WO2010135935A1 - 一种频偏估计装置和方法 - Google Patents
一种频偏估计装置和方法 Download PDFInfo
- Publication number
- WO2010135935A1 WO2010135935A1 PCT/CN2010/071769 CN2010071769W WO2010135935A1 WO 2010135935 A1 WO2010135935 A1 WO 2010135935A1 CN 2010071769 W CN2010071769 W CN 2010071769W WO 2010135935 A1 WO2010135935 A1 WO 2010135935A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- training sequence
- channel estimation
- offset
- cell
- frequency offset
- Prior art date
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L27/00—Modulated-carrier systems
- H04L27/0014—Carrier regulation
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B1/00—Details of transmission systems, not covered by a single one of groups H04B3/00 - H04B13/00; Details of transmission systems not characterised by the medium used for transmission
- H04B1/69—Spread spectrum techniques
- H04B1/707—Spread spectrum techniques using direct sequence modulation
- H04B1/7097—Interference-related aspects
- H04B1/7103—Interference-related aspects the interference being multiple access interference
Definitions
- the present invention relates to a wireless communication system, and in particular to a frequency offset estimation apparatus and method.
- both the base station and the terminal transmit and receive at a nominal carrier frequency. Due to the limitations of the device, there is a certain deviation between the actual transmitted and received frequencies.
- the TD-SCDMA system requires that the carrier frequency error of the base station is less than 0.05 PPM, and the carrier frequency error of the user terminal is required to be less than 0.1 PPM.
- the frequency accuracy of the oscillator can meet the requirements because the temperature, volume, power consumption, and cost of the device are relatively small. On the user side, due to various reasons, the frequency accuracy of the selected crystal oscillator usually does not meet the standard requirements.
- a general frequency offset estimation method is to perform correlation calculation on received training sequence data (such as Midamble code partial data) and a local training sequence (local Midamble code) to obtain an angular deviation within a certain period of time and then convert it into a frequency deviation.
- TD-SCDMA uses multiple Midamble code offsets to simultaneously transmit traffic
- multiple offset Midamble codes interfere with each other, especially the Midamble code has cyclic correlation characteristics; in addition, TD-SCDMA
- the system can use the same-frequency networking.
- the same-frequency networking will bring about the same-frequency interference while improving the spectrum utilization.
- the same-frequency interference exists in some data of the Midamble code. For the above reasons, if you directly receive the partial data of the received Midamble code with the local The Midamble code calculates the frequency offset through the correlation operation, which has a large error, which greatly reduces the terminal receiving performance, and even does not work normally.
- the technical problem to be solved by the present invention is to provide a frequency offset estimating apparatus and method for improving the accuracy of frequency offset estimation.
- the present invention provides a frequency offset estimation method, including: performing multi-cell channel estimation to obtain channel estimation values of each cell; detecting an activated training sequence offset according to channel estimation values of each cell; The frequency offset estimation value of the activated training sequence offset; wherein, obtaining the frequency offset estimation value of the activated training sequence offset, the method includes: removing, in the received training sequence data, the training sequence offset except the one Obtaining interference of the other activated training sequence offsets, obtaining a clean signal corresponding to the activated training sequence offset; acquiring the activated training by using the activated training sequence offset corresponding clean signal The frequency offset estimation value of the sequence offset; the final frequency offset estimation value is obtained according to the frequency offset estimation value of the plurality of activated training sequence offsets.
- the step of obtaining the clean signal corresponding to the activated training sequence offset is performed in the frequency domain or the time domain, and the step of obtaining the clean signal corresponding to the activated training sequence offset comprises: reconstructing the other activated The training sequence offsets the corresponding signal, and removes the signal corresponding to the other activated training sequence offsets in the received training sequence data to obtain a clean signal corresponding to the activated training sequence offset.
- the step of obtaining the final frequency offset estimation value according to the frequency offset estimation values of the plurality of activated training sequence offsets includes: deviating the frequency offset estimation values of the plurality of activated training sequence offsets from the training sequence respectively The energy values of the channel estimation window corresponding to the shifted frequency offset estimation are multiplied and then added, and divided by the sum of the energy values of the channel estimation windows corresponding to the plurality of activated training sequence offsets to obtain a final frequency offset estimation. value.
- the frequency offset estimation values of the several training sequence offsets are the frequency offset estimation values of all the activated training sequence offsets, or the activation training
- the frequency offset estimation value of the first M active training sequence offsets with the largest channel estimation window energy value corresponding to the sequence offset is M, and the M is a preset value.
- Performing multi-cell channel estimation, and obtaining channel estimation values of each cell includes: a) reconstructing signals of other cells except the currently processed cell by using a basic training sequence of each cell and channel estimation values of each cell, initially The channel estimation value of each cell is 0; b) removing the reconstructed signals of other cells in the received training sequence data to obtain a clean signal of the currently processed cell; c) using the clean signal of the currently processed cell for channel estimation, The channel estimation value of the current processing cell is repeated; steps a to c are repeatedly performed, the number of repetitions is ⁇ , and T is a predetermined number of processing times, and finally the channel estimation value of each cell is obtained.
- the present invention further provides a frequency offset estimating apparatus, comprising: a multi-cell channel estimating unit configured to perform multi-cell channel estimation to obtain channel estimation values of each cell; and an offset detecting unit configured to perform channel estimation according to each cell a value, detecting an activated training sequence offset; an interference cancellation unit configured to obtain a clean signal corresponding to the plurality of activated training sequence offsets as follows: an activated training sequence offset, in the received training sequence data Removing interference of other activated training sequence offsets except for an activated training sequence offset, obtaining a clean signal corresponding to an activated training sequence offset; a frequency offset estimating unit configured to acquire several activations as follows The frequency offset estimation value of the training sequence offset: for an activated training sequence offset, using the activated signal sequence offset corresponding clean signal, obtaining the frequency offset estimation value of the activated training sequence offset And a frequency offset processing unit configured to obtain a plurality of frequency offset estimates obtained according to the frequency offset estimating unit The final frequency offset estimate is obtained.
- the interference cancellation unit is configured to obtain a clean signal corresponding to an activated training sequence offset in the time domain or the frequency domain as follows: reconstructing the signal corresponding to the other activated training sequence offset, in the received training sequence The signal corresponding to the other activated training sequence offset is removed from the data to obtain a clean signal corresponding to the activated training sequence offset.
- the frequency offset processing unit is configured to obtain a final frequency offset estimation value as follows: a frequency offset estimation value of the plurality of activated training sequence offsets and a channel estimation window corresponding to the training sequence offset respectively The energy values are multiplied and then added, and then divided by the sum of the energy values of the channel estimation windows corresponding to the plurality of activated training sequence offsets to obtain a final frequency offset estimation value.
- the frequency offset estimation unit is configured to acquire frequency offset estimation values of the plurality of training sequence offsets as follows: acquire frequency offset estimation values of all activated training sequence offsets, or obtain corresponding motion training offsets corresponding to each
- M is a preset value.
