourier transform to obtain a channel response value of the scattered pilot frequency sub-carriers; the time domain interpolation filter is used for carrying out interpolation estimation in the time direction according to the channel response value of the scattered pilot frequency subcarrier to obtain a channel response value in the time direction; the time delay estimation unit is used for transforming the channel response value in the time direction from the frequency domain to the time domain through inverse fast Fourier transform and estimating the maximum multipath time delay of the channel; and the frequency domain interpolation filter group is provided with a plurality of filter banks, each filter bank is provided with a plurality of filters and is used for selecting the filter bank corresponding to the maximum multipath time delay according to the maximum multipath time delay, and then carrying out interpolation estimation in the frequency direction according to the channel response value in the time direction to obtain the response value of the whole channel.
Optionally, the interpolation estimation in the time direction is to perform interpolation by using two adjacent scattered pilot subcarriers in the time direction, and a plurality of OFDM symbols before and after the OFDM symbol are used to complete the interpolation of one OFDM symbol.
Optionally, when the OFDM symbols included in the OFDM interpolation have different timings, removing/applying timing adjustment is performed.
Optionally, the pilot estimation unit is a least squares estimator, and the least squares estimator is a single-order equalizer.
Optionally, the estimating, by the delay estimation unit, the maximum multipath delay specifically includes: performing multiple interpolation and inverse fast Fourier transform on the time direction channel response value to obtain channel impact response in a time domain, and then calculating the average energy of the channel impact response in the inverse fast Fourier transform length; sliding backwards by the length of a guard interval by taking the initial position of the OFDM symbol determined by the fine synchronization of the time sequence as a starting point, and detecting and recording the last point of the sample value energy which is greater than the average energy in the channel impulse response in the interval, wherein the last point is the last path of the multi-path time delay; and subtracting the recorded last path from the time corresponding to the starting position of the OFDM symbol to obtain the maximum multi-path time delay. The multiple interpolation is triple interpolation.
Optionally, after obtaining the maximum multipath delay, selecting a corresponding filter group in the filter group according to the ratio of the maximum multipath delay to the guard interval. The frequency domain interpolation filter group has four filter banks, and each filter bank has three filters respectively.
Compared with the prior art, the technical scheme is that a filter bank with a self-adaptive filter coefficient is provided, a time direction channel response value is obtained by adopting linear interpolation estimation in the time direction through a discrete pilot frequency subcarrier channel response value obtained based on channel response estimation, then the maximum multipath time delay is estimated in the time domain, a filter corresponding to the maximum multipath time delay is selected in a self-adaptive mode, interpolation estimation is carried out in the frequency direction, and finally the response value of the whole channel is obtained.
Drawings
Fig. 1 is a diagram illustrating a frame structure of a general OFDM symbol having pilot subcarriers;
FIG. 2 is a basic flow chart of one embodiment of the channel estimation method of the present invention;
FIG. 3 is a schematic diagram of a frame structure for interpolation estimation in the time direction according to an embodiment of the present invention;
FIG. 4 is a diagram illustrating a frame structure for interpolation estimation in the frequency direction according to an embodiment of the present invention;
fig. 5 is a detailed flow step of the method for estimating the maximum multipath delay in step S25 shown in fig. 2;
fig. 6 is a schematic structural diagram of an embodiment of the channel estimation device of the present invention.
Detailed Description
The technical scheme is applied to a receiving system of a ground digital television broadcasting (DVB) standard, in particular to a DVB-H standard, and mainly comprises the steps of obtaining a time direction channel response value by adopting linear interpolation estimation in a time direction based on a discrete pilot frequency subcarrier channel response value obtained in advance by channel response estimation, estimating the maximum multipath time delay in a time domain, adaptively selecting a filter corresponding to the maximum multipath time delay and carrying out interpolation estimation in a frequency direction to obtain a response value of the whole channel. The technical solution of the present invention is described in detail below with reference to the accompanying drawings and examples.
The channel estimation method in the embodiment of the present invention is shown in fig. 2, where a transmission signal on the channel is modulated by an OFDM technique, the OFDM symbol includes data subcarriers and pilot subcarriers, and the pilot subcarriers include continuous pilot subcarriers and scattered pilot subcarriers. Before implementing channel estimation, the receiver performs synchronization processing such as carrier synchronization, symbol synchronization, clock synchronization and the like after receiving signals, so that the receiving end obtains accurate synchronization. In addition, Fast Fourier Transform (FFT) is performed on the received signal, and the FFT transforms the received signal from the time domain to the frequency domain, and has three transmission modes of 2k, 4k, and 8k, where k is 1024. As shown in fig. 2, the channel estimation method first proceeds to step S21.
