[go: up one dir, main page]

CN107979454B - Method for rapidly extracting OFDMA distributed mode downlink comb spectrum - Google Patents

Method for rapidly extracting OFDMA distributed mode downlink comb spectrum Download PDF

Info

Publication number
CN107979454B
CN107979454B CN201711170826.2A CN201711170826A CN107979454B CN 107979454 B CN107979454 B CN 107979454B CN 201711170826 A CN201711170826 A CN 201711170826A CN 107979454 B CN107979454 B CN 107979454B
Authority
CN
China
Prior art keywords
block
split
butterfly
stage
data
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.)
Active
Application number
CN201711170826.2A
Other languages
Chinese (zh)
Other versions
CN107979454A (en
Inventor
余华
裴欣月
邹游
季飞
陈芳炯
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
South China University of Technology SCUT
Original Assignee
South China University of Technology SCUT
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by South China University of Technology SCUT filed Critical South China University of Technology SCUT
Priority to CN201711170826.2A priority Critical patent/CN107979454B/en
Publication of CN107979454A publication Critical patent/CN107979454A/en
Application granted granted Critical
Publication of CN107979454B publication Critical patent/CN107979454B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L5/00—Arrangements affording multiple use of the transmission path
    • H04L5/0001—Arrangements for dividing the transmission path
    • H04L5/0003—Two-dimensional division
    • H04L5/0005—Time-frequency
    • H04L5/0007—Time-frequency the frequencies being orthogonal, e.g. OFDM(A) or DMT
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L5/00—Arrangements affording multiple use of the transmission path
    • H04L5/003—Arrangements for allocating sub-channels of the transmission path
    • H04L5/0037—Inter-user or inter-terminal allocation
    • H04L5/0041—Frequency-non-contiguous
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L5/00—Arrangements affording multiple use of the transmission path
    • H04L5/0091—Signalling for the administration of the divided path, e.g. signalling of configuration information
    • H04L5/0094—Indication of how sub-channels of the path are allocated

Landscapes

  • Engineering & Computer Science (AREA)
  • Signal Processing (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Complex Calculations (AREA)

Abstract

本发明提供一种OFDMA分布式模式下行链路梳状谱快速提取方法。本发明针对系统N个时域接收信号、采用离散傅里叶变换提取L×M个梳状频点的快速实现问题，即在L个子频带每段中相同位置开始的M个连续子载波上提取承载的数据，利用变换分解方法将N个输入数据分解为L个长度为P的数据块，对每块分别进行DFT运算，利用基于时间抽取的分裂基裁剪方法简化这L块数据的DFT运算，通过将由输出频移带来的额外旋转运算分散到各级运算本身的旋转因子上，以及在一些级上进行裁剪的蝶形运算，从而在用户终端实现对分配给该用户的所有梳状频点的快速提取，大大降低运算复杂度。本发明能应用在信号处理或者图像处理等领域，实现离散傅里叶变换梳状频点提取。

Figure 201711170826

The present invention provides a method for rapidly extracting downlink comb spectrum in OFDMA distributed mode. The present invention aims at the fast realization problem of extracting L×M comb-shaped frequency points by using discrete Fourier transform for N time-domain received signals in the system, that is, extracting M continuous sub-carriers starting from the same position in each segment of L sub-bands. For the data carried, the N input data is decomposed into L data blocks of length P by the transform decomposition method, and the DFT operation is performed on each block separately, and the split basis clipping method based on time extraction is used to simplify the DFT operation of the L blocks of data. By dispersing the extra rotation operation caused by the output frequency shift to the rotation factors of each level of operation itself, as well as clipping butterfly operations at some stages, the user terminal realizes all comb frequency points allocated to the user. The fast extraction, greatly reducing the computational complexity. The invention can be applied in the fields of signal processing or image processing, etc., and realizes discrete Fourier transform comb frequency point extraction.