- the multi-cell channel estimation unit includes a main control sub-unit, a channel estimation sub-unit, and a storage sub-unit: the main control sub-unit is configured to: start channel estimation and record the current channel estimation processing times, and end channel estimation after reaching T times;
- the storage subunit is configured to: store channel estimation values of each cell, where initially, the channel estimation value of each cell is 0; and the channel estimation subunit is configured to: estimate channel estimation of each cell according to an indication of the main control unit
- the channel estimation subunit includes: a reconstruction module, configured to use a basic training sequence of each cell and a channel estimation value of each cell stored by the cell channel estimation value storage subunit, and reconstruct other than the currently processed cell a signal of each cell; an interference cancellation module configured to remove the reconstructed signals of the other cells in the received training sequence data to obtain a clean signal of the currently processed cell; and a channel estimation module configured to use the cleaned current cell Signal estimation And obtaining an channel estimation value of the currently processed cell; and an output module configured to output channel estimation values of all
- FIG. 1 is a flowchart of a frequency offset estimation method based on a Midamble code of the present invention
- FIG. 2 is a flowchart of a multi-cell channel estimation method according to the present invention
- FIG. 3 is a flowchart of a Midacode code offset detection method according to the present invention
- Fig. 5 is a block diagram of a frequency offset estimation apparatus of the present invention.
- the invention provides a frequency offset estimation method based on a Midamble code of a terminal of a TD-SCDMA system, which first calculates a channel estimation of each cell by using a multi-cell channel estimation process and detects a Midamble code offset used by each current cell, and then estimates according to the channel.
- the result and the Midamble code offset detection result reconstruct the respective Midamble code offset signals, eliminate the interference between the respective Midamble codes, and calculate the frequency offset estimation values corresponding to the respective activated Midamble code offsets by using the interference-cancelled Midamble code partial data.
- the frequency offset estimation method of the present invention includes the following steps: Step 101: Perform multi-cell channel estimation, and calculate channel estimation values of respective cells. Multi-cell channel estimation can use parallel interference cancellation or serial interference cancellation to calculate the channel estimation value of each cell more accurately through multi-stage processing.
- K the number of midamble codes that can be configured for a time slot
- Channel estimation value Calculate the energy of each channel estimation window of each cell. For each channel estimation window of each cell, accumulate its tap energy to obtain the energy of the corresponding channel estimation window, namely:
- the energy maximum value of each channel estimation window of each cell is found, and the maximum value is multiplied by a threshold coefficient to obtain a channel estimation window activation detection threshold, and the channel estimation window whose energy is greater than the threshold is considered to be activated.
- the method for specifically detecting the activated Midamble code offset is shown in Figure 3.
- Step 103 For each activated Midamble code offset, remove the other activated Midamble code offsets from the received training sequence data, and obtain the activated Midamble code offset. Net signal.
- a clean signal corresponding to the offset is calculated.
- the method is to reconstruct signals of other activated Midamble code offsets, and subtract the reconstructed other activated Midamble code offset signals from the received training sequence data to obtain a clean signal corresponding to the Midamble code offset, as follows:
- Received midamble net received midamble- ⁇ conv(mi damble(i , k), h(i , k)) where received_midamble is the received training sequence data (ie, Midamble partial data);
- Z conv(midamble(i,k),h(i,k)) is the reconstructed signal of all other activated Midamble code offsets, which is the interference data for the activated Midamble code offset; received midamble net is A clean signal of the currently active Midamble code offset.
- the above formula represents a method of obtaining a clean signal in the time domain.
- a clean signal can also be calculated in the frequency domain, as follows: First, the interference signal is reconstructed in the frequency domain, and the Midamble code and other activated Midamble codes are respectively shifted.
- the channel estimation is performed by fast Fourier transform to obtain ffl( nidamble(i, k ) and ffl( , k, , , and multiply the two by ffiimidambleii, k)) * fft(h(i, k)), and then perform FFT on received_midamble , get fft (received _ midamble), then perform interference cancellation, subtract in frequency domain, get the frequency i or value of the clean signal of the currently activated Midamble code offset rece wi/_ i/iTM ?
- Received _ midamble _ net ifft received _ midamble _ net _ fft
- Step 104 Perform frequency offset estimation of each activated Midamble code offset to obtain several frequencies. Partial estimate. For an activated Midamble code offset, calculate its frequency offset estimate. The frequency offset estimation is performed by using the clean signal corresponding to the activated Midamble code offset obtained in step 103, and the frequency offset estimation value Fr ⁇ OK corresponding to each activated Midamble code is obtained. FreqOffset ⁇ , k, % i The frequency offset estimation value corresponding to the kth channel estimation window of the cells. The basic principle of frequency offset estimation is to use the cleaned signal corresponding to the active Midamble code offset, correlate with the corresponding local Midamble code, calculate the phase difference of a certain chip degree, and then according to the chip rate of the TD-SCDMA system 1.28M.
- a specific frequency offset estimation method for an activated Midamble code offset is shown in FIG.
- steps 103 and 104 are repeated until all frequency offset estimates for the activated Midamble code offset are calculated, or only the first M of the energy of the channel estimation window of the active Midamble code offset are calculated.
- the frequency offset estimate corresponding to the activated Midamble code offset, M is a preset value, which can be determined according to the performance and the cost of hardware and software implementation. It is also possible to calculate only the frequency offset estimation value of the Midamble code offset with the strongest signal energy, that is, the frequency offset estimation value of the Midamble code offset with the largest channel estimation window energy.
- the number of frequency offset estimates obtained in this step may be the frequency offset estimation value of all activated Midamble code offsets, or may be the frequency offset estimation value of the active Midamble code offset with the highest energy of the first M channel estimation windows. It may also be a frequency offset estimation value of the activated Midamble code offset with the largest channel estimation window energy.
- Step 105 The final frequency offset estimation value is obtained according to the frequency offset estimation values of the plurality of activated Midamble code offsets obtained in step 104.
- the following methods are as follows: 1) Combining the channel estimation window energy corresponding to the activated Midamble code offsets to obtain a final frequency offset estimation value, and the specific combining method is: biasing the plurality of activated Midamble codes
- the shifted frequency offset estimation values are respectively multiplied by the energy values of the corresponding channel estimation windows, and then divided by the sum of the energy values of the channel estimation windows corresponding to the plurality of activated Midamble code offsets.
- the final frequency offset estimate is as follows:
- Fr ⁇ Q ⁇ et ⁇ ) refers to the frequency offset estimation value of several activated Midamble code offsets, which is the channel estimation window energy corresponding to the corresponding Midamble code offset, which is the cell index number; k Is the Midamble code offset index number.