In step S21, channel response estimation is performed on the scattered pilot subcarriers in the OFDM symbol output after the FFT, and channel response values of the scattered pilot subcarriers are obtained. In this embodiment, the length of the FFT includes, for example, 2k, 4k, or 8 k. Referring to fig. 1, as shown in fig. 1, in the OFDM system, each frame includes a plurality of OFDM symbols (e.g. 0 th, 1 th, 2.. 66 th, 67 th OFDM symbols), wherein each OFDM symbol further includes, for example, 1705(2k mode), 3409(4k mode), or 6817(8k mode) subcarriers, and the inserted scattered pilot subcarriers are according to a certain ruleAnd (4) discrete distribution. As shown in fig. 1, the interval N of the scattered pilot subcarriers in the time directiont4 and an interval in the frequency direction of Nf12. In this embodiment, the channel response estimation is a least squares estimation, which specifically includes: assume that the original received signal is
Where x (n) is an original signal sent by the sending end, h (n) is a channel impulse response, and w (n) is channel noise, in this embodiment, w (n) is Additive White Gaussian Noise (AWGN). Thus, after FFT, the received signal is transformed from the time domain to the frequency domain:
according to the OFDM model shown in fig. 1, the frequency domain signal of the scattered pilot subcarrier on the kth subcarrier of the ith OFDM symbol in a frame of OFDM is: y isl，k＝Xl，k□Hl，k+Wl，kDue to the pilot value X of the scattered pilot sub-carrierl，kAre known, and thus the channel response values of the scattered pilot subcarriers can be obtained
It should be noted that, the above channel response estimation is described by taking LS estimation as an example, because the operation amount of LS estimation only increases linearly with the number of pilot points, and the circuit design is relatively simple, but the LS estimation in this embodiment is not intended to limit the present invention, and the channel response estimation may also be other channel estimation such as MMSE or LMMSE.
In step S23, interpolation estimation is performed in the time direction according to the channel response value of the scattered pilot subcarriers to obtain a time direction channelThe response value. The interpolation estimation in the time direction is performed by using a linear interpolation estimation method. Specifically, referring to fig. 3, the channel response values of the scattered pilot subcarriers obtained in step S21 are shownInterpolation estimation is carried out in the time direction, and the estimation in the time direction is obtained as follows:l≤m≤Nt-1, where the scattered pilot spacing N in the time directiont＝4；Representing the channel estimates of the scattered pilot symbols at time/frequency k;
represents time l + NtChannel estimation of scattered pilot symbols at frequency k locations. In this embodiment, according to the frame structure of OFDM, for each OFDM symbol, the number of subcarriers (including data subcarriers and pilot subcarriers) having channel response values obtained by performing linear interpolation on discrete pilot subcarriers in the time direction is 131 × 12/3 (524 (2k mode), 262 × 12/3 (1048 (4k mode), 524 × 12/3 (2096 (8k mode), so that the channel response values obtained by linear interpolation can be rearranged and arranged to HTl，j(0. ltoreq. l.ltoreq.67, 0. ltoreq. j.ltoreq. 523/1047/2095). The process then proceeds to step S25.
In addition, as shown in fig. 3, in the present embodiment, the interval N in the time direction due to the inserted scattered pilot subcarriers istSpecifically, the positions of the scattered pilot subcarriers of the 0 th OFDM symbol and the 4 th OFDM symbol are the same, and the positions of the scattered pilot subcarriers of the 1 st OFDM symbol and the 5 th OFDM symbol are the same, as seen from the time direction, thereby performing interpolation estimation in the time directionBy analogy, it can be seen that every 4 OFDM symbols, one scattered pilot subcarrier appears at the same position, such as P12 and P11, P22 and P21 in the figure. Assuming that the OFDM symbol of reference numeral 3 is currently subjected to channel estimation in the time direction, two scattered pilot subcarriers P11 and P12 are required for point P1, and two scattered pilot subcarriers P22 and P21 are required for point P2. Wherein, point P12 is located on the OFDM symbol of reference numeral 0, point P11 is located on the OFDM symbol of reference numeral 4, point P21 is located on the OFDM symbol of reference numeral 6, point P22 is located on the OFDM symbol of reference numeral 2, and the information of the two scattered pilot subcarriers used by point P3 are located on the OFDM symbols of reference numerals 1 and 5, respectively. It can be seen that the estimation of the channel response of the OFDM symbol at reference numeral 3 in the time direction by using the scattered pilot subcarriers uses the scattered pilot subcarriers in the preceding 3 OFDM symbols and the following 3 OFDM symbols, so that including the estimated OFDM symbol at reference numeral 3, the complete interpolation estimation of the OFDM symbol requires the scattered pilot subcarrier information of 7 OFDM symbols in total.
It should be noted that the channel estimation method according to the embodiment of the present invention is performed on the premise of obtaining accurate synchronization, and in fact, the embodiment utilizes the processing result of timing fine synchronization. In the fine timing synchronization process, because timing adjustment needs to be performed once per frame during system design, all OFDM symbols in a frame have the same timing, and different frames have different timings. Therefore, when the interpolation estimation in the time direction is performed in step S23, if 7 OFDM symbols used before and after the time domain interpolation belong to different frames and have different timings, it is necessary to perform removal/application of timing adjustment on the respective OFDM symbols so that the timings of the respective OFDM symbols concerned coincide. Specifically, it is assumed that point P1 on OFDM symbol No. 3 is to be estimated in the time direction, and OFDM symbol No. 0 where pilot subcarrier P12 participates in estimation with respect to point P1 and symbol No. 4 where pilot subcarrier P11 belongs to two frames with different timings, respectively, P12 has been subjected to timing adjustment, and P11 has not been subjected to timing adjustment. If the OFDM symbol No. 3 at the point P1 and the OFDM symbol No. 0 at the point P12 belong to the same frame, that is, the point P1 has also undergone timing adjustment, then the pilot subcarrier P11 that participates in estimation and has not undergone timing adjustment must exert the influence of timing adjustment; on the contrary, if the OFDM symbol denoted by P1 and the OFDM symbol denoted by P11 belong to the same frame, i.e. the timing adjustment has not been performed at P1, the pilot subcarrier P12 that is already subjected to timing adjustment and is involved in estimation must remove the influence of timing adjustment. Therefore, the subcarriers in each OFDM symbol are kept consistent, and the accuracy of interpolation estimation is ensured.
In step S25, the time direction channel response value obtained by the interpolation estimation in the time direction is transformed from the frequency domain to the time domain by IFFT transformation, and the maximum multipath delay τ of the channel is estimatedmax. In this embodiment, the IFFT length is 2k, 4k, or 8k, where k is 1024. The maximum multipath time delay taumaxThe time difference between the first path and the last path detected within the IFFT length is obtained, and a specific estimation method thereof can be seen in fig. 5. The process then proceeds to step S27.
In step S27, the maximum multipath delay τ is obtainedmaxSelecting the maximum multipath time delay tau from a preset filter groupmaxA corresponding filter bank. The filter group has a plurality of filter banks therein, and each filter bank has a plurality of filters therein. The filter is for example a low pass filter and has corresponding filter coefficients. In this embodiment, the range of possible maximum multipath delays may be divided into four delay sections, each corresponding to a set of filter banks, so that the filter groups have four sets of filter banks corresponding to the estimated maximum multipath delays τmax. Specifically, the maximum multipath delay range and the guard interval are divided into four delay sections 0 to 0.25, 0.25 to 0.5, 0.5 to 0.75, and 0.75 to 1.0, respectively, and each delay section corresponds to a filter bank 0, 1, 2, and 3, respectively. The guard interval is related to the FFT/IFFT length, which may be 2K, 4K or 8K, and the guard intervalThe length may be 1/32, 1/16, 1/8, 1/4 of the FFT/IFFT length. That is, when the maximum multipath delay τ is obtained in step S23maxThen, the maximum multipath delay tau can be usedmaxThe ratio of the maximum multipath time delay tau to the guard interval is judgedmaxThe filter group is selected corresponding to the time delay section according to the judgment result, wherein the bandwidth of the selected filter is slightly larger than the maximum multipath time delay taumax. Meanwhile, each filtering operation should obtain three values, and each value needs one filter, so each filter bank has three filters, and thus the whole filter bank has 12 filters, in this embodiment, each filter in the filter bank is a 12-order filter. Therefore, through the filter bank with the adaptive filter coefficient, the corresponding filter can be adaptively selected according to the maximum multipath time delay obtained by real-time detection and estimation, and the problems that in the prior art, a wiener filter (whether a two-dimensional wiener filter or a one-dimensional wiener filter) is adopted to acquire the relevant characteristics of a channel, the circuit is complex to realize and the estimation performance is unstable, or the circuit is complex due to real-time tracking of the change of the channel and the like can be relatively overcome. The process then proceeds to step S29.
In step S29, according to the time direction channel response value, interpolation estimation is performed in the frequency direction by using the selected corresponding filter bank, so as to obtain a response value of the entire channel. Specifically, referring to fig. 4, as shown in the figure, interpolation estimation is performed in the frequency direction according to the filter coefficients COE1(i), COE2(i) and COE3(i) corresponding to the selected three 12-step filters, where i is 0 to 11, and the estimation in the frequency direction is obtained as follows:
wherein,indicating the channel estimation of the pilot sub-carrier symbols in time/frequency k position in the frequency direction, HTl，jIs the result of channel estimation in the previous time direction. In this way, by two-dimensional interpolation estimation in the time direction and the frequency direction, the channel response values of all positions (including data subcarriers and pilot subcarriers), that is, the entire channel, can be accurately obtained. Compared with the prior art that interpolation estimation is carried out by utilizing a two-dimensional wiener filter or two one-dimensional wiener filters which are cascaded, the method has the advantage that the defect of high algorithm complexity is easily caused.
The multi-path delay described in step S25 is due to the fact that the transmission paths experienced by the path components of the signal during transmission are different, and therefore have different time delays. The maximum multipath time delay is the time difference between the signal component reaching the receiving end firstly and the signal component reaching the receiving end last, the accurate estimation of the maximum multipath time delay can provide reliable judgment basis for a filter selecting proper filter coefficient subsequently. The maximum multipath delay provided by this embodiment is estimated as shown in fig. 5. As shown in fig. 5, the estimation first proceeds to step S51.
In step S51, the time-direction channel response value is subjected to multiple times of interpolation and inverse fast fourier transform to obtain the channel impulse response in the time domain, and then the average energy of the channel impulse response in the inverse fast fourier transform length is calculated. Actually, before performing the inverse fast fourier transform, it is necessary to perform multiple interpolation on the time direction channel response value, in this embodiment, three times of interpolation is adopted according to the distribution position of the scattered pilot subcarriers. As shown in fig. 3, since the time direction channel response value obtained in step S23 is the result of interpolation estimation in the time direction at the position of the scattered pilot subcarrier in the OFDM symbol, the frame structure of the OFDM symbol shows that the interval of the scattered pilot subcarriers in the frequency direction is three, i.e. the scattered pilot subcarriers do not appear at every three frequency points, as shown in fig. 3, K is 3, 6, and 9. The average energy is obtained by adding and averaging sample energies of channel impulse responses within an IFFT length, where the IFFT length corresponds to the FFT length and includes 2k, 4k, or 8k, where k is 1024. Step S53 is then performed.
In step S53, the starting position determined by the fine timing synchronization is used as the starting point, and the length of a guard interval is slid backwards, and the last point in the channel impulse response, where the sample energy is greater than the average energy, is detected and recorded, where the last point is the last path of the multi-path delay. Wherein, the starting position determined by the time sequence fine synchronization is the first path of the multipath time delay. Step S55 is then performed.
In step S55, the recorded last path is subtracted from the time corresponding to the start position determined by the timing synchronization to obtain the maximum multi-path delay τmax. The maximum multipath time delay taumaxIs less than the guard interval GLIn (1). The invention obtains the maximum multi-path time delay according to the interpolation estimation result in the time direction on the premise of precise time sequence synchronization, can select a proper filter corresponding to the filter coefficient for the follow-up, has relatively simple and easy realization method, and effectively reduces the complexity of the system.
In accordance with the above-mentioned channel estimation method, another aspect of the present invention provides a channel estimation apparatus applied to a receiver, as shown in fig. 6, the OFDM symbol received by the receiver in the embodiment of the present invention includes a synchronization unit 100, a fast fourier transform unit 120, and an inverse fast fourier transform unit 140, the synchronization unit 100 is used for performing synchronization processing including carrier synchronization, symbol synchronization, clock synchronization, and the like, the FFT unit 120 is used for transforming a signal from a time domain signal to a frequency domain signal, and the IFFT unit 140 is used for transforming a signal from a frequency domain signal to a time domain signal. The channel estimation device comprises: pilot frequency estimation unit 61, time domain interpolation filter 62, time delay estimation unit 63, frequency domain interpolation filter group 64.
The time domain interpolation filter 62 is used for performing interpolation estimation in the time direction according to the channel response value of the scattered pilot subcarrier obtained by the pilot estimation unit 61 to obtain a time direction channel response value. The time-domain interpolation filter 62 is, for example, a low-pass filter, and can perform linear interpolation in the time direction.
The delay estimation unit 63 is configured to transform the time-domain channel response value output by the IFFT unit 140 from the frequency domain to the time domain by inverse fast fourier transform, and estimate the maximum multipath delay of the channel. The specific estimation method comprises the following steps: performing multiple interpolation (in this embodiment, three times of interpolation is adopted according to the distribution position of the scattered pilot subcarriers) and inverse fast fourier transform on the time direction channel response value to obtain a channel impulse response in a time domain, and then calculating the average energy of the channel impulse response in the inverse fast fourier transform length; taking the initial position determined by the time sequence fine synchronization as a starting point, sliding backwards by the length of a guard interval, and detecting and recording the last point of the sample value energy which is greater than the average energy in the channel impact response in the interval, wherein the last point is the last path of the multi-path time delay; and subtracting the recorded last path from the time corresponding to the initial position determined by the time sequence fine synchronization to obtain the maximum multi-path time delay.
And a frequency domain interpolation filter group 64 having a plurality of filter banks, and each filter bank having a plurality of filters, configured to select a filter having a filter coefficient corresponding to the maximum multipath delay according to the maximum multipath delay, and further perform interpolation estimation in the frequency direction according to the channel response value obtained by performing interpolation estimation in the time direction, so as to obtain a response value of the entire channel. The filter is for example a low pass filter and has corresponding filter coefficients. In the present embodiment, the frequency-domain interpolation filter group 64 has four filter banks 640, 641, 642, 643, which respectively correspond to the maximum multipath delays of four different delay sections. In addition, since three values are obtained for each filtering according to the distribution position of the scattered pilot subcarriers in this embodiment, each filter bank has three filters, i.e. the filter bank 640 has filters 640a, 640b, and 640c, the filter bank 641 has filters 641a, 641b, and 641c, and so on, and the frequency domain interpolation filter group 64 has 12 filters in total. Thus, based on the estimated maximum multipath delay τmaxIs determined, the maximum multipath delay tau is determinedmaxFalls within the range of which time delay section is divided, so that the filter set corresponding to the time delay section can be selected, wherein the bandwidth of the selected filter set is slightly larger than the obtained maximum multipath time delay taumax. Since the frequency domain interpolation filter group 64 has the filter banks 640, 641, 642, 643 with different filter coefficients, the corresponding filter bank can be adaptively selected according to the maximum multipath delay estimated by real-time detection, and the problems of high system complexity caused by the adoption of a wiener filter (no matter a two-dimensional wiener filter or a one-dimensional wiener filter) or circuit complexity caused by real-time tracking of channel changes in the prior art can be relatively overcome.
In summary, the channel estimation method and the channel estimation device of the present invention perform channel response estimation on the scattered pilot subcarriers in advance to obtain channel response values of the scattered pilot subcarriers, perform interpolation estimation in the time direction to obtain time direction channel response values, estimate the maximum multipath delay through inverse fast fourier transform, adaptively select a filter corresponding to the maximum multipath delay and perform interpolation estimation in the frequency direction to obtain response values of the entire channel, which not only can meet performance requirements of channel estimation, but also can achieve low complexity of the system. The method can overcome the defects that the interpolation estimation is carried out by using a two-dimensional wiener filter or two one-dimensional wiener filters which are cascaded in the prior art, the complexity of the system is easily increased, or the circuit is easily complicated when the change of a channel is tracked in real time, and the like. And the accurate maximum channel time delay can be obtained by utilizing the result of the precise time sequence synchronization.
In addition, a simpler LS estimation method is adopted during interpolation estimation in the embodiment of the invention, and compared with MMSE or LMMSE channel estimation, the calculation amount is smaller and the system complexity is lower; the filter selected during subsequent frequency domain interpolation can filter out the out-of-band part of noise added by LS estimation, and effectively makes up for the deficiency of LS estimation performance.
Although the present invention has been described with reference to the preferred embodiments, it is not intended to be limited thereto, and variations and modifications may be made by those skilled in the art without departing from the spirit and scope of the present invention.
Claims (17)
1. A channel estimation method, characterized in that said channel estimation method comprises the steps of:
performing channel response estimation on the scattered pilot frequency sub-carriers in the orthogonal frequency division multiplexing symbols output after the fast Fourier transform to obtain channel response values of the scattered pilot frequency sub-carriers;
performing interpolation estimation in the time direction according to the channel response value of the scattered pilot frequency subcarrier to obtain a time direction channel response value;
transforming the time direction channel response value from the frequency domain to the time domain through inverse fast Fourier transform, and estimating the maximum multipath time delay of the channel;
selecting a filter group corresponding to the maximum multipath time delay from a preset filter group according to the maximum multipath time delay;
and performing interpolation estimation in the frequency direction by adopting the selected filter bank according to the channel response value in the time direction to obtain the response value of the whole channel.
2. The channel estimation method of claim 1, wherein the interpolation estimation in the time direction is performed by using two adjacent scattered pilot subcarriers in the time direction, and a plurality of OFDM symbols before and after an OFDM symbol are used to perform the interpolation of one OFDM symbol.
3. The channel estimation method according to claim 2, wherein when the OFDM symbols included in the completion of the OFDM interpolation have different timings, the removal/application of the timing adjustment is performed.
4. The channel estimation method according to claim 1, wherein the estimating of the maximum multipath delay of the channel specifically comprises:
performing multiple interpolation and inverse fast Fourier transform on the time direction channel response value to obtain channel impact response in a time domain, and then calculating the average energy of the channel impact response in the inverse fast Fourier transform length;
taking the initial position determined by the time sequence fine synchronization as a starting point, sliding backwards by the length of a guard interval, and detecting and recording the last point of the sample value energy which is greater than the average energy in the channel impact response in the interval, wherein the last point is the last path of the multi-path time delay;
and subtracting the recorded last path from the time corresponding to the initial position determined by the time sequence fine synchronization to obtain the maximum multi-path time delay.
5. The channel estimation method according to claim 4, wherein the multiple interpolation is triple interpolation.
6. The channel estimation method according to claim 4, wherein after obtaining the maximum multipath delay, selecting a corresponding filter set from the filter set according to a ratio of the maximum multipath delay to a guard interval.
7. The channel estimation method according to claim 6, wherein the filter group has four filter banks, and each filter bank has three filters.
8. A channel estimation apparatus applied to a receiver, the channel estimation apparatus comprising:
a pilot frequency estimation unit, configured to perform channel response estimation on a scattered pilot frequency subcarrier in an orthogonal frequency division multiplexing symbol output after fast fourier transform, to obtain a channel response value of the scattered pilot frequency subcarrier;
the time domain interpolation filter is used for carrying out interpolation estimation in the time direction according to the channel response value of the scattered pilot frequency subcarrier to obtain a channel response value in the time direction;
the time delay estimation unit is used for transforming the channel response value in the time direction from the frequency domain to the time domain through inverse fast Fourier transform and estimating the maximum multipath time delay of the channel;
and the frequency domain interpolation filter group is provided with a plurality of filter banks, each filter bank is provided with a plurality of filters and is used for selecting the filter bank with the filtering corresponding to the maximum multipath time delay according to the maximum multipath time delay, and then carrying out interpolation estimation in the frequency direction according to the channel response value in the time direction to obtain the response value of the whole channel.
9. The channel estimation device of claim 8, wherein the interpolation estimation in the time direction is performed by using two adjacent scattered pilot subcarriers in the time direction, and a plurality of OFDM symbols before and after the OFDM symbol are used to perform the interpolation of one OFDM symbol.
10. The channel estimation apparatus as claimed in claim 9, wherein the removal/application timing adjustment is performed when the OFDM symbols included in the completion of the OFDM interpolation have different timings.
11. The channel estimation device of claim 8 wherein the pilot estimation unit is a least squares estimator.
12. The channel estimation apparatus of claim 11, wherein the least squares estimator is a single order equalizer.
13. The channel estimation device of claim 8, wherein the time-domain interpolation employs linear interpolation.
14. The channel estimation apparatus according to claim 8, wherein the estimating of the maximum multipath delay by the delay estimation unit specifically comprises:
performing multiple interpolation and inverse fast Fourier transform on the time direction channel response value to obtain channel impact response in a time domain, and then calculating the average energy of the channel impact response in the inverse fast Fourier transform length;
taking the initial position determined by the time sequence fine synchronization as a starting point, sliding backwards by the length of a guard interval, and detecting and recording the last point of the sample value energy which is greater than the average energy in the channel impact response in the interval, wherein the last point is the last path of the multi-path time delay;
and subtracting the recorded last path from the time corresponding to the initial position determined by the time sequence fine synchronization to obtain the maximum multi-path time delay.
15. The channel estimation apparatus of claim 14, wherein the multiple-time interpolation is triple-time interpolation.
16. The apparatus according to claim 14, wherein after obtaining the maximum multipath delay, a corresponding filter bank in the filter bank is selected according to a ratio of the maximum multipath delay to a guard interval.
17. The apparatus of claim 16, wherein the frequency-domain interpolation filter group has four filter banks, and each filter bank has three filters.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201010022872XA CN102130871A (en) | 2010-01-15 | 2010-01-15 | Channel estimation method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201010022872XA CN102130871A (en) | 2010-01-15 | 2010-01-15 | Channel estimation method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN102130871A true CN102130871A (en) | 2011-07-20 |
Family
ID=44268763
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201010022872XA Pending CN102130871A (en) | 2010-01-15 | 2010-01-15 | Channel estimation method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102130871A (en) |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103501284A (en) * | 2013-10-22 | 2014-01-08 | 湖南国科微电子有限公司 | Method of synchronizing precise timing of DVB-T2 (Digital Video Broadcasting-Terrestrial 2) system |
| CN103905350A (en) * | 2014-04-08 | 2014-07-02 | 电子科技大学 | S-IFFT-based channel estimation method for OFDM system |
| CN104348765A (en) * | 2013-08-05 | 2015-02-11 | 普天信息技术研究院有限公司 | Channel estimation method |
| CN104883327A (en) * | 2014-11-21 | 2015-09-02 | 广东省电信规划设计院有限公司 | Channel estimation method and system based on two-dimensional interpolation |
| CN107733830A (en) * | 2016-08-12 | 2018-02-23 | 中兴通讯股份有限公司 | Method, apparatus and system caused by a kind of multi-carrier signal |
| CN108234364A (en) * | 2018-01-18 | 2018-06-29 | 重庆邮电大学 | Channel estimation methods based on cell reference signals in a kind of lte-a system |
| CN109495409A (en) * | 2018-12-29 | 2019-03-19 | 京信通信系统（中国）有限公司 | Channel estimation methods, device, radio receiver and storage medium |
| CN111600816A (en) * | 2020-05-12 | 2020-08-28 | 国网河北省电力有限公司电力科学研究院 | Two-dimensional interpolation channel estimation method for power line carrier communication |
| CN112333643A (en) * | 2020-11-16 | 2021-02-05 | 武汉大学 | Wireless positioning method, system, medium and intelligent terminal of 5G downlink signal |
| CN112399550A (en) * | 2020-08-25 | 2021-02-23 | 中兴通讯股份有限公司 | Method, device and equipment for determining time delay compensation value and storage medium |
| CN109525521B (en) * | 2018-11-22 | 2021-03-23 | 南京铁道职业技术学院 | Channel estimation method |
| CN112866150A (en) * | 2021-01-13 | 2021-05-28 | 上海复旦通讯股份有限公司 | Wireless channel feature extraction and dimension reduction method and system |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101127745A (en) * | 2006-08-16 | 2008-02-20 | 大唐移动通信设备有限公司 | A chancel estimation method and device |
| CN101247376A (en) * | 2007-02-15 | 2008-08-20 | 北京三星通信技术研究有限公司 | OFDM channel estimation method for eliminating noise combined with wavelet transformed domain |
| CN101478507A (en) * | 2008-12-10 | 2009-07-08 | 北京创毅视讯科技有限公司 | Channel estimation method and terminal |
-
2010
- 2010-01-15 CN CN201010022872XA patent/CN102130871A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101127745A (en) * | 2006-08-16 | 2008-02-20 | 大唐移动通信设备有限公司 | A chancel estimation method and device |
| CN101247376A (en) * | 2007-02-15 | 2008-08-20 | 北京三星通信技术研究有限公司 | OFDM channel estimation method for eliminating noise combined with wavelet transformed domain |
| CN101478507A (en) * | 2008-12-10 | 2009-07-08 | 北京创毅视讯科技有限公司 | Channel estimation method and terminal |
Non-Patent Citations (1)
| Title |
|---|
| 王勇等: "低复杂度无线OFDM自适应信道估计算法", 《系统仿真学报》 * |
Cited By (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104348765A (en) * | 2013-08-05 | 2015-02-11 | 普天信息技术研究院有限公司 | Channel estimation method |
| CN104348765B (en) * | 2013-08-05 | 2017-08-29 | 普天信息技术研究院有限公司 | Channel estimation methods |
| CN103501284B (en) * | 2013-10-22 | 2016-09-28 | 湖南国科微电子股份有限公司 | A kind of method of fine timing synchronization in DVB-T2 system |
| CN103501284A (en) * | 2013-10-22 | 2014-01-08 | 湖南国科微电子有限公司 | Method of synchronizing precise timing of DVB-T2 (Digital Video Broadcasting-Terrestrial 2) system |
| CN103905350A (en) * | 2014-04-08 | 2014-07-02 | 电子科技大学 | S-IFFT-based channel estimation method for OFDM system |
| CN103905350B (en) * | 2014-04-08 | 2017-01-04 | 电子科技大学 | Channel estimation methods based on S IFFT in a kind of ofdm system |
| CN104883327A (en) * | 2014-11-21 | 2015-09-02 | 广东省电信规划设计院有限公司 | Channel estimation method and system based on two-dimensional interpolation |