Figure 201711170826

Description

Method for rapidly extracting OFDMA distributed mode downlink comb spectrum
Technical Field
The invention relates to the field of communication signal processing, in particular to a method for quickly extracting comb spectrums when a downlink user terminal receives and demodulates in a distributed subcarrier allocation mode aiming at an Orthogonal Frequency Division Multiple Access (OFDMA) communication system.
Background
Orthogonal Frequency Division Multiple Access (OFDMA) is the mainstream multiple access scheme for the downlink of mobile communication systems such as LTE, 4G, and IEEE 802.16 m. OFDMA divides the entire frequency band into many subcarriers, converts a frequency selective fading channel into several flat fading subchannels, and thus can effectively resist frequency selective fading in a wireless mobile environment. By allocating different sub-carriers to different users, the sub-carriers between the users meet the requirement of mutual orthogonality, and the users can access the system at the same time. OFDMA may employ two subcarrier allocation patterns: centralized and distributed. Several continuous sub-carriers are distributed to one user in a centralized mode, and sub-carriers distributed to one user in a distributed mode are dispersed to the whole frequency band, so that a comb-shaped frequency spectrum is formed. The difficulty of channel estimation can be reduced in a centralized manner, but the frequency diversity gain obtained by the method is small, and the average performance of users is slightly poor; and the distributed mode can make the frequency selective fading correlation between the subcarriers weaker by utilizing the farther interval between the subcarriers, thereby obtaining stronger diversity gain and better user average performance, but the channel estimation is more complex in the mode. The two modes can be flexibly selected according to actual conditions. The patent mainly researches the extraction problem of comb spectrum when a user terminal receives and demodulates in a distributed subcarrier allocation mode downlink of an OFDMA communication system. The current main method is to adopt the conventional FFT methods such as radix-2 or split radix to extract the data carried on all subcarriers, and the invention cuts the conventional FFT method and realizes the rapid extraction of the data on the comb-shaped frequency point.
Disclosure of Invention
The present invention is directed to overcome the drawbacks and disadvantages of the prior art, and to provide a method for rapidly extracting a comb spectrum when a downlink user terminal receives and demodulates a distributed subcarrier allocation pattern for an Orthogonal Frequency Division Multiple Access (OFDMA) communication system.
Assuming that the total number of subcarriers in the system is N, all subcarriers are divided into subcarriers of length P (P ═ 2)m) L blocks of (i.e., N ═ L × P), each user starts occupying M (M ═ 2) contiguous at the same position in P carriers of each blockd<P) sub-carriers, user signals are comb-shaped frequency spectrum, the user terminal only needs to extract the data loaded on the L multiplied by M sub-carriers allocated to the user terminal for receiving and processing, and the data on the other sub-carriers are irrelevant to the user and do not need to be extracted. The conventional Fast Fourier Transform (FFT) methods such as a radix-2 or a split radix are adopted to extract data borne by all subcarriers in the conventional method, and the conventional FFT method is cut, so that the fast extraction of the data on the required frequency point is realized; the case where both P and M are powers of 2 is limited. The invention realizes DFT comb shape based on transform decomposition and butterfly graph cuttingAnd (4) rapidly solving the spectrum. The specific method is that N input data are decomposed into L blocks with the length of P points by a transform decomposition method to be respectively subjected to DFT operation, the L DFT blocks with the length of P are simply calculated by a split-based cutting method based on time extraction DIT, extra rotation operation brought by output frequency shift is dispersed on twiddle factors of each stage of operation, and butterfly operation of cutting is carried out on some stages, so that a fast operation method is realized, all comb frequency points allocated to a user are quickly extracted at a user terminal, and operation complexity is greatly reduced.
Specifically, for N input data [ x (0), x (1), x (2), …, x (N-1)]Calculating the comb output of the discrete Fourier transform; note that N points DFT of N input data are [ X (0), X (1), X (2), …, X (N-1)]Dividing the output N data into L blocks with the length of P, and setting M continuous frequency points which are occupied by a user from the Kth position to the K + M-1 position in each block; for the user, the total L multiplied by M output data of the L blocks are recorded as [ X (K), X (K +1), …, X (K + M-1); x (P + K), X (P + K +1), …, X (P + K + M-1); …, respectively; x ((L-1) P + K), X ((L-1) P + K +1), …, X ((L-1) P + K + M-1)](ii) a Particularly, in the present invention, the arbitrary continuous M data in each block of output data also includes the case that the M data are continuous in a circular loop, that is, the M data connected end to end in each block of output data also belong to the case to be considered; in the technical scheme, input data is equally divided into L blocks with the length of P by using a transformation decomposition mode, and the r-th block is marked as [ x ]r-1(0),xr-1(1),…,xr-1(P-1)]R is 1, …, L; the calculation of each block is realized in a hierarchical mode, and each block of input data has m levels, namely P is 2mWhere d-level operations cannot be clipped, i.e. M2d(ii) a The output of the t-th stage computation of the r-th block is denoted as [ x ](r-1),t(0),x(r-1),t(1),…,x(r-1),t(P-1)]The output of the t stage is the input of the t +1 stage, and the first M continuous data output by the mth stage of the mth block is the operation result of the block;
any continuous M data in each block of output data are solved, and the data are shifted to the left by cyclic shift and K bits to be changed to the 0 th bit to the M-1 th bit of the output end; cycling of the output according to the nature of the discrete Fourier transformLeft shifting by K bits is equivalent to rotating each data at the input of each block, i.e. multiplying the nth input data by a frequency shift twiddle factor
Figure GDA0002749165130000021
Shifting the frequency by a rotation factor
Figure GDA0002749165130000022
Dispersing the data to original twiddle factors of each level of a butterfly diagram of a split-radix method, and quickly solving the segmented DFT comb spectrum output, thereby realizing the quick extraction of all comb frequency points allocated to users, and specifically comprising the following steps:
step 1, dividing N time-lapse input data into L blocks with the length of P by using a transformation decomposition mode;
step 2, calculating the total series m of the split-radix butterfly diagram of each block of data and the series d which does not need to be cut for L blocks with the length of P, and carrying out bit reverse order on the serial number of each block of input data;