- Step 201 Using a basic Midamble code (ie, a chip sequence of 128 chip length defined by a protocol) and an existing channel estimation value (ie, calculated by channel estimation) The 128-tap channel estimation value is obtained, and the signals of the cells other than the current processing cell are reconstructed, and the initial channel estimation value of each cell is all 0; and Z conv (channel, basic_midamble) is used to indicate other than the currently processed cell. The signal of each cell. Where channel is the current channel estimate, initially 0; basic_midamble is the basic Midamble code.
- Step 202 Perform interference cancellation according to the signal reconstructed in step 201, cancel out the signals of the reconstructed other cells, and obtain a relatively clean time domain signal.
- Received_midamble_netl received—midamble- ⁇ conv(channel,basic_midamble) where,
- Received_midamble_netl is a relatively clean time i or signal after interference with 4 s; the received_midamble is the received training sequence data (ie, the midamble partial data).
- Step 203 Perform channel estimation by using the relatively clean time domain signal obtained in step 202. First calculate the frequency domain value of the 128-chip (chip) Midamble partial data received_midamble_netl.
- Channel' ifft(received_midamble_netl_fft - /basic_midamble_fft) where ⁇ / indicates the division of the corresponding number of two arrays, basic_midamble_fft is the frequency domain value of the currently used basic Midamble code (that is, the value obtained by FFT the basic midamble code).
- the channel estimation value channel' is subjected to channel estimation and denoised to obtain a channel estimation value channel of each cell; steps 201 to 203 are repeated until all levels 4 are processed.
- the multi-cell channel estimation value can also be obtained by other methods, and the example shown in FIG. 2 is only an example, which is not limited by the present invention.
- Step 304 Set the energy threshold according to the energy maximum:
- Step 330055 The root judges the activation behavior of the MMiiddaammbbllee code offset shift according to the threshold of the energy energy threshold. . For example, if ⁇ (( )) ⁇ ⁇ , then the kkth MMiiddaammbbllee code code offset migration activation activity of the iith small cell area is judged, and otherwise, no activation is activated. live. .
- FIG. 44 is a flow chart showing the frequency offset estimation method of the MMiiddaammbbllee code code offset shift of each of the activation activations of the present invention.
- Step 440011 : Calculate the correlation correlation between the two sections of the front and back of the MMIiddaammbbllee before and after the calculation of the path of each of the strips, according to the public formula below. ::
- Ii is the small cell area index quotation mark;
- kk is the MMiiddaammbbllee code code offset migration index quotation mark when the current front processing is activated;
- rreecceeiivveedd mmiiddaammbbllee nneett is when the current pre-activated
- the live MMiiddaammbbllee code offsets the clean signal signal.
- the 1155 PPrreevvCCoorrrr table indicates the result of the correlation of the first half of the MMiiddaammbbllee.
- the SSuuccccCCoorrrr table indicates the result of the phase-related correlation of the second half of the MMiiddaammbbllee, and the ssuumm table indicates the summation.
- Step 440022 The calculation calculation table represents the number of complex phase differences of the phase correlations of the phase correlation before and after the two segments.
- MM is the largest road to support and merge
- the number of path paths is limited by the source resource limit of the actual implementation.
- phase deviation sign calculated for each time slot also requires a moving average to filter the effects of noise.
- the length of the sliding average window is indicated by N. Assuming that the current subframe number for the frequency offset estimation is ⁇ , the result of the multi-subframe average is:
- the present invention further provides a frequency offset estimation apparatus, as shown in FIG. 5, including a multi-cell channel estimation unit, an offset detection unit, an interference cancellation unit, a frequency offset estimation unit, and a frequency offset processing unit, where: a multi-cell channel estimation unit And configured to perform multi-cell channel estimation to obtain channel estimation values of each cell; and an offset detecting unit configured to detect an activated training sequence offset according to channel estimation values of each cell; and an interference cancellation unit configured to be An activated training sequence offset, in the received training sequence data, removing interference of other activated training sequence offsets in the received training sequence data, and obtaining a clean signal corresponding to the activated training sequence offset; specifically, The interference cancellation unit reconstructs signals corresponding to the other activated training sequence offsets, and removes signals corresponding to the other activated training sequence offsets in the received training sequence data, to obtain a clean signal corresponding to the activated training sequence offset; a unit that is set to obtain a frequency offset estimate of a number of activated training sequence offsets And acquiring, by using the activated
- the multi-cell channel estimation unit includes a main control sub-unit, a channel estimation sub-unit, and a storage sub-unit: a main control sub-unit configured to start channel estimation and record the current channel estimation processing times, and end channel estimation after reaching T times, T a pre-specified number of processing times; a storage subunit configured to store a channel estimation value of each cell, where the channel estimation value of each cell is initially 0; a channel estimation subunit configured to estimate each according to an indication of the main control unit
- the channel estimation value of the cell including:
- a reconstruction module configured to use a basic training sequence of each cell and a channel estimation value of each cell stored in the storage subunit to reconstruct signals of other cells except the currently processed cell; and an interference cancellation module: configured to receive The training sequence data is used to remove the reconstructed signals of the other cells to obtain a clean signal of the currently processed cell; the channel estimation module: configured to perform channel estimation using the clean signal of the currently processed cell, to obtain a channel estimation value of the currently processed cell; Output Module: It is arranged to output the channel estimation values of all cells estimated by the channel estimation module to the storage subunit.
- the frequency offset estimating apparatus and method according to the present invention removes interference between Midamble codes, performs frequency offset estimation using a clean signal, and improves frequency offset estimation accuracy.