| CN104883327B (en) * | 2014-11-21 | 2019-01-08 | 广东省电信规划设计院有限公司 | Channel estimation methods and system based on two-dimensional interpolation |
| CN107733830A (en) * | 2016-08-12 | 2018-02-23 | 中兴通讯股份有限公司 | Method, apparatus and system caused by a kind of multi-carrier signal |
| CN108234364B (en) * | 2018-01-18 | 2020-10-27 | 重庆邮电大学 | A kind of channel estimation method based on cell reference signal in LTE-A system |
| CN108234364A (en) * | 2018-01-18 | 2018-06-29 | 重庆邮电大学 | Channel estimation methods based on cell reference signals in a kind of lte-a system |
| CN109525521B (en) * | 2018-11-22 | 2021-03-23 | 南京铁道职业技术学院 | Channel estimation method |
| CN109495409A (en) * | 2018-12-29 | 2019-03-19 | 京信通信系统（中国）有限公司 | Channel estimation methods, device, radio receiver and storage medium |
| CN109495409B (en) * | 2018-12-29 | 2021-06-25 | 京信通信系统（中国）有限公司 | Channel estimation method, device, wireless receiving equipment and storage medium |
| CN111600816A (en) * | 2020-05-12 | 2020-08-28 | 国网河北省电力有限公司电力科学研究院 | Two-dimensional interpolation channel estimation method for power line carrier communication |
| CN112399550A (en) * | 2020-08-25 | 2021-02-23 | 中兴通讯股份有限公司 | Method, device and equipment for determining time delay compensation value and storage medium |
| WO2022042333A1 (en) * | 2020-08-25 | 2022-03-03 | 中兴通讯股份有限公司 | Method and apparatus for determining time delay compensation value, and device and storage medium |
| CN112333643A (en) * | 2020-11-16 | 2021-02-05 | 武汉大学 | Wireless positioning method, system, medium and intelligent terminal of 5G downlink signal |
| CN112866150A (en) * | 2021-01-13 | 2021-05-28 | 上海复旦通讯股份有限公司 | Wireless channel feature extraction and dimension reduction method and system |
| CN112866150B (en) * | 2021-01-13 | 2024-05-17 | 上海复旦通讯股份有限公司 | Wireless channel feature extraction and dimension reduction method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102130871A (en) | Channel estimation method and device | |
| EP2200238B1 (en) | Methods and systems to estimate channel frequency response in multi-carrier signals | |
| US20070297522A1 (en) | Method for Signal Processing and a Signal Processor in an Ofdm System | |
| CN101378371A (en) | Method for channel estimation of wideband wireless mobile commutation system and channel estimator | |
| JP2008501272A (en) | Channel estimation in OFDM systems with high Doppler shift | |
| WO2006134578A1 (en) | Receiver apparatus for receiving a multicarrier signal | |
| US8396144B2 (en) | Method and system for OFDM symbol timing recovery | |
| CN107483373B (en) | Anti-multipath iterative weighting LMMSE channel estimation method and device | |
| CN101345726A (en) | Channel estimation method for reducing storage space | |
| CN101552758B (en) | A Fine Symbol Timing Synchronization Method in OFDM Systems | |
| Yeh et al. | Efficient channel estimation based on discrete cosine transform | |
| KR101390317B1 (en) | Apparatus and method for compensation of channel impulse response estimation error in orthogonal frequency division multiplexing systems | |
| James et al. | Channel estimation for OFDM systems | |
| KR100656384B1 (en) | Channel Estimation Method and Apparatus Using Linear Prediction in Open FM Communication System with Virtual Subcarrier | |
| CN1984109A (en) | Channel estimater and channel estimating method in telecommunication system | |
| CN102487364B (en) | Channel estimation method and apparatus thereof | |
| CN103428126A (en) | OFDM self-adaptive channel estimation method based on comb-type pilot frequency | |
| KR20000066459A (en) | Apparatus for channel estimating of ofdm) | |
| KR100849493B1 (en) | Channel Estimation System and Method Using Linear or Nonlinear Average in Wireless Communication System | |
| Yao | Research on denoise methods of channel estimation in ofdm system with high-speed multipath channels | |
| Yücek | Self-interference handling in OFDM based wireless communication systems | |
| Husen et al. | Simple Doppler compensation for DVB-T | |
| Wu | On comparison of pilot-aided ofdm channel estimation algorithms | |
| Lee et al. | Channel estimation based on path separation for DVB-T in long delay situations | |
| Jang et al. | DFT-based decision directed channel estimation for OFDM systems in very large delay spread channels |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20110720 |