step 3, calculating all time extraction radix-2 butterflies in the level 1 of each block for L blocks with the length of P;
step 4, for L blocks with the length of P, calculating all the time from the 2 nd level to the d-th level of each block to extract split-radix butterflies;
step 5, calculating all the cut split-radix butterflies from the d +1 th level to the m th level of the last level of each block with L blocks with the length of P according to a split-radix cutting method;
and 6, recombining the L output blocks with the length of P obtained in the step to obtain the required comb-shaped output.
Further, in step 1, for the input of N points, the sequence { x (Ln)1)},n1P-1 maps to a first block of length P, denoted x respectively, …0(0),x0(1),…,x0(P-1); will sequence { x (Ln)1+1)},n1P-1 maps to a second block of length P, denoted x respectively, …1(0),x1(1),…,x1(P-1); …, respectively; will sequence { x (Ln)1+L-1)},n1P-1 maps to the lth block of length P, denoted as P, 0, …xL-1(0),xL-1(1),…,xL-1(P-1)。
In step 2, the total stage number of the DIT split radix butterfly diagram for the input of the P point in each block
Figure GDA0002749165130000035
M-point arbitrary continuous output, no need of cutting stage number
Figure GDA0002749165130000036
The number of stages to be clipped is m-d. Input to the r-th block { xr-1(0),xr-1(1),xr-1(2),…,xr-1The serial number of (P-1) is subjected to bit reversal, specifically, P serial numbers of 0,1, … and P-1 are represented by m-level binary, and then the binary numbers are inverted and represented as new decimal serial numbers. Reordering the order of elements in the input sequence by the new decimal sequence number and noting as { x(r-1)0(0),x(r-1)0(1),…,x(r-1)0(P-1), which will be the input to the 1 st stage operation of the r-th block.
In step 3, all time extraction radix-2 butterflies in level 1 of each of the L blocks with the length of P are respectively calculated. For the r-th block, all of the time-decimated radix-2 butterflies in stage 1 are performed. The formula for the time-decimated radix-2 butterfly may be expressed as:
Figure GDA0002749165130000031
where p and q +1 denote the serial numbers of the upper and lower nodes operated by the butterfly unit. Only p is 2 × 4a-2+b×4a+1The radix-2 butterfly at (a) needs to be computed, where a is 0,1,2, …; b is 0,1,2, …; so that P is less than P-1.
In step 4, all time extraction split-radix butterflies from the 2 nd level to the d-th level of each block are calculated respectively. For the r-th block, stages 2 through d perform a time-decimating split-radix butterfly operation. T in the split-radix butterfly denotes the t-th level, NiDenotes the size of the twiddle factor block in which the split base is located, where Ni＝2tAnd t is more than or equal to 2 and less than or equal to m. Input terminalFrequency shift twiddle factor
Figure GDA0002749165130000032
The twiddle factors distributed after the intrinsic twiddle factors of the 2 nd to d nd stages are
Figure GDA0002749165130000033
Wherein
q(u)＝(u-K)·2(m-t)，u＝0,1,…,2t-2-1, t ═ 2, …, m equation (3)
In particular, when u is 0, phi (u) and phi (3u) are the twiddle factors of the first split-base butterfly of the twiddle factor block in which the t-th split-base is located, respectively; when u is 1, phi (u) and phi (3u) are respectively twiddle factors of the second split-base butterfly of the twiddle factor block where the t-th split base is located, and so on. The formula for the computation of the time-decimated split-radix butterfly can be expressed as:
Figure GDA0002749165130000034
where j is the unit of an imaginary number,
n＝Ni×4a+1-Ni+2b×Ni×4a+1+ u formula (5)
Wherein a is-1, 0,1,2, …; b is 0,1,2, …; u is 0,1, …,2t-2-1, such that N < P-3Ni/4。
And extracting the split radix butterfly from the 2 nd stage to the d-th stage at all times to perform complete split radix butterfly operation according to the formula.
In step 5, all the clipped split radix butterflies from the d +1 th level to the m-th level of the last level of each block are calculated respectively. For the r block, according to a time-extracted split-radix clipping method, butterfly operations for clipping are performed on the split bases from the (d +1) th level to the m-th level of the last level in an output clipping manner, and the specific clipping manner is as follows:
in the d +1 th stage, only the first two of the time-decimated split-radix butterflies are computedOutputting, cutting time extracting the last two outputs of the split-radix butterfly, wherein the first two outputs are xt(n) and
Figure GDA0002749165130000041
t in the split-radix butterfly denotes the t-th level, NiDenotes the size of the twiddle factor block in which the split base is located, where Ni＝2d+1Frequency shift twiddle factor
Figure GDA0002749165130000042
The twiddle factors φ (u) and φ (3u) dispersed to the d +1 th stage original twiddle factors are calculated according to the equations (2) and (3). n is calculated according to equation (5). The formula for computing the split-radix butterfly in this step can be expressed as:
Figure GDA0002749165130000043
in each of the stages from the (d + 2) th stage to the mth stage of the last stage, the first M split radix butterflies are calculated for each rotation factor block of each stage, and each split radix butterfly calculates only the first output. Wherein N isi＝2tD +2, …, m, frequency shift twiddle factor
Figure GDA0002749165130000044
The twiddle factors phi (u) and phi (3u) distributed after the original twiddle factors from the d +2 th stage to the M-th stage of the last stage are calculated according to the formula (2) and the formula (3), except that the value range of u is 0,1, …, and M-1. The formula for computing the split-radix butterfly in this step can be expressed as:
Figure GDA0002749165130000045
wherein n is calculated according to formula (5), except that u is in the range of 0,1, …, and M-1.
In step 6, L blocks with length P are recombined to obtain comb-shaped output. In L blocks with the length of P, recombining output required by the user by using the first M data in each block, and calculating comb spectrum output [ X (K), X (K +1), … and X (K + M-1) required finally by using the following formula; x (P + K), X (P + K +1), …, X (P + K + M-1); …, respectively; x ((L-1) P + K), X ((L-1) P + K +1), …, X ((L-1) P + K + M-1) ].
Figure GDA0002749165130000046
Compared with the prior art, the invention has the following advantages and beneficial effects:
1. the invention provides a method for quickly extracting comb spectrum when a distributed subcarrier allocation mode downlink user terminal of an Orthogonal Frequency Division Multiple Access (OFDMA) communication system receives and demodulates, which can quickly evaluate the continuous comb spectrum with the output of DFT being equal space, and greatly reduces time complexity compared with the conventional FFT methods such as radix-2 and split radix which must completely solve all point values.