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Mobile Radio Communication Systems (AREA)
- Cable Transmission Systems, Equalization Of Radio And Reduction Of Echo (AREA)
Abstract
本发明提供了一种频偏估计方法，包括：进行多小区信道估计，获得各小区的信道估计值；根据各小区的信道估计值，检测激活的训练序列偏移；获取若干个激活的训练序列偏移的频偏估计值；其中，获取一激活的训练序列偏移的频偏估计值包括：在接收的训练序列数据中去除其他激活的训练序列偏移的干扰，得到所述一激活的训练序列偏移对应的干净信号；使用一激活的训练序列偏移对应的干净信号，获取一激活的训练序列偏移的频偏估计值；根据若干个激活的训练序列偏移的频偏估计值得到最终的频偏估计值。本发明还提供一种频偏估计装置。本发明所述方法，通过去除 Midamble码之间的干扰使用干净信号进行频偏估计，提高了频偏估计精度。
Description
一种频偏估计装置和方法
技术领域 本发明涉及无线通信系统， 具体涉及一种频偏估计装置和方法。
背景技术
在时分同步码分多址接入 ( Time Division Synchronous Code Division Multiple Access, TD-SCDMA ) 系统中， 基站和终端均以标称的载波频率进 行发送和接收。受器件的限制， 实际发送和接收的频率之间会有一定的偏差。 TD-SCDMA系统要求基站的载波频率误差小于 0.05PPM,要求用户端的载波 频率误差小于 0.1PPM。 在基站端， 由于器件的温度、 体积、 功耗和成本等的 限制比较小， 振荡器的频率精度可以满足要求。 而在用户端， 受到前面各种 原因限制， 所选用的晶体振荡器的频率精度通常不满足标准要求。 为了获得较高的通讯信号质量， 需要对频率偏差进行估计， 并且通过自 动频率控制进行频率跟踪和补偿， 使得用户端的载波频率误差能够满足系统 要求。 一般的频偏估计方法是对接收的训练序列数据 (比如 Midamble码部分 数据）与本地的训练序列 （本地 Midamble码）进行相关运算， 得到一定时 间内的角度偏差， 再换算成频率偏差。
发明内容 当 TD-SCDMA系统在同时使用多个 Midamble码偏移来传输业务的情况 下， 多个偏移的 Midamble码互为干扰， 特别是 Midamble码存在循环相关特 性； 另夕卜， TD-SCDMA系统为了提高频谱利用率， 可釆用同频组网， 但是同 频组网在提高频谱利用率的同时，会带来同频干扰， Midamble码部分数据上 也存在同频干扰。 基于以上原因， 如果直接将接收到的 Midamble 码部分数据与本地的
Midamble码通过相关运算计算频偏，会有较大的误差，极大的降低终端接收 性能， 甚至不能正常工作。 本发明要解决的技术问题是提供一种频偏估计装置和方法， 提高频偏估 计精度。
为了解决上述问题， 本发明提供了一种频偏估计方法， 包括： 进行多小区信道估计， 获得各小区的信道估计值； 根据各小区的信道估计值， 检测激活的训练序列偏移； 获取若干个激活的训练序列偏移的频偏估计值； 其中， 获取一激活的训 练序列偏移的频偏估计值， 其包括： 在接收的训练序列数据中去除除所述一 激活的训练序列偏移之外的其他激活的训练序列偏移的干扰， 得到所述一激 活的训练序列偏移对应的干净信号； 使用所述一激活的训练序列偏移对应的 干净信号， 获取所述一激活的训练序列偏移的频偏估计值； 才艮据所述若干个激活的训练序列偏移的频偏估计值得到最终的频偏估计 值。 得到所述一激活的训练序列偏移对应的干净信号的步骤是在频域或时域 进行， 得到所述一激活的训练序列偏移对应的干净信号的步骤包括： 重构所 述其他激活的训练序列偏移对应的信号， 在接收的训练序列数据中去除所述 其他激活的训练序列偏移对应的信号， 得到所述一激活的训练序列偏移对应 的干净信号。 才艮据所述若干个激活的训练序列偏移的频偏估计值得到最终的频偏估计 值的步骤包括： 将所述若干个激活的训练序列偏移的频偏估计值分别和训练序列偏移的 频偏估计值对应的信道估计窗的能量值相乘后相加， 再除以所述若干个激活 的训练序列偏移对应的信道估计窗的能量值的和， 得到最终的频偏估计值。 获取若干个训练序列偏移的频偏估计值的步骤中， 若干个训练序列偏移 的频偏估计值是所有激活的训练序列偏移的频偏估计值， 或者是各激活的训
练序列偏移对应的信道估计窗能量值最大的前 M个激活的训练序列偏移的 频偏估计值， M为一预设值。 进行多小区信道估计， 获得各小区的信道估计值的步骤包括： a)使用各小区的基本训练序列和各小区的信道估计值， 重构除当前处理 小区外的其他各小区的信号， 初始时各小区的信道估计值为 0; b)在接收的训练序列数据中去除重构的其他各小区的信号， 得到当前处 理小区的干净信号； c)使用当前处理小区的干净信号进行信道估计， 得到当前处理小区的信 道估计值； 重复执行步骤 a至 c, 重复次数为 Τ, T为预先指定的处理次数， 最终得 到各小区的信道估计值。
本发明还提供一种频偏估计装置， 包括： 多小区信道估计单元， 其设置为进行多小区信道估计， 获得各小区的信 道估计值； 偏移检测单元， 其设置为根据各小区的信道估计值， 检测激活的训练序 列偏移； 干扰消除单元， 其设置为按如下方式获得若干个激活的训练序列偏移对 应的干净信号： 对一激活的训练序列偏移， 在接收的训练序列数据中去除除 一激活的训练序列偏移之外的其他激活的训练序列偏移的干扰， 得到一激活 的训练序列偏移对应的干净信号； 频偏估计单元， 其设置为按如下方式获取若干个激活的训练序列偏移的 频偏估计值： 对一激活的训练序列偏移， 使用所述一激活的训练序列偏移对 应的干净信号， 获取所述一激活的训练序列偏移的频偏估计值； 以及 频偏处理单元， 其设置为根据所述频偏估计单元获取的若干个频偏估计 值得到最终的频偏估计值。
所述干扰消除单元是设置为在时域或频域按如下方式得到一激活的训练 序列偏移对应的干净信号： 重构所述其他激活的训练序列偏移对应的信号， 在接收的训练序列数据 中去除所述其他激活的训练序列偏移对应的信号， 得到一激活的训练序列偏 移对应的干净信号。 所述频偏处理单元是设置为按如下方式得到最终的频偏估计值： 将所述 若干个激活的训练序列偏移的频偏估计值分别和所述训练序列偏移对应的信 道估计窗的能量值相乘后相加， 再除以所述若干个激活的训练序列偏移对应 的信道估计窗的能量值的和， 得到最终的频偏估计值。 所述频偏估计单元是设置为按如下方式获取若干个训练序列偏移的频偏 估计值： 获取所有激活的训练序列偏移的频偏估计值， 或者获取各激活的训 练序列偏移对应的信道估计窗能量值最大的前 M个激活的训练序列偏移的 频偏估计值， M为一预设值。 所述多小区信道估计单元包括主控子单元、 信道估计子单元和存储子单 元： 所述主控子单元设置为： 启动信道估计和记录当前信道估计处理次数， 到达 T次后结束信道估计； 所述存储子单元设置为： 存储各小区的信道估计值， 初始时各小区的信 道估计值为 0; 以及 所述信道估计子单元设置为： 根据主控单元的指示， 估计各小区的信道 估计值， 所述信道估计子单元包括： 重构模块， 其设置为使用各小区的基本训练序列和小区信道估计值 存储子单元存储的各小区的信道估计值， 重构除当前处理小区外的其他 各小区的信号； 干扰抵消模块， 其设置为在接收的训练序列数据中去除重构的其他 各小区的信号， 得到当前处理小区的干净信号； 信道估计模块， 其设置为使用当前处理小区的干净信号进行信道估
计， 得到当前处理小区的信道估计值； 以及 输出模块， 其设置为将信道估计模块估计的所有小区的信道估计值 输出至存储单元。
附图概述 图 1是本发明基于 Midamble码的频偏估计方法流程图； 图 2是本发明多小区信道估计方法流程图； 图 3是本发明 Midamble码偏移检测方法流程图； 图 4是本发明各 Midamble码偏移的频偏估计流程图； 图 5是本发明频偏估计装置系统框图。
本发明的较佳实施方式 以下将结合附图， 对本发明各较佳实施例进行较为详细的说明。 本发明提出了一种 TD-SCDMA系统的终端基于 Midamble码的频偏估计 方法， 首先通过多小区信道估计处理计算各个小区的信道估计和检测当前各 个小区使用的 Midamble码偏移， 然后根据信道估计结果和 Midamble码偏移 检测结果重构各个 Midamble码偏移信号 ,消除各个 Midamble码之间的干扰， 再用干扰消除后的 Midamble码部分数据计算各个激活的 Midamble码偏移对 应的频偏估计值， 最后通过合并各个激活的 Midamble码偏移对应的频偏估
下面通过一实施例进一步详细说明本发明。 本实施例以 = = 8为
例， 其中， P = 128为基本 Midamble码长度， = 16为信道估计窗长度； 干扰 ^氐消为 4级， 1个本小区 3个邻小区， 一个小区分配第 1、 2两个 Midamble 码偏移。
如图 1所示， 本发明的频偏估计方法包括如下步骤： 步骤 101 : 进行多小区信道估计， 计算出各个小区的信道估计值。 多小区信道估计可以釆用并行干扰消除或者串行干扰消除， 通过多级处 理， 来比较准确的计算各个小区的信道估计值。 若当前处理 N个小区（N值可以根据性能和软硬件开销折中选取， 通常 为 2 ~ 5 , 本实施例中处理 1个本小区 3个邻小区， 即 Ν=4 ) , 系统参数 K 值为 8 (其中， K为一个时隙可以配置使用的 midamble码数目 ） ， 多小区信 道估计结果为 channel(i ), 其中， i=l~N, 是小区索引号， j=l〜： 128, 是抽头 索引号。 多小区信道估计的具体方法见图 2。
步骤 102: 检测激活的 Midamble码偏移； 各个小区信道估计分窗处理： h(i, k) = channel , (A - 1) * 16 + (1： 16》 其中， i=l,2,... ,N,为小区索引号； k=l,2,... ,8,为信道估计窗索引号； h(i, k) 是指第 i个小区的第 k个信道估计窗的 16个信道估计值。 计算各个小区各个信道估计窗的能量。 对各小区的各信道估计窗， 将其 抽头能量累加， 得到对应信道估计窗的能量， 即：
P(i, A) = h(i, k). * conj(h(i, k)) 是指第 i个小区的第 k个信道估计窗的能量， 其中， 跳 /z ^)的共轭， i=l, 2,... , N; k=l,2,... ,8。 找出各小区的各信道估计窗的能量最大值， 将该最大值乘以一个门限系 数，得到信道估计窗激活检测门限， 能量大于该门限的信道估计窗认为激活。 具体检测激活的 Midamble码偏移的方法见图 3。
步骤 103: 对每一激活的 Midamble码偏移， 在接收的训练序列数据中去 除其他激活的 Midamble码偏移的干 4尤 ,得到该激活的 Midamble码偏移的干
净信号。
对于某一个小区的某一个激活的 Midamble码偏移， 计算该偏移对应的 干净信号。 方法是重构其他激活的 Midamble码偏移的信号， 在接收的训练 序列数据中减去重构的其他激活的 Midamble 码偏移的信号， 得到该 Midamble码偏移对应的干净信号， 方法如下：
received midamble net = received midamble- ^ conv(mi damble(i , k), h(i , k)) 其中， received_midamble是接收的训练序列数据（即 Midamble部分数据）；
//( )是除待计算的激活的 Midamble码偏移外的其他激活的 Midamble 码偏移对应的信道估计； midamble(i,k)是其对应的 Midamble码， 即第 i个小 区的 Midamble码。
Z conv(midamble(i,k),h(i,k))是重构的所有其他激活的 Midamble码偏移的 信号， 该信号是对本激活的 Midamble码偏移的干扰数据； received midamble net是当前激活的 Midamble码偏移的干净信号。 上式表示的是在时域中得到干净信号的方法， 当然也可以在频域中计算 干净信号， 具体如下： 首先在频域重构干扰信号， 分别对 Midamble码和其它激活的 Midamble 码偏移的信道估计进行快速傅立叶变换得到 ffl( nidamble(i, k )和 ffl( , k、、 , 将 二者相乘 ffiimidambleii, k)) * fft(h(i, k))后， 对 received_midamble进行 FFT , 得到 fft(received _ midamble) , 再进行干扰抵消 , 在频域相减 , 得到当前激活的 Midamble码偏移的干净信号的频 i或值 rece wi/_ i/i™ ? _w _ received _ midamble _ net _ fft = fft received _ midamble) - ^ fft midamble( , k)) * fft h i, k)) received _ midamble _ net _ fft变换到时域，从而得到当前激活的 Midamble码偏 移的干净信号
received _ midamble _ net = ifft received _ midamble _ net _ fft)
步骤 104: 进行各个激活的 Midamble码偏移的频偏估计， 得到若干个频
偏估计值。 对于一个激活的 Midamble码偏移， 计算其频偏估计值。 用步骤 103中 得到的该激活的 Midamble码偏移对应的干净信号进行频偏估计， 得到各个 激活的 Midamble码对应的频偏估计值 Fr^OK )。 FreqOffset{ , k、 % i个 小区第 k个信道估计窗对应的频偏估计值。 频偏估计的基本原理是用激活的 Midamble码偏移对应的干净信号， 与 对应的本地 Midamble 码相关， 再计算一定码片程度上的相位差， 再根据 TD-SCDMA系统的码片速率 1.28M, 可以计算 lchip (码片 ) 的时间， 再将 相位差换算到频率。 频偏估计的具体实现方法比较成熟， 这里不再赘述。 具 体的一种激活的 Midamble码偏移的频偏估计方法见图 4。 对于各个激活的 Midamble码偏移， 重复步骤 103和 104, 直到所有的激 活 Midamble码偏移的频偏估计计算完，或者只计算激活的 Midamble码偏移 的信道估计窗的能量最大的前 M个激活的 Midamble码偏移对应的频偏估计 值， M为一预设值， 可以根据性能和软硬件实现的代价确定。 也可以只计算 信号能量最强的 Midamble码偏移的频偏估计值， 即信道估计窗能量最大的 Midamble码偏移的频偏估计值。
即本步骤中得到的若干个频偏估计值， 可以是所有激活的 Midamble码 偏移的频偏估计值， 也可以是前 M个信道估计窗能量最大的激活 Midamble 码偏移的频偏估计值， 也可以是信道估计窗能量最大的激活 Midamble码偏 移的频偏估计值。
步骤 105:才艮据步骤 104中得到的该若干个激活的 Midamble码偏移的频 偏估计值得到最终的频偏估计值。
具体有如下几种方法： 1 )根据该若干个激活的 Midamble码偏移对应的信道估计窗能量进行合 并， 得到最后的频偏估计值， 具体合并方法为： 将该若干个激活的 Midamble 码偏移的频偏估计值分别和其对应的信道估计窗的能量值相乘后相加， 再除 以该若干个激活的 Midamble码偏移对应的信道估计窗的能量值的和， 得到
最终的频偏估计值， 如下式所示:
―∑∑響 TO
∑∑P(^) 其中， Fr^Q^et ^)指若干个激活的 Midamble码偏移的频偏估计值， 为对应的 Midamble码偏移对应的信道估计窗能量， 是小区索引号； k 是 Midamble码偏移索引号。
2 )也可以不合并， 将信号能量最强的 Midamble码偏移对应的计算出来 的频偏估计值作为最后的频偏估计值， 或者釆用其他处理方法得到最终的频 偏估计值。
图 2所示为多小区信道估计方法流程图， 具体包括： 步骤 201 : 用基本 Midamble码（即协议定义的一个长度为 128chip的码 片序列）和已有的信道估计值（即通过信道估计计算出来的 128抽头的信道 估计值） ， 重构除当前处理小区以外的其他各小区的信号， 各个小区的初始 信道估计值为全 0; 用 Z conv(channel,basic_midamble)表示当前处理小区以外的其他各小区的 信号。 其中， channel是当前的信道估计值， 初始为 0 ; basic_midamble是基本 Midamble码。 步骤 202: 根据步骤 201 中重构的信号做干扰抵消， 抵消掉重构的其他 小区的信号， 得到相对干净的时域信号；
received_midamble_netl = received—midamble- ^ conv(channel,basic_midamble) 其中，
received_midamble_netl是干扰 4氏消之后的相对干净的时 i或信号； received_midamble是接收的训练序列数据 (即 midamble部分数据 ) 。
步骤 203: 用步骤 202得到的相对干净的时域信号做信道估计； 先计算 128chip (码片） 的 Midamble部分数据 received_midamble_netl的频 域值 received midamble net 1 fft： received midamble net 1 fft = _ ft(received_midamble_netl)
channel' = ifft(received_midamble_netl_fft - /basic_midamble_fft) 其中 · /表示两个数组对应数相除， basic_midamble_fft是当前使用的基本 Midamble码的频域值 (即对基本 midamble码进行 FFT后得到的值 ) 。 信道估计值 channel'经过信道估计后处理去噪得到各个小区的信道估计值 channel; 重复步骤 201 ~ 203直到 4级全部处理完。 多小区信道估计值也可釆用其他方法得到， 图 2所示仅为示例， 本发明 对此不作限定。
图 3所示为 Midamble码偏移检测方法流程图， 具体包括： 步骤 301 : 对各个小区信道估计分窗处理： h(i, k) = channel , (A - 1) * 16 + (1： 16》 其中， i=l,2,... ,4, 为小区索引号; k=l,2,... ,8, 为信道窗索引号； 步骤 302: 计算各个小区各个信道估计窗的能量： P(i, A) = h(i, k). * conj(h(i, k)) 步骤 303: 找出各小区各信道估计窗能量最大值：
步骤 304: 根据能量最大值设置能量门限 ：
其中， Th是可配置门限系数， 默认值 0.25 , 可以根据仿真或者测试作适
当调整。 步步骤骤 330055:: 根根据据能能量量门门限限 判判断断 MMiiddaammbbllee码码偏偏移移的的激激活活情情况况。。 如如果果 ΡΡ(( ))≥≥ ,, 则则判判决决第第 ii个个小小区区的的第第 kk个个 MMiiddaammbbllee码码偏偏移移激激活活，， 否否 则则未未激激活活。。
图图 44所所示示为为本本发发明明各各个个激激活活的的 MMiiddaammbbllee码码偏偏移移的的频频偏偏估估计计方方法法流流程程图图 ,, 具具体体包包括括：： 步步骤骤 440011：： 对对每每一一条条路路径径计计算算接接收收 MMiiddaammbbllee前前后后两两段段的的相相关关 按按照照下下面面的的公公式式计计算算：：
1100 DDeeSSccrrDDaattaa == rreecceeiivveedd mmiiddaammbbllee nneett ** ccoonnjj((MMiiddaammbbllee((ii,,kk))));;
ii 是是小小区区索索引引号号；； kk 是是当当前前处处理理的的激激活活的的 MMiiddaammbbllee 码码偏偏移移索索引引号号；； rreecceeiivveedd mmiiddaammbbllee nneett是是当当前前激激活活的的 MMiiddaammbbllee码码偏偏移移的的干干净净信信号号。。
PPrreevvCCoorrrr == ssuumm((DDeeSSccrrDDaattaa((ll ::6644))));;
SSuuccccCCoorrrr == ssuumm((DDeeSSccrrDDaattaa((6655:: 112288))));;
1155 PPrreevvCCoorrrr表表示示 MMiiddaammbbllee前前半半段段相相关关结结果果，， SSuuccccCCoorrrr表表示示 MMiiddaammbbllee后后 半半段段相相关关结结果果，， ssuumm表表示示求求和和。。 步步骤骤 440022:: 计计算算表表示示前前后后两两段段的的相相关关的的相相位位差差的的复复数数
DDaattaaOOffffsseett == ccoonnjj ((PPrreevvCCoorrrr ))** SSuuccccCCoorrrr;; 步步骤骤 440033 :: 计计算算多多条条路路径径合合并并的的结结果果
2200 假假设设不不同同路路径径计计算算的的相相位位差差复复数数为为 DDaattaaOOffffsseett((mm))，，其其中中 mm表表示示选选取取第第 mm 条条无无线线路路径径
AAddddOOffffsseett == DDaattaaOOffffsseett((ll)) ++ DDaattaaOOffffsseett((22)) ++ ......DDaattaaOOffffsseett ((MM));; 其其中中 MM是是支支持持合合并并的的最最大大路路径径数数目目，，受受到到实实现现的的资资源源限限制制，，缺缺省省取取 ΛΛ//==22,, 支支持持两两条条路路径径合合并并。。 合合并并方方式式釆釆用用等等增增益益合合并并。。
2255 *
每个时隙计算得到的相位偏差符号还需要进行滑动平均以过滤噪声的影 响。 滑动平均窗的长度以 N表示。 假设当前进行频偏估计的子帧号是 ^ 则 多子帧平均的结果为：
AvgOffset(k) = (AddOffset(k) + AddOffset(k-l)...+ AddOffset(k-N+l); 其中 N是滑动平均的子帧数目， 取值可以配置为 1 , 2或 4。 步骤 405: 计算相位偏差 第 i个小区的第 k个激活的 Midamble码偏移对应的相位偏差计算如下： PhaseOffset(i,k) = arctan(imag(AvgOffset)/ real (AvgOffset))/64; 步骤 406: 计算频偏 FreqOffset(i,k) = PhaseOffset(i,k)/ (2*π*(128/2)) * 1280000。
本发明还提供一种频偏估计装置， 如图 5所示， 包括多小区信道估计单 元、 偏移检测单元、 干扰消除单元、 频偏估计单元和频偏处理单元， 其中： 多小区信道估计单元， 其设置为进行多小区信道估计， 获得各小区的信 道估计值； 偏移检测单元， 其设置为根据各小区的信道估计值， 检测激活的训练序 列偏移； 干扰消除单元， 其设置为对一激活的训练序列偏移， 在接收的训练序列 数据中在接收的训练序列数据中去除其他激活的训练序列偏移的干扰， 得到 该激活的训练序列偏移对应的干净信号； 具体是指， 干扰消除单元重构其他 激活的训练序列偏移对应的信号， 在接收的训练序列数据中去除其他激活的 训练序列偏移对应的信号， 得到激活的训练序列偏移对应的干净信号； 频偏估计单元，其设置为获取若干个激活的训练序列偏移的频偏估计值， 对一激活的训练序列偏移， 使用该激活的训练序列偏移对应的干净信号， 获 取该激活的训练序列偏移的频偏估计值； 所述频偏估计单元获取若干个训练 序列偏移的频偏估计值是指， 获取全部激活的训练序列偏移的频偏估计值，