2. The invention provides a method for decomposing N-point DFT into smaller blocks by means of transform decomposition to respectively carry out DFT operation, and in addition, the rotation of the frequency shift of the DFT operation output end of each block to the input end is ingeniously dispersed to the original twiddle factors of each level of a time extraction DIT butterfly diagram, and the output cutting of a split basis is combined, so that the operation complexity is greatly reduced, the value of DFT at a comb spectrum output point is required in the fields of signal processing or image processing and the like, and when the requirement on the time complexity is higher, the method can be used for replacing the traditional FFT operation.
Drawings
Fig. 1 is a downlink user signal comb spectrum of an OFDMA distributed subcarrier allocation pattern according to the present invention.
FIG. 2 is a graph of the temporal decimation base 2 and the split-base butterfly according to the present invention.
FIG. 3 is a two cropping butterflies of the time-decimated split-radix of the present invention.
FIG. 4 is a 16-point time-decimated DIT split-radix butterfly diagram of the present invention.
Fig. 5 is a schematic flow chart of comb spectrum output extraction according to the present invention.
Detailed Description
The present invention will be described in further detail with reference to examples and drawings, but the present invention is not limited thereto.
Assuming that the total number of subcarriers of an OFDMA communication system is N, all subcarriers are divided into L blocks with a length of P, each user occupies M consecutive data from the K-th position to the K + M-1-th position in each P carriers of each block, and a user signal has a comb-shaped spectrum, as shown in fig. 1. The user terminal only needs to extract the data loaded on the L multiplied by M sub-carriers allocated to the user terminal, and the data on the rest sub-carriers are irrelevant to the user terminal and do not need to be extracted.
Specifically, a method for rapidly extracting a comb spectrum when a downlink user terminal receives and demodulates in a distributed subcarrier allocation mode in an OFDMA communication system is described by taking an example where N is 48 points, a decomposition factor L is 3, a block length P is 16, and each block continuously outputs M is 4 points from a K-6 point, including the steps of:
step 1, for the input of N points, mapping the N points to L blocks with the length of P by using a transformation decomposition mode.
Further, in step 1, for the input of N points, the sequence { x (Ln)1)},n1P-1 maps to a first block of length P, denoted x, 0, …, respectively0(0),x0(1),…,x0(P-1); will sequence { x (Ln)1+1)},n1P-1 maps to a second block of length P, denoted x respectively, 0, …1(0),x1(1),…,x1(P-1); …, respectively; will sequence { x (Ln)1+L-1)},n1P-1 maps to the lth block of length P, denoted x, 0, …, respectivelyL-1(0),xL-1(1),…,xL-1(P-1)。
Further, in step 1, as shown in the process from (a) to (b) in fig. 5, x (0), x (3), x (6), x (9), …, x (45) are mapped to the first block with length of 16, which is denoted as x0(0),x0(1),x0(2),x0(3),…,x0(15) (ii) a Mapping x (1), x (4), x (7), x (10), …, x (46) into a second block of length 16, denoted x1(0),x1(1),x1(2),x1(3),…,x1(15) (ii) a Mapping x (2), x (5), x (8), x (11), …, x (47) into a third block of length 16, denoted x2(0),x2(1),x2(2),x2(3),…,x2(15)。
And 2, calculating the total series of each split radix butterfly graph and the series required to be cut for L blocks with the length of P, and performing bit reverse order on the input point of each block.
Further, in step 2, the total number of stages of the input DIT split radix butterfly map for the P point in each block is
Figure GDA0002749165130000062
M points are output randomly and continuously, and the stage number without cutting is
Figure GDA0002749165130000063
The number of stages to be clipped is m-d. Input to the r-th block { xr-1(0),xr-1(1),xr-1(2),…,xr-1The serial number of (P-1) is subjected to bit reversal, specifically, P serial numbers of 0,1, … and P-1 are represented by m-level binary, and then the binary numbers are inverted and represented as new decimal serial numbers. Reordering the order of elements in the input sequence by the new decimal sequence number and noting as { x(r-1)0(0),x(r-1)0(1),…,x(r-1)0(P-1), which will be the input to the 1 st stage operation of the r-th block.
Further, in step 2, the data of 16 points in the first block is used for explanation, and as shown in fig. 4, the 16-point input M is a time-extracted DIT butterfly map which is output at 4 points arbitrarily and continuously. The 16-point DIT butterfly has 4 levels, the levels which do not need to be clipped are the 1 st level and the 2 nd level, and the levels which need to be clipped are the 3 rd level and the 4 th level. The 16 input points are subjected to bit reversal sequence, and then the input values are adjusted according to the new decimal serial numbers to obtain corresponding new decimal serial numbers x respectively00(0)(＝x0(0)),x00(1)(＝x0(8)),x00(2)(＝x0(4)),x00(3)(＝x0(12)),x00(4)(＝x0(2)),x00(5)(＝x0(10)),x00(6)(＝x0(6)),x00(7)(＝x0(14)),x00(8)(＝x0(1)),x00(9)(＝x0(9)),x00(10)(＝x0(5)),x00(11)(＝x0(13)),x00(12)(＝x0(3)),x00(13)(＝x013(11)),x00(14)(＝x0(7)),x00(15)(＝x0(15))。
And 3, calculating all time extraction radix-2 butterflies in the level 1 of each block for L blocks with the length of P.
Further, in step 3, all the time-decimated radix-2 butterflies in level 1 of each block are computed separately, and for the r-th block, all the time-decimated radix-2 butterflies in level 1 are performed. The formula for the time-decimated radix-2 butterfly may be expressed as:
Figure GDA0002749165130000061
where p and q +1 denote the serial numbers of the upper and lower nodes operated by the butterfly unit. Only p is 2 × 4a-2+b×4a+1The radix-2 butterfly at (a) needs to be computed, where a is 0,1,2, …; b is 0,1, 2; so that P is less than P-1.
Further, in step 3, all the time-decimated radix-2 butterflies in level 1 of each block are computed separately, for block 1, according to the formula for the computation of p, x in FIG. 400(0) And x00(1),x00(4) And x00(5),x00(6) And x00(7)、x00(8) And x00(9)、x00(12) And x00(13) Corresponding to the two input terminals of the radix-2 butterfly, respectively. When K is 6, the twiddle factor of radix-2 butterfly combined with offset coefficient is (-1)6The output value of each radix-2 butterfly of the first stage can be calculated according to equation (1) as 1.
And 4, for L blocks with the length of P, calculating all the time from the 2 nd stage to the d-th stage of each block to extract the split radix butterfly.
Further, in step 4, all time-decimated split-radix butterflies in level 2 through level d are computed for each block, respectively, as for the r-th block, level 2 through level dThe d-th stage performs a time-decimation split-radix butterfly operation. T in the split-radix butterfly denotes the t-th level, NiDenotes the size of the twiddle factor block in which the split base is located, where Ni＝2tAnd t is more than or equal to 2 and less than or equal to m. Input frequency shift twiddle factor
Figure GDA0002749165130000071
Twiddle factors distributed after intrinsic twiddle factors of 2 nd to d nd stages
Figure GDA0002749165130000072
Wherein