或者，获取各激活的训练序列偏移对应的信道估计窗能量值最大的前 M个激 活的训练序列偏移的频偏估计值， M为一预设值； 频偏处理单元， 其设置为对频偏估计单元获取的若干个频偏估计值进行 处理， 得到最终的频偏估计值。 对频偏估计值的处理方法见方法实施例中所 述。 所述多小区信道估计单元包括主控子单元、 信道估计子单元和存储子单 元： 主控子单元， 其设置为启动信道估计和记录当前信道估计处理次数， 到 达 T次后结束信道估计， T为预先指定的处理次数； 存储子单元， 其设置为存储各小区的信道估计值， 初始时各小区的信道 估计值为 0; 信道估计子单元， 其设置为根据主控单元的指示， 估计各小区的信道估 计值， 包括：
重构模块： 其设置为使用各小区的基本训练序列和存储子单元存储 的各小区的信道估计值， 重构除当前处理小区外的其他各小区的信号； 干扰抵消模块： 其设置为在接收的训练序列数据中去除重构的其他 各小区的信号， 得到当前处理小区的干净信号； 信道估计模块： 其设置为使用当前处理小区的干净信号进行信道估 计， 得到当前处理小区的信道估计值； 输出模块： 其设置为将信道估计模块估计的所有小区的信道估计值 输出至存储子单元。
工业实用性 本发明所述的频偏估计装置和方法， 去除 Midamble码之间的干扰， 使 用干净信号进行频偏估计， 提高了频偏估计精度。
Claims
1、 一种频偏估计方法， 其包括: 进行多小区信道估计， 获得各小区的信道估计值； 根据各小区的信道估计值， 检测激活的训练序列偏移； 获取若干个激活的训练序列偏移的频偏估计值； 其中， 获取一激活的训 练序列偏移的频偏估计值包括： 在接收的训练序列数据中去除除所述一激活 的训练序列偏移之外的其他激活的训练序列偏移的干扰， 得到所述一激活的 训练序列偏移对应的干净信号； 使用所述一激活的训练序列偏移对应的干净 信号， 获取所述一激活的训练序列偏移的频偏估计值； 才艮据所述若干个激活的训练序列偏移的频偏估计值得到最终的频偏估计 值。
2、 如权利要求 1所述的方法， 其中， 得到所述一激活的训练序列偏移对应的干净信号的步骤是在频域或时域 进行； 得到所述一激活的训练序列偏移对应的干净信号的步骤包括： 重构所述其他激活的训练序列偏移对应的信号， 在接收的训练序列数据 中去除所述其他激活的训练序列偏移对应的信号， 得到所述一激活的训练序 列偏移对应的干净信号。
3、如权利要求 1或 2所述的方法， 其中，根据所述若干个激活的训练序 列偏移的频偏估计值得到最终的频偏估计值的步骤包括： 将所述若干个激活的训练序列偏移的频偏估计值分别和训练序列偏移的 频偏估计值对应的信道估计窗的能量值相乘后相加， 再除以所述若干个激活的训练序列偏移对应的信道估计窗的能量值的 和,
得到最终的频偏估计值。
4、如权利要求 1或 2所述的方法， 其中， 获取若干个训练序列偏移的频 偏估计值的步骤中， 若干个训练序列偏移的频偏估计值是所有激活的训练序 列偏移的频偏估计值， 或者是各激活的训练序列偏移对应的信道估计窗能量 值最大的前 M个激活的训练序列偏移的频偏估计值， M为一预设值。
5、 如权利要求 1所述的方法， 其中， 进行多小区信道估计， 获得各小区 的信道估计值的步骤包括：
a)使用各小区的基本训练序列和各小区的信道估计值， 重构除当前处理 小区外的其他各小区的信号， 初始时各小区的信道估计值为 0;
b)在接收的训练序列数据中去除重构的其他各小区的信号， 得到当前处 理小区的干净信号； c)使用当前处理小区的干净信号进行信道估计， 得到当前处理小区的信 道估计值； 重复执行步骤 a至 c, 重复次数为 Τ, T为预先指定的处理次数， 最终得到各小区的信道估计值。
6、 一种频偏估计装置， 其包括： 多小区信道估计单元， 其设置为进行多小区信道估计， 获得各小区的信 道估计值； 偏移检测单元， 其设置为根据各小区的信道估计值， 检测激活的训练序 列偏移； 干扰消除单元， 其设置为按如下方式获得若干个激活的训练序列偏移对 应的干净信号： 对一激活的训练序列偏移， 在接收的训练序列数据中去除除 一激活的训练序列偏移之外的其他激活的训练序列偏移的干扰， 得到一激活 的训练序列偏移对应的干净信号； 频偏估计单元， 其设置为按如下方式获取若干个激活的训练序列偏移的 频偏估计值： 对一激活的训练序列偏移， 使用所述一激活的训练序列偏移对
应的干净信号， 获取所述一激活的训练序列偏移的频偏估计值； 以及 频偏处理单元， 其设置为根据所述频偏估计单元获取的若干个频偏估计 值得到最终的频偏估计值。
7、如权利要求 6所述的装置， 其中， 所述干扰消除单元是设置为在时域 或频域按如下方式得到一激活的训练序列偏移对应的干净信号： 重构所述其他激活的训练序列偏移对应的信号， 在接收的训练序列数据 中去除所述其他激活的训练序列偏移对应的信号， 得到一激活的训练序列偏 移对应的干净信号。
8、如权利要求 6或 7所述的装置， 其中， 所述频偏处理单元是设置为按 如下方式得到最终的频偏估计值：
将所述若干个激活的训练序列偏移的频偏估计值分别和所述训练序列偏 移对应的信道估计窗的能量值相乘后相加， 再除以所述若干个激活的训练序列偏移对应的信道估计窗的能量值的
9、如权利要求 6所述的装置， 其中， 所述频偏估计单元是设置为按如下 方式获取若干个训练序列偏移的频偏估计值： 获取所有激活的训练序列偏移的频偏估计值， 或者 获取各激活的训练序列偏移对应的信道估计窗能量值最大的前 M个激 活的训练序列偏移的频偏估计值， M为一预设值。
10、 如权利要求 6所述的装置， 其中， 所述多小区信道估计单元包括主 控子单元、 信道估计子单元和存储子单元： 所述主控子单元设置为： 启动信道估计和记录当前信道估计处理次数， 到达 T次后结束信道估计；
所述存储子单元设置为： 存储各小区的信道估计值， 初始时各小区的信 道估计值为 0; 以及 所述信道估计子单元设置为： 根据主控单元的指示估计各小区的信道估 计值， 所述信道估计子单元包括： 重构模块， 其设置为： 使用各小区的基本训练序列和存储子单元存储的 各小区的信道估计值， 重构除当前处理小区外的其他各小区的信号； 干扰抵消模块， 其设置为在接收的训练序列数据中去除重构的其他各小 区的信号， 得到当前处理小区的干净信号； 信道估计模块， 其设置为使用当前处理小区的干净信号进行信道估计， 得到当前处理小区的信道估计值； 以及 输出模块， 其设置为用于将信道估计模块估计的所有小区的信道估计值 输出至存储子单元。
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200910143732.5A CN101902249B (zh) | 2009-05-25 | 2009-05-25 | 一种频偏估计装置和方法 |
| CN200910143732.5 | 2009-05-25 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2010135935A1 true WO2010135935A1 (zh) | 2010-12-02 |
Family
ID=43222144
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/CN2010/071769 WO2010135935A1 (zh) | 2009-05-25 | 2010-04-14 | 一种频偏估计装置和方法 |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN101902249B (zh) |
| WO (1) | WO2010135935A1 (zh) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105791195A (zh) * | 2014-12-25 | 2016-07-20 | 中兴通讯股份有限公司 | 微波通信系统信道频偏估计方法及装置 |
| CN106487735A (zh) * | 2015-09-01 | 2017-03-08 | 中兴通讯股份有限公司 | 一种频偏估计方法及装置 |
| CN105897643B (zh) * | 2016-04-08 | 2019-04-16 | 中国人民解放军国防科学技术大学 | Ofdm系统中基于连续循环移位训练序列的联合窄带干扰检测、消除与信道估计方法 |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101026408A (zh) * | 2007-01-19 | 2007-08-29 | 重庆重邮信科股份有限公司 | 时分－同步码分多址系统中频率精细校正的方法及装置 |
| CN101404518A (zh) * | 2008-11-21 | 2009-04-08 | 北京天碁科技有限公司 | 一种用于无线通信系统的频偏估计方法及装置 |
| CN101420248A (zh) * | 2008-12-05 | 2009-04-29 | 北京天碁科技有限公司 | 一种td-scdma终端频偏估计的方法及装置 |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101147371B (zh) * | 2005-01-12 | 2012-11-28 | Nxp股份有限公司 | 信道估计的方法和设备 |
| CN100544333C (zh) * | 2005-06-08 | 2009-09-23 | 重庆重邮信科通信技术有限公司 | 移动通信系统中的频偏估计方法及装置 |
| CN1929353B (zh) * | 2005-09-09 | 2012-04-25 | 展讯通信（上海）有限公司 | 同频小区信道估计方法及装置 |
-
2009
- 2009-05-25 CN CN200910143732.5A patent/CN101902249B/zh active Active
-
2010
- 2010-04-14 WO PCT/CN2010/071769 patent/WO2010135935A1/zh active Application Filing
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101026408A (zh) * | 2007-01-19 | 2007-08-29 | 重庆重邮信科股份有限公司 | 时分－同步码分多址系统中频率精细校正的方法及装置 |
| CN101404518A (zh) * | 2008-11-21 | 2009-04-08 | 北京天碁科技有限公司 | 一种用于无线通信系统的频偏估计方法及装置 |
| CN101420248A (zh) * | 2008-12-05 | 2009-04-29 | 北京天碁科技有限公司 | 一种td-scdma终端频偏估计的方法及装置 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101902249B (zh) | 2014-04-09 |
| CN101902249A (zh) | 2010-12-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Fernandez et al. | Performance of the 802.11 p physical layer in vehicle-to-vehicle environments | |
| TWI352531B (en) | Apparatus,method and computer program product for | |
| US20070211827A1 (en) | Channel Estimation in an Ofdm System With High Doppler Shift | |
| KR20080083259A (ko) | 파일럿을 이용한 채널 추정 방법 및 장치 | |
| JPWO2003032541A1 (ja) | Ｏｆｄｍ受信方法及びｏｆｄｍ受信装置 | |
| RU2006146805A (ru) | Способ обработки сигналов и процессор сигналов в системе ofdm | |
| US20080186948A1 (en) | Method and device for frame synchronization and coarse frequency offset estimation (cfe) | |
| RU2007137486A (ru) | Коррекции хронирования в системе с множественными несущими и распространение на временной фильтр оценки канала | |
| EP2127283A2 (en) | Frequency domain equalization for time varying channels | |
| JP2022064954A (ja) | 送信機、受信機およびその対応方法 | |
| KR100611170B1 (ko) | 수신 장치 및 수신 타이밍 검출 방법 | |
| TW201106638A (en) | Methods and systems using FFT window tracking algorithm | |
| CN109302208B (zh) | 一种交织Gold映射序列的并行组合扩频水声通信方法 | |
| CN102664839B (zh) | 信道估计方法及装置 | |
| RU2414083C2 (ru) | Оценка канала связи | |
| WO2010135935A1 (zh) | 一种频偏估计装置和方法 | |
| CN103581065B (zh) | 一种维纳滤波信道估计方法和装置 | |
| JP2008113240A (ja) | マルチキャリア受信装置およびチャネル推定値補間方法 | |
| WO2006106474A2 (en) | Method and apparatus for estimating channel in mobile communication system | |
| CN102710564B (zh) | 一种信道时域冲激响应滤波方法及装置 | |
| CN102369707B (zh) | 消除导频上的同频干扰的方法和装置 | |
| JP4640870B2 (ja) | 受信装置 | |
| CN102143098B (zh) | 一种正交频分复用系统中的信道估计方法及装置 | |
| CN107276953B (zh) | 定时同步方法、装置和系统 | |
| JP5001427B2 (ja) | ルックアップテーブルによる共分散の効率的な算出 |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| 121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 10780017 Country of ref document: EP Kind code of ref document: A1 |
|
| NENP | Non-entry into the national phase |
Ref country code: DE |
|
| 122 | Ep: pct application non-entry in european phase |
Ref document number: 10780017 Country of ref document: EP Kind code of ref document: A1 |