q(u)＝(u-K)·2(m-t)，u＝0,1,…,2t-2-1, t ═ 2, …, m equation (3)
In particular, when u is 0, phi (u) and phi (3u) are the twiddle factors of the first split-base butterfly of the twiddle factor block in which the t-th split-base is located, respectively; when u is 1, phi (u) and phi (3u) are respectively twiddle factors of the second split-base butterfly of the twiddle factor block where the t-th split base is located, and so on. The formula for the computation of the time-decimated split-radix butterfly can be expressed as:
Figure GDA0002749165130000073
where j is the unit of an imaginary number,
n＝Ni×4a+1-Ni+2b×Ni×4a+1+ u formula (5)
Wherein a is-1, 0,1,2, …; b is 0,1,2, …; u is 0,1, …,2t-2-1, such that N < P-3Ni/4。
The split radix butterfly is extracted from the 2 nd stage to the d nd stage at all times, and the complete split radix butterfly operation is carried out according to the formula.
Further, in step 4, computing the extracted split-radix butterfly at all times in each of the 2 nd to d-th stages, respectively, for the 1 st block, 16 points input the DIT split-radix butterfly with 4 points output in any sequence,only the second stage is a complete time-decimated radix-butterfly, the size N of the twiddle factor block in which the second stage's split bases are locatediThe values of n are 0, 8, 12, and u is 0 according to equation (5). When n is 0, 8, 12, the corresponding twiddle factors are expressed according to the formula (2) and the formula (3)
Figure GDA0002749165130000074
And
Figure GDA0002749165130000075
the output of each split-radix butterfly of the second stage can now be calculated according to equation (4).
And 5, calculating all the cut split-radix butterflies from the (d +1) th level to the (m) th level of the last level of each block for the L blocks with the length of P according to a split-radix cutting method.
Further, in step 5, all the tailored split-radix butterflies from the d +1 th level to the m th level of the last level of each block are respectively calculated, for the r-th block, the butterfly operation of tailoring is performed on the split bases from the d +1 th level to the m th level of the last level of the block in an output tailoring manner according to a split-radix tailoring method extracted in time, and the specific tailoring manner is as follows: in the d +1 stage, only the first two outputs of the time-decimated split-radix butterfly are calculated, and the second two outputs of the time-decimated split-radix butterfly are clipped, where the first two outputs are xt(n) and
Figure GDA0002749165130000076
t in the split-radix butterfly denotes the t-th level, NiDenotes the size of the twiddle factor block in which the split base is located, where Ni＝2d+1Frequency shift twiddle factor
Figure GDA0002749165130000081
The twiddle factors φ (u) and φ (3u) dispersed to the d +1 th stage original twiddle factors are calculated according to the equations (2) and (3). n is calculated according to equation (5). The formula for computing the split-radix butterfly in this step can be expressed as:
Figure GDA0002749165130000082
in each of the stages from the (d + 2) th stage to the mth stage of the last stage, the first M split radix butterflies are calculated for each rotation factor block of each stage, and each split radix butterfly calculates only the first output. Wherein N isi＝2tD +2, …, m, frequency shift twiddle factor
Figure GDA0002749165130000083
The twiddle factors phi (u) and phi (3u) distributed after the original twiddle factors from the d +2 th stage to the M-th stage of the last stage are calculated according to the formula (2) and the formula (3), except that the value range of u is 0,1, …, and M-1. The formula for computing the split-radix butterfly in this step can be expressed as:
Figure GDA0002749165130000084
wherein n is calculated according to formula (5), except that u is in the range of 0,1, …, and M-1.
Furthermore, in step 5, all clipped split-radix butterflies from the d +1 th level to the m-th level of the last level of each block are calculated according to the split-radix clipping method. For the first block, the 3 rd and 4 th split bases are subjected to butterfly operation of clipping by adopting an output clipping mode, and the specific mode is as follows: in stage 3, only the first two outputs of the time-decimated split-radix butterfly are computed. The size of the twiddle factor block where the 3 rd level splitting base is positioned is NiWhen u is 0, n is 0; when u is 1, n is 1. Thus, the two clipped outputs of the split-radix butterfly are x03(0) And x03(2)，x03(1) And x03(3). The output is x03(0) And x03(2) Has a rotation factor of
Figure GDA0002749165130000085
And
Figure GDA0002749165130000086
the output is x03(1) And x03(3) Butterfly shape ofThe twiddle factor is expressed by the following formula (2) and formula (3)
Figure GDA0002749165130000087
And
Figure GDA0002749165130000088
the first two outputs of the level 3 split radix butterfly can be calculated using equation (6). While at stage 4, only the first output of the time-decimated split-radix butterfly is computed. Wherein the size N of the twiddle factor block in which the 4 th level of the splitting base is locatediIn the 4 th stage, the value of n is obtained according to the formula (5), and when u is 0, n is 0, and the corresponding output is x04(0) (ii) a When u is 1, n is 1, corresponding to an output of x04(1) (ii) a When u is 2, n is 2, corresponding to an output of x04(2) (ii) a When u is 3, n is 3, corresponding to output x04(3). According to the formula (2) and the formula (3), the output is x04(0) Has a rotation factor of
Figure GDA0002749165130000089
And
Figure GDA00027491651300000810
the output is x04(1) Has a rotation factor of
Figure GDA00027491651300000811
And
Figure GDA00027491651300000812
the output is x04(2) Has a rotation factor of
Figure GDA00027491651300000813
And
Figure GDA00027491651300000814
the output is x04(3) Has a rotation factor of
Figure GDA00027491651300000815
And
Figure GDA00027491651300000816
the first output of the 4 th stage, 4 split radix butterfly, can be calculated using equation (7). All other outputs are not calculated.
And 6, carrying out recombination operation on the L blocks with the length of P to obtain the comb-shaped output required by us.
Further, in step 6, L blocks with length P are recombined to obtain the comb-like output we need. In L blocks with the length of P, recombining output required by the user by using the first M data in each block, and calculating comb spectrum output [ X (K), X (K +1), … and X (K + M-1) required finally by using the following formula; x (P + K), X (P + K +1), …, X (P + K + M-1); …, respectively; x ((L-1) P + K), X ((L-1) P + K +1), …, X ((L-1) P + K + M-1) ].
Figure GDA0002749165130000091
The above embodiments are preferred embodiments of the present invention, but the present invention is not limited to the above embodiments, and any other changes, modifications, substitutions, combinations, and simplifications which do not depart from the spirit and principle of the present invention should be construed as equivalents thereof, and all such changes, modifications, substitutions, combinations, and simplifications are intended to be included in the scope of the present invention.

Claims (5)

1.OFDMA分布式模式下行链路梳状谱快速提取方法，其特征在于：在OFDMA通信系统分布式子载波分配模式下行链路中，假设系统总的子载波个数为N个，所有子载波分成长度为P的L块，每个用户在每块P个载波中相同位置开始占用连续的M个子载波，用户信号成梳状频谱，用户终端接收处理只需提取出分配给它的L×M个子载波上承载的数据，其余子载波上的数据与该用户无关无需提取；对常规FFT方法上进行裁剪，实现了所需频点上数据的快速提取，限制P与M均为2的幂的情况；1. A method for rapidly extracting a comb spectrum in an OFDMA distributed mode downlink, characterized in that: in the downlink of a distributed subcarrier allocation mode of an OFDMA communication system, assuming that the total number of subcarriers in the system is N, all subcarriers Divided into L blocks of length P, each user starts to occupy consecutive M sub-carriers at the same position in each block of P carriers, the user signal becomes a comb-shaped spectrum, and the user terminal only needs to extract the L×M allocated to it for reception processing. The data carried on one sub-carrier, and the data on the remaining sub-carriers have nothing to do with the user and do not need to be extracted; the conventional FFT method is tailored to realize the rapid extraction of data on the required frequency points, and it is limited that P and M are both powers of 2 Happening; 对N个输入数据[x(0),x(1),x(2),…,x(N-1)]，计算其离散傅里叶变换的梳状输出；记N个输入数据的N点DFT为[X(0),X(1),X(2),…,X(N-1)]，将输出的N个数据分为长度为P的L块，设某用户在每块中占据从第K个位置到第K+M-1位置的M个连续频点；对该用户来说L块总的L×M个输出数据记为[X(K),X(K+1),…,X(K+M-1)；X(P+K),X(P+K+1),…,X(P+K+M-1)；…；X((L-1)P+K),X((L-1)P+K+1),…,X((L-1)P+K+M-1)]；每块输出数据中任意连续的M个数据也包括这M个数据为圆周循环连续的情况，即在每块输出数据中首尾相连的M个数据也属于需考虑的情况；将输入数据利用变换分解方式同样分为长度为P的L块，第r块记为[xr-1(0),xr-1(1),…,xr-1(P-1)]，r＝1,…,L；每块的计算采用分级方式实现，每块输入数据一共有m级，即P＝2m，其中有d级运算不能裁剪，即M＝2d；第r块的第t级计算的输出记为[x(r-1),t(0),x(r-1),t(1),…,x(r-1),t(P-1)]，t＝1～m，其中第t级的输出即为第t+1级的输入，第r块第m级输出的前M个连续数据是该块的运算结果；For N input data [x(0),x(1),x(2),...,x(N-1)], calculate the comb output of its discrete Fourier transform; record N of N input data The point DFT is [X(0), X(1), X(2),...,X(N-1)], and the output N data is divided into L blocks of length P, and a user is assumed to be in each block. Occupies M consecutive frequency points from the Kth position to the K+M-1th position; for the user, the total L×M output data of the L block is recorded as [X(K), X(K+1 ),…,X(K+M-1);X(P+K),X(P+K+1),…,X(P+K+M-1);…;X((L-1 )P+K),X((L-1)P+K+1),…,X((L-1)P+K+M-1)]; any consecutive M data in each block of output data It also includes the case that the M data are continuous in a circular cycle, that is, the M data that are connected end to end in each block of output data is also a situation that needs to be considered; the input data is also divided into L blocks of length P by the transformation decomposition method, The rth block is recorded as [x r-1 (0), x r-1 (1),...,x r-1 (P-1)], r=1,...,L; the calculation of each block adopts a hierarchical method Realization, each block of input data has a total of m levels, that is, P=2 m , and there are d-level operations that cannot be cut, that is, M=2 d ; the output of the t-th level calculation of the r-th block is denoted as [x (r-1) ,t (0),x (r-1),t (1),…,x (r-1),t (P-1)], t=1～m, where the output of the t-th stage is the For the input of stage t+1, the first M continuous data output by the mth stage of the rth block is the operation result of the block; 所求的每块输出数据中任意连续的M个数据，将它向左循环移位移动K位变到输出端的第0到M-1位；根据离散傅里叶变换的性质，输出端的循环左移K位等价于对每块的输入端每个数据进行旋转，即对第n个输入数据乘以频移旋转因子
Figure FDA0002749165120000011
将该频移旋转因子
Figure FDA0002749165120000012
分散到分裂基方法蝶形图各级本来的旋转因子上，快速求解分块DFT梳状谱输出，从而实现对分配给用户的所有梳状频点的快速提取，具体包括以下步骤：For any continuous M data in each block of output data, move it to the left by cyclic shift by K bits to the 0th to M-1th bits of the output; according to the properties of discrete Fourier transform, the cyclic left of the output is Shifting K bits is equivalent to rotating each data at the input of each block, that is, multiplying the nth input data by the frequency shift twiddle factor
Figure FDA0002749165120000011
the frequency shift twiddle factor
Figure FDA0002749165120000012
Disperse to the original twiddle factors at all levels of the butterfly diagram of the split basis method, and quickly solve the block DFT comb spectrum output, so as to realize the rapid extraction of all comb frequency points assigned to the user, including the following steps: 步骤1.对N个时输入数据，利用变换分解的方式分成L个长度为P的块：对N个输入数据[x(0),x(1),x(2),…,x(N-1)]，将序列{x(Ln1)},n1＝0,…,P-1，映射到第一个长度为P的块中，分别记为x0(0),x0(1),…,x0(P-1)；将序列{x(Ln1+1)},n1＝0,…,P-1，映射到第二个长度为P的块中，分别记为x1(0),x1(1),…,x1(P-1)；……；将序列{x(Ln1+L-1)},n1＝0,…,P-1，映射到第L个长度为P的块中，分别记为xL-1(0),xL-1(1),…,xL-1(P-1)；Step 1. For N input data, divide it into L blocks of length P by transforming and decomposing: for N input data [x(0), x(1), x(2),...,x(N -1)], map the sequence {x(Ln 1 )}, n 1 =0,...,P-1 to the first block of length P, denoted as x 0 (0), x 0 ( 1),...,x 0 (P-1); map the sequence {x(Ln 1 +1)},n 1 =0,...,P-1 to the second block of length P, respectively record is x 1 (0), x 1 (1),...,x 1 (P-1);...; the sequence {x(Ln 1 +L-1)},n 1 =0,...,P-1 , mapped to the Lth block of length P, denoted as x L-1 (0), x L-1 (1),...,x L-1 (P-1); 步骤2.对L个长度为P的块，计算每块数据的分裂基蝶形图总级数m和不需要裁剪的级数d，并将每块输入数据的序号进行比特逆序：对每个数据块，DIT分裂基蝶形图总的级数为
Figure FDA0002749165120000013
M点任意连续的输出，不需要裁剪的级数为
Figure FDA0002749165120000014
需要裁剪的级数为m-d；对第r块的输入{xr-1(0),xr-1(1),xr-1(2),…,xr-1(P-1)}的序号进行比特逆序，具体就是将0,1,…,P-1这P个序号以m级二进制表示，再将这些二进制数码进行翻转，并表示成新的十进制序号；按照新的十进制序号重新排序输入序列中元素的顺序，并记为{x(r-1)0(0),x(r-1)0(1),…,x(r-1)0(P-1)}，这将是第r块第1级运算的输入；Step 2. For L blocks of length P, calculate the total number of stages m of the split-base butterfly diagram and the number of stages d that do not need to be cut for each block of data, and perform bit reverse order on the serial number of each block of input data: for each Data block, the total number of series of DIT split basis butterfly diagram is
Figure FDA0002749165120000013
Arbitrary continuous output of M points, the number of series that does not need to be clipped is
Figure FDA0002749165120000014
The number of series to be clipped is md; for the input to the rth block {x r-1 (0), x r-1 (1), x r-1 (2),...,x r-1 (P-1) The serial number of } is bit reversed, specifically, the P serial numbers of 0, 1, ..., P-1 are represented by m-level binary, and then these binary numbers are inverted and represented as a new decimal serial number; according to the new decimal serial number Reorder the elements in the input sequence and denote {x (r-1)0 (0),x (r-1)0 (1),…,x (r-1)0 (P-1)} , which will be the input to the 1st stage operation of the rth block; 步骤3.对L个长度为P的块，计算每块第1级中所有的时间抽取基2蝶形；Step 3. For L blocks of length P, calculate all the time extraction radix 2 butterflies in the first level of each block; 步骤4.对L个长度为P的块，计算每块第2级到第d级中的所有时间抽取分裂基蝶形；Step 4. For L blocks of length P, calculate the split-base butterfly at all times from the 2nd to the dth level of each block; 步骤5.对L个长度为P的块，按照分裂基裁剪方法计算每块第d+1级到最后一级第m级的所有裁剪的分裂基蝶形；Step 5. For L blocks of length P, according to the split basis cropping method, calculate the split basis butterflies of all trimmings from the d+1th level to the last level mth level of each block; 步骤6.将上述步骤得到的L个长度为P的输出块进行重组，得到所需要的梳状的输出。Step 6. Reorganize the L output blocks of length P obtained in the above steps to obtain the required comb-shaped output. 2.根据权利要求1所述的OFDMA分布式模式下行链路梳状谱快速提取方法，其特征在于：步骤3中，分别计算L个长度为P的块中每块的第1级中所有的时间抽取基2蝶形；对于第r块而言，执行第1级中所有的时间抽取基2蝶形运算；时间抽取基2蝶形的计算公式表示为：2. The method for rapidly extracting downlink comb spectrum in OFDMA distributed mode according to claim 1, characterized in that: in step 3, respectively calculating all L blocks in the first level of each block of length P Time extraction radix 2 butterfly; for the rth block, all time extraction radix 2 butterfly operations in the first stage are performed; the calculation formula of time extraction radix 2 butterfly is expressed as:
Figure FDA0002749165120000021
Figure FDA0002749165120000021
 其中p和q＝p+1表示蝶形单元运算的上下节点的序号；只有p＝2×4a-2+b×4a+1处的基2蝶形需要计算，其中a＝0,1,2，…；b＝0,1,2,…；使p＜P-1。Where p and q=p+1 represent the sequence numbers of the upper and lower nodes of the butterfly unit operation; only the radix-2 butterfly at p=2×4 a -2+b×4 a+1 needs to be calculated, where a=0,1 ,2,...; b=0,1,2,...; so that p<P-1. 3.根据权利要求1所述的OFDMA分布式模式下行链路梳状谱快速提取方法，其特征在于：步骤4中，分别计算每块第2级到第d级中的所有时间抽取分裂基蝶形；对于第r块而言，第2级到第d级执行时间抽取分裂基蝶形运算；分裂基蝶形中t表示所在第t级，Ni表示分裂基所在的旋转因子块的大小，其中Ni＝2t，2≤t≤m；输入端频移旋转因子
Figure FDA0002749165120000022
分散到第2级到第d级本来的旋转因子之后新的旋转因子为3. The method for rapidly extracting the downlink comb spectrum in OFDMA distributed mode according to claim 1, wherein in step 4, the split-base butterfly is calculated separately for all times in the 2nd to the dth stage of each block. For the rth block, the second stage to the dth stage perform the time extraction split-base butterfly operation; in the split-base butterfly, t represents the t- th stage, and Ni represents the size of the twiddle factor block where the split basis is located, where N i =2 t , 2≤t≤m; input frequency shift twiddle factor
Figure FDA0002749165120000022
The new twiddle factor after dispersing to the original twiddle factor of the 2nd to the dth stage is
Figure FDA0002749165120000023
Figure FDA0002749165120000023
 其中in q(u)＝(u-K)·2(m-t)，u＝0,1,…,2t-2-1，t＝2,…,m. 公式(3)q(u)=(uK)·2 (mt) , u=0,1,...,2 t-2 -1, t=2,...,m. Equation (3) 特别地，当u＝0时，φ(u)与φ(3u)分别为第t级分裂基所在旋转因子块的第一个分裂基蝶形的旋转因子；u＝1时，φ(u)与φ(3u)分别为第t级分裂基所在旋转因子块的第二个分裂基蝶形的旋转因子，以此类推；时间抽取的分裂基蝶形的计算公式可表示为:In particular, when u=0, φ(u) and φ(3u) are respectively the twiddle factors of the first split basis butterfly of the twiddle factor block where the t-th-level split basis is located; when u=1, φ(u) and φ(3u) are respectively the twiddle factor of the second split basis butterfly of the twiddle factor block where the t-th level split basis is located, and so on; the calculation formula of the split basis butterfly extracted by time can be expressed as:
Figure FDA0002749165120000031
Figure FDA0002749165120000031
 这里的j为虚数单位，Here j is the imaginary unit, n＝Ni×4a+1-Ni+2b×Ni×4a+1+u 公式(5)n=N i ×4 a+1 -N i +2b×N i ×4 a+1 +u Formula (5) 其中a＝-1,0,1,2，…；b＝0,1,2,…；u＝0,1,…,2t-2-1，使得n＜P-3Ni/4；where a=-1,0,1,2,...; b=0,1,2,...; u=0,1,...,2 t- 2-1, so that n<P-3N i /4; 对第2级到第d级的所有时间抽取分裂基蝶形按照这个公式进行完整的分裂基蝶形运算。Extract the split-base butterfly at all times from the second stage to the d-th stage and perform a complete split-based butterfly operation according to this formula. 4.根据权利要求1所述的OFDMA分布式模式下行链路梳状谱快速提取方法，其特征在于：步骤5中，分别计算每块第d+1级到最后一级第m级的所有裁剪的分裂基蝶形；对于第r块而言，按照时间抽取的分裂基裁剪方法，对第d+1级到最后一级第m级的分裂基均采用输出裁剪的方式进行裁剪的蝶形运算，具体的裁剪方式为：4. The method for rapidly extracting downlink comb spectrum in OFDMA distributed mode according to claim 1, wherein in step 5, all trimmings from the d+1th level to the last level mth level of each block are calculated respectively. For the r-th block, according to the time-extracted split-base clipping method, the output clipping method is adopted for the split-basis of the d+1th level to the last level m-th level. , the specific cutting method is: 在第d+1级中，仅仅计算时间抽取分裂基蝶形的前两个输出，裁剪时间抽取分裂基蝶形的后两个输出，这前两个输出为xt(n)与
Figure FDA0002749165120000032
分裂基蝶形中t表示所在第t级，Ni表示分裂基所在的旋转因子块的大小，其中Ni＝2d+1，频移旋转因子
Figure FDA0002749165120000033
分散到第d+1级本来的旋转因子之后新的旋转因子φ(u)与φ(3u)按公式(2)及公式(3)计算，n按公式(5)计算；本步裁剪分裂基蝶形的计算公式可表示为:In stage d+1, only the first two outputs of the split-base butterfly are extracted at computing time, and the last two outputs of the split-base butterfly are extracted at clipping time. The first two outputs are x t (n) and
Figure FDA0002749165120000032
In the split-base butterfly, t represents the t- th stage, and Ni represents the size of the twiddle factor block where the split-base is located, where N i = 2 d+1 , the frequency shift twiddle factor
Figure FDA0002749165120000033
The new twiddle factors φ(u) and φ(3u) are calculated according to formula (2) and formula (3) after dispersing to the original twiddle factor of the d+1th stage, and n is calculated according to formula (5); this step is to cut and split the basis The calculation formula of butterfly can be expressed as:
Figure FDA0002749165120000034
Figure FDA0002749165120000034
 第d+2级到最后一级第m级的各级中，对于每一级的每一个旋转因子块计算前M个分裂基蝶形，而每个分裂基蝶形仅仅计算第一个输出；其中Ni＝2t，t＝d+2,…,m，频移旋转因子
Figure FDA0002749165120000035
分散到第d+2级到最后一级第m级本来的旋转因子之后的旋转因子φ(u)与φ(3u)按公式(2)及公式(3)计算，只是u的取值范围为u＝0,1,…,M-1；本步裁剪分裂基蝶形的计算公式表示为:In the stages from the d+2th stage to the last stage of the mth stage, the first M split-base butterflies are calculated for each twiddle factor block of each stage, and each split-base butterfly only calculates the first output; where N i =2 t , t=d+2,...,m, frequency shift twiddle factor
Figure FDA0002749165120000035
The twiddle factors φ(u) and φ(3u) after the original twiddle factor from the d+2th stage to the last stage of the mth stage are calculated according to formula (2) and formula (3), except that the value range of u is u=0,1,...,M-1; the calculation formula of the cutting and splitting base butterfly in this step is expressed as:
Figure FDA0002749165120000036
Figure FDA0002749165120000036
 其中，n按公式(5)计算，只是u的取值范围为u＝0,1,…,M-1。Among them, n is calculated according to formula (5), except that the value range of u is u=0,1,...,M-1. 5.根据权利要求1所述的OFDMA分布式模式下行链路梳状谱快速提取方法，其特征在于：步骤6中，将L个长度为P的块进行重组操作，得到所需要的梳状的输出；L个长度为P的块中，利用每块中前M个数据重组出需要的输出，利用下式求出最终所需要的梳状谱输出[X(K),X(K+1),…,X(K+M-1)；X(P+K),X(P+K+1),…,X(P+K+M-1)；…；X((L-1)P+K),X((L-1)P+K+1),…,X((L-1)P+K+M-1)]；5. The method for rapidly extracting the downlink comb spectrum in OFDMA distributed mode according to claim 1, wherein in step 6, a reorganization operation is performed on L blocks of length P to obtain the required comb-shaped spectrum. Output: In L blocks of length P, use the first M data in each block to reconstruct the required output, and use the following formula to find the final required comb spectrum output [X(K), X(K+1) ,…,X(K+M-1);X(P+K),X(P+K+1),…,X(P+K+M-1);…;X((L-1) P+K),X((L-1)P+K+1),…,X((L-1)P+K+M-1)];
Figure FDA0002749165120000041
Figure FDA0002749165120000041
CN201711170826.2A 2017-11-22 2017-11-22 Method for rapidly extracting OFDMA distributed mode downlink comb spectrum Active CN107979454B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201711170826.2A CN107979454B (en) 2017-11-22 2017-11-22 Method for rapidly extracting OFDMA distributed mode downlink comb spectrum

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201711170826.2A CN107979454B (en) 2017-11-22 2017-11-22 Method for rapidly extracting OFDMA distributed mode downlink comb spectrum

Publications (2)

Publication Number Publication Date
CN107979454A CN107979454A (en) 2018-05-01
CN107979454B true CN107979454B (en) 2021-01-19

Family

ID=62010789

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201711170826.2A Active CN107979454B (en) 2017-11-22 2017-11-22 Method for rapidly extracting OFDMA distributed mode downlink comb spectrum

Country Status (1)

Country Link
CN (1) CN107979454B (en)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP1766798A4 (en) * 2004-06-30 2008-11-12 Motorola Inc Frequency-hopped ifdma communication system
CN102036135A (en) * 2011-01-25 2011-04-27 上海交通大学 Orthogonal frequency division multiple access (OFDMA) passive optical network transmission system
CN103402147A (en) * 2013-07-30 2013-11-20 电子科技大学 Orthogonal frequency division multiple access (OFDMA) metropolitan area access fusion network compatible with optical virtual private network (OVPN) communication, and communication method thereof
CN104780128A (en) * 2015-04-14 2015-07-15 哈尔滨工程大学 Sparse channel estimation and pilot optimization method for underwater sound OFDMA uplink communication
US9294218B2 (en) * 2004-07-16 2016-03-22 Qualcomm Incorporated Rate prediction in fractional reuse systems
CN106528495A (en) * 2016-11-23 2017-03-22 华南理工大学 Method for obtaining output value of FFT in any continuous interval

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP1766798A4 (en) * 2004-06-30 2008-11-12 Motorola Inc Frequency-hopped ifdma communication system
US9294218B2 (en) * 2004-07-16 2016-03-22 Qualcomm Incorporated Rate prediction in fractional reuse systems
CN102036135A (en) * 2011-01-25 2011-04-27 上海交通大学 Orthogonal frequency division multiple access (OFDMA) passive optical network transmission system
CN103402147A (en) * 2013-07-30 2013-11-20 电子科技大学 Orthogonal frequency division multiple access (OFDMA) metropolitan area access fusion network compatible with optical virtual private network (OVPN) communication, and communication method thereof
CN104780128A (en) * 2015-04-14 2015-07-15 哈尔滨工程大学 Sparse channel estimation and pilot optimization method for underwater sound OFDMA uplink communication
CN106528495A (en) * 2016-11-23 2017-03-22 华南理工大学 Method for obtaining output value of FFT in any continuous interval

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
"a general resource allocation algorithm with fairness for for SDMA/MISO/OFDMA systems";weishan lu,fei ji,hua yu;《IEEE》;20110908;全文 *
"lagrange multiplier based fair resource allocation for mimo-ofdma system";dou ni,fangjiong chen,cunmei ji;《IEEE》;20090505;全文 *

Also Published As

Publication number Publication date
CN107979454A (en) 2018-05-01

Similar Documents

Publication Publication Date Title
He et al. Designing pipeline FFT processor for OFDM (de) modulation
EP1554831B1 (en) Frequency hopping ofdma method using symbols of comb pattern
KR101012990B1 (en) Improved Frequency Division Multiple Access for Wireless Communications
KR101984151B1 (en) Fbmc signal transmitting method, receiving method, transmitter and receiver
JP2008533854A (en) Partial FFT processing and demodulation for systems with multiple subcarriers
US6169723B1 (en) Computationally efficient analysis and synthesis of real signals using discrete fourier transforms and inverse discrete fourier transforms
CN1507727A (en) Channel estimation for orthogonal frequency division multiplexed (OFDM) signals
WO2007022831A1 (en) Combined ofdm and wavelet multi-carrier transceiver
Chafii et al. A necessary condition for waveforms with better PAPR than OFDM
KR100628303B1 (en) Variable point IFT / FTF method and apparatus
US8351525B2 (en) Orthogonal frequency division multiplexing transmitting and receiving device
CN106130946A (en) A kind of biorthogonal frequency-division multiplex multi-carrier modulation/demodulation methods and system
CN106059732A (en) Pilot-frequency optimization method in OFDM sparse channel estimation based on compressive sensing
CN107979454B (en) Method for rapidly extracting OFDMA distributed mode downlink comb spectrum
CN106656891B (en) Data processing equipment in LTE system
CN111291315A (en) Data processing method, device and equipment
Adiono et al. Low latency parallel-pipelined configurable FFT-IFFT 128/256/512/1024/2048 for LTE
KR100810490B1 (en) Fast Fourier Transform Device and Orthogonal Frequency Division Multiplexing Receiver Comprising the Same
Bunsri et al. FPGA Implementation of IFFT-2048 Points for High Speed Transmitter OFDM Communication Systems with 64-QAM Mapping
CN1980114A (en) Channel estimation emitting-receiving device and method
CN108566355B (en) A Peak-to-Average Ratio Reduction Algorithm for Discrete Multitone Modulation Systems
Jang et al. Low latency IFFT design for OFDM systems supporting full-duplex FDD
Rajni et al. Distinctive approach to design tree in wavelet packet based OFDM system
Tseng et al. A reduced-complexity PTS scheme for peak-to-average power ratio reduction in OFDM systems
Ibing et al. On hardware implementation of multiuser multiplexing for SC-FDMA

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant