WO2016015592A1 - 一种图拨检测方法及其装置 - Google Patents
一种图拨检测方法及其装置 Download PDFInfo
- Publication number
- WO2016015592A1 WO2016015592A1 PCT/CN2015/084932 CN2015084932W WO2016015592A1 WO 2016015592 A1 WO2016015592 A1 WO 2016015592A1 CN 2015084932 W CN2015084932 W CN 2015084932W WO 2016015592 A1 WO2016015592 A1 WO 2016015592A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- information
- reliability
- feedback information
- external information
- priori
- Prior art date
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L1/00—Arrangements for detecting or preventing errors in the information received
Definitions
- the present invention relates to the field of communications technologies, and in particular, to a dial detection method and apparatus therefor.
- 3GPP 3rd Generation Partnership Project
- LTE Long Term Evolution
- UMTS Universal Mobile Telecommunications System
- MIMO Multiple-Input Multiple-Output
- OFDM Orthogonal Frequency Division Multiplex
- the main performance objectives of 3GPP LTE include: providing a downlink rate of 100 Mbps and an uplink of 50 Mbps in a 20 MHz spectrum bandwidth (maximum support for downlink 3 Gbps and uplink peak rate of 150 M); improving base station edge user performance; increasing base station capacity; and reducing system delay.
- the unidirectional transmission delay in the user plane is less than 5ms, the migration time from the sleep state to the active state is less than 50ms, the migration time from the resident state to the active state is less than 100ms, and the base station coverage of 100km radius is supported; Unpaired spectrum and flexible configuration of multiple bandwidths from 1.25MHz to 20MHz.
- the LTE system implements multi-layer data transmission through MIMO technology to achieve higher throughput.
- multi-layer transmission will cause serious inter-layer and inter-antenna interference.
- UE User Equipment
- MMSE Minimum Mean Square Error
- the LTE system will consider the use of enhanced interference cancellation algorithms such as Turbo detection and Maximum Likelihood (ML) detection.
- Turbo detection has better performance than ML in the low to medium signal-to-noise ratio, and can provide a better performance experience for the UE.
- Turbo detection has a large computational complexity and processing delay, and is difficult to use for devices with limited complexity.
- the embodiment of the invention provides a method for detecting dials and a device thereof, which are used for improving the robustness of Turbo detection, reducing the computational complexity and processing delay of Turbo detection.
- a map dial detection method includes:
- the prior information based on the maximum a posteriori criterion Decoding to obtain external information ⁇ ;
- the first feedback information Performing rate matching, coding block concatenation, and scrambling processing to obtain the second feedback information
- a map dial detecting device includes:
- a decoding unit configured to receive a signal vector y and second feedback information from the second processing unit And acquiring an equivalent frequency domain channel estimation matrix ⁇ of each channel that receives the signal vector; according to the second feedback information And the equivalent frequency domain channel estimation matrix ⁇ , demodulating the signal vector y to obtain a log likelihood ratio ⁇ ;
- a first processing unit configured to descramble the log likelihood ratio ⁇ , obtain a prior information by coding block cascading and de-rate matching
- a decoding unit configured to a priori information obtained by the first processing unit based on a maximum a posteriori criterion Decoding to obtain external information ⁇ ;
- a reliability determining unit configured to determine the a priori information And whether the reliability of the external information ⁇ satisfies the requirements
- a feedback information processing unit configured to determine a prior information if the reliability determining unit determines And the reliability of the external information ⁇ does not meet the requirement, then the first feedback information Set to 0;
- the second processing unit is configured to use the first feedback information obtained by the feedback information processing unit. Performing rate matching, coding block concatenation, and scrambling processing to obtain the second feedback information
- the embodiment of the present invention has the following advantages: in the process of detecting the dialing, filtering out a priori information whose reliability does not meet the requirements And the external information ⁇ , the first feedback information to be used for demodulation Set to 0 to indicate that no a priori information is passed to the demodulation module.
- the influence of the inaccurate part on the demodulation module increases the reliability of each loop information, and the number of iterations of the Turbo detection can be reduced in the case of ensuring the Turbo detection effect. Therefore, the above scheme improves the robustness of the Turbo detection, thereby It can reduce the computational complexity and processing delay of Turbo detection.
- FIG. 1 is a schematic structural diagram of a turbo detecting device according to an embodiment of the present invention.
- FIG. 2 is a schematic flowchart of a method according to an embodiment of the present invention.
- FIG. 3 is a schematic structural diagram of a turbo detecting device according to an embodiment of the present invention.
- FIG. 4 is a schematic flowchart of a method according to an embodiment of the present invention.
- FIG. 5 is a schematic structural diagram of a turbo detecting device according to an embodiment of the present invention.
- FIG. 6 is a schematic structural diagram of a turbo detecting device according to an embodiment of the present invention.
- FIG. 7 is a schematic structural diagram of a turbo detecting device according to an embodiment of the present invention.
- the channel model can be expressed as:
- Turbo detection device is shown in Figure 1.
- the demodulation module and the decoding module are connected together in an iterative manner, that is, the demodulation operation is first performed, then the decoding operation is performed (decoding internal iteration D times), and then the second operation is performed.
- the demodulation operation, and the second decoding operation (the same iteration is internally performed D times), continues until the Lth demodulation and decoding is completed.
- L be the number of iterations of Turbo detection.
- the demodulation module receives external H and y, H from channel estimation; the demodulation module includes MIMO detection and soft demodulation (constellation point de-mapping), and the demodulation module calculates a bit soft value, that is, a log likelihood ratio ( Log-Likelihood Ratio, LLR).
- a bit soft value that is, a log likelihood ratio ( Log-Likelihood Ratio, LLR).
- the first item on the right side of the above formula (2) The posterior information obtained for the demodulation module, the second item For the LLR returned by the decoding module (external information ⁇ j, i is rate matched, coded block (CB) cascaded and scrambled
- the external information ⁇ j,i is obtained by the following formula (3).
- the second item A value of 0 indicates that there is no prior information, because no decoding has been done at this time, and there is no prior information.
- the decoding module uses the maximum a posteriori criterion, and the output LLR can be expressed as (corresponding to the i-th bit b j,i of the j-th layer symbol)
- the a posteriori information obtained for the decoding module can be obtained by different algorithms, such as BCJR (Bahl, Cocke, Jelinek, Raviv, an algorithm on the trellis diagram used to maximize the posterior probability of error correction coding)
- BCJR Bohl, Cocke, Jelinek, Raviv, an algorithm on the trellis diagram used to maximize the posterior probability of error correction coding
- the algorithm, the final test result of the Turbo receiver is obtained by this hard decision.
- LLR For the demodulation module output LLR ( ⁇ j, i through descrambling, CB de-cascade and de-rate matching ).
- L and D values usually power consumption and delay
- Embodiments of the present invention will provide a solution capable of reducing the value of L and/or D while ensuring Turbo detection performance, thereby reducing the complexity of Turbo detection.
- An embodiment of the present invention provides a method for detecting dials, as shown in FIG. 2, including:
- the receiving vector is an N-dimensional receiving vector composed of signals of N antennas.
- demodulating the signal vector y to obtain a log likelihood ratio ⁇ comprising: performing maximum likelihood ratio detection on the signal vector y according to the equivalent frequency domain channel estimation matrix ⁇ Using the result of the detection and the second feedback information described above Soft demodulation is performed to obtain the above log likelihood ratio ⁇ .
- the detection result is also obtained in the decoding process of step 204.
- the solution of the embodiment of the present invention does not involve the detection result, and therefore is not described in more detail.
- step 206 The first feedback information mentioned above Perform rate matching, coding block concatenation, and scrambling processing to obtain the above second feedback information. This step 206 obtains the second feedback information. Second feedback It will be passed to step 201 to perform the next iteration.
- the specific number of iterations may be preset, or it may be stopped after determining that the detection result meets the requirements.
- the specific manner of how to control the number of iterations is not limited in the embodiment of the present invention.
- the a priori information whose reliability is not satisfactory is filtered out.
- the above external information ⁇ the first feedback information to be used for demodulation Set to 0 to indicate that no a priori information is passed to the demodulation module.
- the embodiment of the present invention further provides a solution capable of further improving the Turbo detection efficiency, as follows:
- the first feedback information above Perform rate matching, coding block concatenation, and scrambling processing to obtain the above second feedback information.
- the embodiment of the present invention further provides how to determine the a priori information.
- the specific implementation manner of whether the reliability of the external information ⁇ meets the requirements and it should be noted that those skilled in the art may determine the a priori information according to other manners.
- the reliability of the above-mentioned external information ⁇ so the manner of determining the reliability is not limited to the manner given in the embodiment, and an example of a more convenient judgment manner given in this embodiment is not to be understood as an embodiment of the present invention.
- the uniqueness is defined as follows: the above a priori information is determined And whether the reliability of the above external information ⁇ satisfies the requirements, including:
- the a priori information is determined according to a preset rule. And the reliability of the above external information ⁇ satisfies the requirement, or the above a priori information And the reliability of the above external information ⁇ does not meet the requirements;
- the threshold value ⁇ is obtained by querying the above modulation and coding scheme MCS.
- the threshold value T in this embodiment can be determined by off-line simulation.
- the threshold value T is associated with the encoding code rate and the modulation order of the data for the Turbo detection for offline simulation.
- the threshold value ⁇ is stored in the table for the query.
- the values of the above ⁇ (n) and ⁇ (n) are positively correlated with the current number of iterations n.
- the number of times, so the above scheme further improves the robustness of Turbo detection, which can reduce the computational complexity and processing delay of Turbo detection.
- ⁇ (n) and ⁇ (n) can be linear processing factors.
- ⁇ (n) and ⁇ (n) can be linear processing factors.
- the feedback information can be made more efficient by dividing the processing factor with a linear processing factor and the nonlinear processing factor.
- Other weighting calculation methods are also achievable.
- the calculation method used in the embodiments of the present invention is a preferred example and should not be construed as limiting the uniqueness of the embodiments of the present invention.
- the processing factors ⁇ (n) , ⁇ (n) can be preset by a technician, in the embodiment of the present invention, the processing factors ⁇ (n) , ⁇ (n) , with It can be determined by off-line simulation, in the offline simulation process, the processing factors ⁇ (n) , ⁇ (n) , with Offline simulation is performed in association with the code rate and modulation order of the data for which the Turbo detection is directed.
- the factors ⁇ (n) , ⁇ (n) with Store in a table for querying.
- the value of the number of iterations in the above decoding may be preset by a technician.
- the number of internal iterations may be determined by offline simulation. In the offline simulation process, the number of internal iterations and Turbo detection are used. The coded code rate and the modulation order of the data are correlated for off-line simulation. After determining the number of iterations, the number of inner iterations is stored in the table for the query.
- Embodiments of the present invention will propose solutions for reducing L and reducing D, respectively, which are described in detail below.
- the performance of the receiver is guaranteed.
- the difference from FIG. 1 is only that the information input by the demodulation module is obtained by processing the decoded prior information and the external information, which ensures that the L value is smaller when the value is smaller.
- the performance of the receiver is ensured by the feedback information provided by each decoding.
- the information processing module in Figure 3 contains two types of operations, which are described separately below.
- Operation 1 In order to avoid the influence of the inaccurate part of the decoding feedback information on the demodulation module, it is necessary to identify unreliable information.
- the specific criteria are: when satisfied
- the log-Likelihood Ratio (LLR) information is unreliable (that is, the LLR has a small absolute value and is considered unreliable), and the corresponding feedback information is set to zero.
- LLR log-Likelihood Ratio
- Setting the feedback information to 0 means that no a priori information is passed to the demodulation module.
- the above absolute value is usually small, which means that when the judgments of demodulation and decoding are completely opposite, we also consider it unreliable. This special case also shows that the above criteria are reasonable. For LLRs that do not satisfy (4), we do nothing here.
- the above threshold T is related to the code rate and the modulation order of the code, and can be determined in advance by offline simulation.
- MCS Modulation Coding Scheme
- Operation 2 For the LLR that has been processed, the operation 2 is no longer processed further; the other LLRs are processed as follows:
- ⁇ (i) and ⁇ (i) are linear processing factors
- n represents the nth iteration (ie, the coefficients are different for each iteration.
- n the larger the linear factor is, because the lower the number of iterations.
- the purpose of doing so is to control the weight of the feedback information of the a priori information and the external information, and to ensure the quality of the feedback information, because the role of the prior information and the external information are different.
- Yes It is a nonlinear processing factor, and the purpose is to treat different LLR sizes differently, because the LLR information with larger values and the LLR information with smaller values have different contributions to the feedback information. same, Yes It is related to the current iterations.
- the values of the above four processing factors are related to the code rate and the modulation order of the code, and can be determined in advance by offline simulation.
- Turbo detection When performing Turbo detection, first find out the corresponding processing factor according to MCS, and then perform operation 2.
- the operation flow of the information processing module is as shown in FIG. 4 .
- the embodiment of the present invention proposes a scheme of using the number of iterations in the unbalanced manner, that is, the number of internal iterations per decoding is different, and D n is used. Said.
- the number of decodings should follow the principle of less before and after.
- D n is 0, it means that the iteration does not decode, and the intermediate de-scrambling, CB block cascading, rate matching, etc. do not need to be done.
- the LLR required for demodulation can directly feedback from the demodulation module. Return to the demodulation module.
- the value of the above D n is related to the code rate of the code and the modulation order, and can be determined in advance by offline simulation.
- Turbo detection you can first find the corresponding D n according to the MCS, and then perform the inner iteration.
- the solution proposed by the embodiment of the present invention can greatly reduce the complexity of the UE implementing Turbo detection on the premise that the performance of the receiver is substantially unchanged.
- the method of the embodiment of the present invention can at least take the L and the D. The value is reduced by half, and the processing power consumption and delay of UE Turbo detection are greatly reduced.
- the receiver of the embodiment of the present invention can be extended to other communication systems, such as GSM (Group Special Mobile communication), TDS-CDMA (Time Division-Synchronous Code Division Multiple Access), WCDMA (Wideband Code Division Multiple Access), CDMA2000 (Code Division Multiple Access 2000), and the like.
- GSM Global System for Mobile communication
- TDS-CDMA Time Division-Synchronous Code Division Multiple Access
- WCDMA Wideband Code Division Multiple Access
- CDMA2000 Code Division Multiple Access 2000
- the receiver framework of the embodiment of the present invention can not only eliminate inter-layer interference, but also can eliminate various types of interference such as inter-cell interference, inter-user interference, inter-code crosstalk, and inter-subcarrier interference.
- the embodiment of the present invention further provides a map detecting device, as shown in FIG. 5, in the direction indicated by the arrow in FIG. 5, the data flow direction includes:
- a decoding unit 501 configured to receive a signal vector y and second feedback information from the second processing unit 506 And obtaining an equivalent frequency domain channel estimation matrix ⁇ of each channel that receives the signal vector; according to the second feedback information And the above-mentioned equivalent frequency domain channel estimation matrix ⁇ , demodulating the signal vector y to obtain a log likelihood ratio ⁇ ;
- the first processing unit 502 is configured to descramble the log likelihood ratio ⁇ , obtain a prior information by coding block cascading and de-rate matching
- a decoding unit 503 configured to obtain a priori information obtained by the first processing unit 502 based on a maximum a posteriori criterion Decoding to obtain external information ⁇ ;
- the reliability determining unit 504 is configured to determine the a priori information And whether the reliability of the above external information ⁇ satisfies the requirements;
- the feedback information processing unit 505 is configured to determine the a priori information if the reliability determining unit 504 determines And the reliability of the above external information ⁇ does not satisfy the requirement, then the first feedback information Set to 0;
- the second processing unit 506 is configured to use the first feedback information obtained by the feedback information processing unit 505. Perform rate matching, coding block concatenation, and scrambling processing to obtain the above second feedback information.
- the a priori information whose reliability is not satisfactory is filtered out.
- the above external information ⁇ the first feedback information to be used for demodulation Set to 0 to indicate that no a priori information is passed to the demodulation module.
- the embodiment of the present invention further provides a solution that can further improve the Turbo detection efficiency, and is specifically as follows: the feedback information processing unit 505 is further configured to: if the reliability determining unit 504 determines the prior information And the reliability of the above external information ⁇ satisfies the requirements, then passes The above prior information And weighting the external information ⁇ to obtain the first feedback information.
- the reliability determining unit 504 determines the prior information And the reliability of the above external information ⁇ satisfies the requirements, then passes The above prior information And weighting the external information ⁇ to obtain the first feedback information.
- ⁇ (n) , ⁇ (n) with The query is obtained by the modulation coding scheme MCS preset by the above-mentioned dial detection device, where n is the current number of iterations.
- the embodiment of the present invention further provides how to determine the a priori information.
- the specific implementation manner of whether the reliability of the external information ⁇ meets the requirements and it should be noted that those skilled in the art may determine the a priori information according to other manners.
- the reliability of the above-mentioned external information ⁇ so the manner of determining the reliability is not limited to the manner given in the embodiment, and an example of a more convenient judgment manner given in this embodiment is not to be understood as an embodiment of the present invention.
- the uniqueness is defined as follows: As shown in FIG. 6, the reliability determining unit 504 includes:
- An absolute value calculation unit 601 configured to calculate the a priori information The absolute value of the sum of the above external information ⁇ ;
- a reliability determining subunit 602 configured to compare the foregoing absolute value with a preset threshold value ;
- the a priori information is determined according to a preset rule. And the reliability of the above external information ⁇ satisfies the requirement, or the above a priori information And the reliability of the above external information ⁇ does not meet the requirements;
- the threshold value ⁇ is obtained by querying the above modulation and coding scheme MCS.
- the threshold value T in this embodiment can be determined by off-line simulation.
- the threshold value T is associated with the encoding code rate and the modulation order of the data for the Turbo detection for offline simulation.
- the threshold value ⁇ is stored in the table for the query.
- the values of the above ⁇ (n) and ⁇ (n) are positively correlated with the current number of iterations n.
- ⁇ (n) and ⁇ (n) can be linear processing factors.
- ⁇ (n) and ⁇ (n) can be linear processing factors.
- the feedback information can be made more efficient by dividing the processing factor with a linear processing factor and the nonlinear processing factor.
- Other weighting calculation methods are also achievable.
- the calculation method used in the embodiments of the present invention is a preferred example and should not be construed as limiting the uniqueness of the embodiments of the present invention.
- the processing factors ⁇ (n) , ⁇ (n) can be preset by a technician, in the embodiment of the present invention, the processing factors ⁇ (n) , ⁇ (n) , with It can be determined by off-line simulation, in the offline simulation process, the processing factors ⁇ (n) , ⁇ (n) , with Offline simulation is performed in association with the code rate and modulation order of the data for which the Turbo detection is directed.
- the factors ⁇ (n) , ⁇ (n) with Store in a table for querying.
- the embodiment of the present invention further provides a specific implementation scheme of how to obtain a log-release ratio, as follows:
- the decoding unit 501 is configured to perform maximum likelihood on the signal vector y according to the equivalent frequency domain channel estimation matrix. Ratio detection, using the result of the detection and the second feedback information described above Soft demodulation is performed to obtain the above log likelihood ratio ⁇ .
- the above decoding unit 503 pairs the a priori information
- the number of decodings to be decoded is positively correlated with the current number of iterations n.
- the embodiment of the present invention further provides another Turbo detecting device, as shown in FIG. 7, comprising: a receiver 701, a transmitter 702, a processor 703, a memory 704, and a decoder 705;
- the processor 703 is configured to perform an operation in an iterative manner: receiving a signal vector y and second feedback information. And acquiring an equivalent frequency domain channel estimation matrix ⁇ of each channel that receives the signal vector; according to the second feedback information And the above-mentioned equivalent frequency domain channel estimation matrix ⁇ , demodulating the signal vector y to obtain a log likelihood ratio ⁇ ; descrambling the log likelihood ratio ⁇ , coding block cascading and de-rate matching to obtain Information Prior information Sending to the decoder 705; receiving the external information ⁇ returned by the decoder 705; determining the a priori information And whether the reliability of the above external information ⁇ satisfies the requirement, if the above a priori information And the reliability of the above external information ⁇ does not satisfy the requirement, then the first feedback information Set to 0; the first feedback information above Perform rate matching, coding block concatenation, and scrambling processing to obtain the above second feedback information.
- a decoder 705 configured to receive a priori information from the processor 703 Based on the maximum a posteriori criterion, the above a priori information Decoding is performed to obtain the outer information ⁇ ; and sent to the processor 703.
- the detection result is also obtained in the decoding process of the decoder 705.
- the solution of the embodiment of the present invention does not involve the detection result, and thus is not described in more detail.
- the a priori information whose reliability is not satisfactory is filtered out.
- the above external information ⁇ the first feedback information to be used for demodulation Set to 0 to indicate that no a priori information is passed to the demodulation module.
- the processor 703 is configured to use the second feedback information according to the second feedback information.
- the above-mentioned equivalent frequency domain channel estimation matrix ⁇ demodulating the signal vector y to obtain a log likelihood ratio ⁇ , comprising: performing maximum likelihood ratio detection on the signal vector y according to the equivalent frequency domain channel estimation matrix ⁇ Using the result of the detection and the second feedback information described above Soft demodulation is performed to obtain the above log likelihood ratio ⁇ .
- the above embodiment can filter out a priori information whose reliability does not meet the requirements.
- the embodiment of the present invention further provides a solution capable of further improving the Turbo detection efficiency, as follows: in the process of performing the operation in an iterative manner by the processor 703, if the foregoing a priori information And the reliability of the above external information ⁇ satisfies the requirements, then passes The above prior information And weighting the external information ⁇ to obtain the first feedback information.
- ⁇ (n) , ⁇ (n) with Obtaining by the query for the modulation coding scheme MCS preset by the above-mentioned Turbo detection device, where n is the current number of iterations;
- the first feedback information above Perform rate matching, coding block concatenation, and scrambling processing to obtain the above second feedback information.
- the embodiment of the present invention further provides how to determine the a priori information. And the specific implementation manner of whether the reliability of the external information ⁇ meets the requirements, and it should be noted that those skilled in the art may determine the a priori information according to other manners. And the reliability of the above-mentioned external information ⁇ , so the manner of determining the reliability is not limited to the manner given in the embodiment, and an example of a more convenient judgment manner given in this embodiment is not to be understood as an embodiment of the present invention.
- the uniqueness is defined as follows: the processor 703 is configured to determine the a priori information. And whether the reliability of the above external information ⁇ satisfies the requirements, including: calculating the a priori information described above And an absolute value of the sum of the above external information ⁇ , and comparing the above absolute value with a preset threshold value ;;
- the a priori information is determined according to a preset rule. And the reliability of the above external information ⁇ satisfies the requirement, or the above a priori information And the reliability of the above external information ⁇ does not meet the requirements;
- the threshold value ⁇ is obtained by querying the above modulation and coding scheme MCS.
- the threshold value T in this embodiment can be determined by off-line simulation.
- the threshold value T is associated with the encoding code rate and the modulation order of the data for the Turbo detection for offline simulation.
- the threshold value ⁇ is stored in the table for the query.
- the values of the above ⁇ (n) and ⁇ (n) are positively correlated with the current number of iterations n.
- ⁇ (n) and ⁇ (n) can be linear processing factors.
- ⁇ (n) and ⁇ (n) can be linear processing factors.
- the feedback information can be made more efficient by dividing the processing factor with a linear processing factor and the nonlinear processing factor.
- Other weighting calculation methods are also achievable.
- the calculation method used in the embodiments of the present invention is a preferred example and should not be construed as limiting the uniqueness of the embodiments of the present invention.
- the processing factors ⁇ (n) , ⁇ (n) can be preset by a technician, in the embodiment of the present invention, the processing factors ⁇ (n) , ⁇ (n) , with It can be determined by off-line simulation, in the offline simulation process, the processing factors ⁇ (n) , ⁇ (n) , with Offline simulation is performed in association with the code rate and modulation order of the data for which the Turbo detection is directed.
- the factors ⁇ (n) , ⁇ (n) with Store in a table for querying.
- the value of the number of iterations in the above decoding may be preset by a technician.
- the number of internal iterations may be determined by offline simulation. In the offline simulation process, the number of internal iterations and Turbo detection are used. The coded code rate and the modulation order of the data are correlated for off-line simulation. After determining the number of iterations, the number of inner iterations is stored in the table for the query.
- the included units are only divided according to the functional logic, but are not limited to the above-mentioned division, as long as the corresponding functions can be realized;
- the specific names of the units are also for convenience of distinguishing from each other and are not intended to limit the scope of the present invention.
- the storage medium may be a read only memory, a magnetic disk or an optical disk or the like.
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Error Detection And Correction (AREA)
Abstract
本发明实施例公开了一种图拨检测方法及其装置，其中方法的实现包括：以迭代的方式执行如下操作：接收信号向量y和第二反馈信息并获取接收信号向量的各个信道的等效频域信道估计矩阵Η；根据第二反馈信息和等效频域信道估计矩阵Η，对信号向量y进行解调得到对数似然比λ；对对数似然比λ进行解扰，编码块级联及解速率匹配后得到先验信息基于最大后验准则，对先验信息进行译码，得到外信息γ；确定先验信息和外信息γ的可靠性是否满足要求，若先验信息和外信息γ的可靠性不满足要求，则将第一反馈信息置0；对第一反馈信息进行速率匹配、编码块级联以及加扰处理，以得到第二反馈信息可以降低Turbo检测的运算复杂度和处理时延。
Description
本申请要求于2014年7月28日提交中国专利局、申请号为201410364464.0、发明名称为“一种图拨检测方法及其装置”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
本发明涉及通信技术领域，特别涉及一种图拨检测方法及其装置。
随着无线宽带通信技术的发展，用户对通信系统的性能的要求不断提高。2004年底，第3代合作伙伴计划(3rd Generation Partnership Project,3GPP)开始了通用移动通信系统(Universal Mobile Telecommunications System,UMTS)技术的长期演进(Long Term Evolution,LTE)项目。多输入多输出(Multiple-Input Multiple-Output，MIMO)和正交频分复用(Orthogonal Frequency Division Multiplex,OFDM)是LTE最关键的两个技术。3GPP LTE主要性能目标包括：在20MHz频谱带宽能够提供下行100Mbps、上行50Mbps的峰值速率(最大可支持下行3Gbps，上行150M的峰值速率)；改善基站边缘用户的性能；提高基站容量；降低系统延迟，用户平面内部单向传输时延低于5ms，控制平面从睡眠状态到激活状态迁移时间低于50ms，从驻留状态到激活状态的迁移时间小于100ms；支持100km半径的基站覆盖；支持成对或非成对频谱，并可灵活配置1.25MHz到20MHz多种带宽。
LTE系统通过MIMO技术实现多层数据的传输，以此获得更高的吞吐量。然而，多层传输会带来严重的层间和天线间干扰，为了克服此问题，用户设备(User Equipment,UE)一般会采用某种线性干扰消除算法，如最小均方误差(Minimum Mean Square Error,MMSE)算法。由于线性干扰消除算法抑制干扰的能力有限，LTE系统将考虑使用增强的干扰消除算法，如Turbo(图拨)检测和最大似然(Maximum Likelihood，ML)检测等算法。实际上，Turbo检测相对ML在中低信噪比具有更好的性能，能给UE提供更好的性能体验。然而，Turbo检测为了获得较好的性能，其运算复杂度和处理时延较大，难以用于复杂度受限的设备。
发明内容
本发明实施例提供了一种图拨检测方法及其装置，用于提升Turbo检测的鲁棒性，降低Turbo检测的运算复杂度以及处理时延。
一种图拨检测方法，包括：
以迭代的方式执行如下操作：
接收信号向量y和第二反馈信息
并获取接收所述信号向量的各个信道的等效频域信道估计矩阵Η；
根据所述第二反馈信息
和所述等效频域信道估计矩阵Η，对所述信号向量y进行解调得到对数似然比λ；
对所述对数似然比λ进行解扰，编码块级联及解速率匹配后得到先验信息
基于最大后验准则，对所述先验信息
进行译码，得到外信息γ；
确定所述先验信息
和所述外信息γ的可靠性是否满足要求，若所述先验信息
和所述外信息γ的可靠性不满足要求，则将第一反馈信息
置0；
对所述第一反馈信息
进行速率匹配、编码块级联以及加扰处理，以得到所述第二反馈信息
一种图拨检测装置，包括：
解码单元，用于接收信号向量y和来自第二处理单元的第二反馈信息
并获取接收所述信号向量的各个信道的等效频域信道估计矩阵Η；根据所述第二反馈信息
和所述等效频域信道估计矩阵Η，对所述信号向量y进行解调得到对数似然比λ；
第一处理单元，用于对所述对数似然比λ进行解扰，编码块级联及解速率匹配后得到先验信息
译码单元，用于基于最大后验准则，对所述第一处理单元得到的先验信息
进行译码，得到外信息γ；
可靠性确定单元，用于确定所述先验信息
和所述外信息γ的可靠性是否满足要求；
反馈信息处理单元，用于若所述可靠性确定单元确定先验信息
和所述外信息γ的可靠性不满足要求，则将第一反馈信息
置0；
上述第二处理单元，用于对所述反馈信息处理单元得到的第一反馈信息
进行速率匹配、编码块级联以及加扰处理，以得到所述第二反馈信息
从以上技术方案可以看出，本发明实施例具有以下优点：在图拨检测过程中，滤掉可靠性不满足要求的先验信息
和所述外信息γ，将用于解调的第一反馈信息
置0，表
示不把任何先验信息传递给解调模块。从而避免第一反馈信息
中不准确的部分对解调模块的影响，增加每次循环信息的可靠性，在保证Turbo检测效果的情况下可以减少Turbo检测的迭代次数，因此以上方案提升了Turbo检测的鲁棒性，从而可以降低Turbo检测的运算复杂度以及处理时延。
为了更清楚地说明本发明实施例中的技术方案，下面将对实施例描述中所需要使用的附图作简要介绍，显而易见地，下面描述中的附图仅仅是本发明的一些实施例，对于本领域的普通技术人员来讲，在不付出创造性劳动性的前提下，还可以根据这些附图获得其他的附图。
图1为本发明实施例Turbo检测装置结构示意图；
图2为本发明实施例方法流程示意图；
图3为本发明实施例Turbo检测装置结构示意图；
图4为本发明实施例方法流程示意图；
图5为本发明实施例Turbo检测装置结构示意图；
图6为本发明实施例Turbo检测装置结构示意图；
图7为本发明实施例Turbo检测装置结构示意图。
为了使本发明的目的、技术方案和优点更加清楚，下面将结合附图对本发明作进一步地详细描述，显然，所描述的实施例仅仅是本发明一部份实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其它实施例，都属于本发明保护的范围。
对于Turbo检测复杂度高的原因，发明人进行了详细的分析如下：
在LTE MIMO-OFDM系统中，假设UE侧接收天线个数为N，用x、y和H分别表示某个资源单元(Resource Element,RE)的发射符号向量、接收向量(在本发明实施例中接收向量是N个天线的信号构成的N维接收向量)和等效频域信道估计矩阵，那么信道模型可以表示为：
y＝Hx+n (1)
其中假设n是N维加性高斯白噪声向量(均值为0协方差矩阵为σ2I)；x为M维的
列向量(M表示发射符号总的层数)，即x＝[x1,x2,…,xM]T。
Turbo检测装置如图1所示，解调模块和译码模块通过迭代的方式连接在一起，即首先进行解调操作，然后进行译码操作(译码内部迭代D次)，之后再进行第二次解调操作，以及第二次译码操作(译码内部同样迭代D次)，如此下去，直到做完第L次解调和译码。把L叫做Turbo检测的迭代次数。
其中，解调模块接收外部的H以及y，H来自信道估计；解调模块包括MIMO检测和软解调(星座点解映射)，解调模块计算得到比特软值，即对数似然比(Log-Likelihood Ratio,LLR)。用bj,i表示待估计符号xj对应的第i个比特，则其对应的LLR可以写为：
上式(2)中右边第一项
为解调模块得到的后验信息，第二项
为译码模块返回的LLR(外信息γj,i经过速率匹配、编码块(Coded Block，CB)级联和加扰得到
其中外信息γj,i由下面的(3)式得到)。初次解调时，第二项
为0，表示没有任何先验信息，因为此时还没有做过译码，无先验信息。
译码模块采用最大后验准则，其输出的LLR可以表示为(对应第j层符号第i个比特bj,i)
其中，第一项
为译码模块得到的后验信息，可以由不同的算法得到，如BCJR(Bahl，Cocke，Jelinek，Raviv，一种在网格图上的用来最大化纠错编码的后验概率的算法)算法，Turbo接收机最终的检测结果通过此项硬判决获得。第二项
为解调模块输出LLR(λj,i经过解扰、CB解级联和解速率匹配得到
)。
为了保证Turbo接收机的性能，通常Turbo检测的迭代次数需要比较大，比如L＝5；另一方面，译码的内迭代次数通常也需要比较大，例如D＝8。对于复杂度受限的UE
而言，上述L和D取值较大所带来的复杂度(通常为功耗和时延)是无法接受的。本发明实施例将提供能够在保证Turbo检测性能的情况下，降低L和/或D的取值的方案，从而降低Turbo检测的复杂度。
本发明实施例提供了一种图拨检测方法，如图2所示，包括：
以迭代的方式执行图2所示各步骤的操作：
201：接收信号向量y和第二反馈信息
并获取接收上述信号向量的各个信道的等效频域信道估计矩阵Η；
假设Turbo检测一侧的设备的接收天线个数为N，那么接收向量是N个天线的信号构成的N维接收向量。
202：根据上述第二反馈信息
和上述等效频域信道估计矩阵Η，对上述信号向量y进行解调得到对数似然比λ；
更具体地，本步骤中根据上述第二反馈信息
和上述等效频域信道估计矩阵Η，对上述信号向量y进行解调得到对数似然比λ，包括：根据上述等效频域信道估计矩阵Η对上述信号向量y进行最大似然比检测，利用检测得到的结果和上述第二反馈信息
进行软解调，得到上述对数似然比λ。
203：对上述对数似然比λ进行解扰，编码块级联及解速率匹配后得到先验信息
204：基于最大后验准则，对上述先验信息
进行译码，得到外信息γ；
在步骤204的译码过程中还会得到检测结果，本发明实施例的方案不涉及该检测结果，故未作更详细的说明。
205：确定上述先验信息
和上述外信息γ的可靠性是否满足要求，若上述先验信息
和上述外信息γ的可靠性不满足要求，则将第一反馈信息
置0；
206：对上述第一反馈信息
进行速率匹配、编码块级联以及加扰处理，以得到上述第二反馈信息
本步骤206得到第二反馈信息
后第二反馈信息
会被传递到步骤201执行下一次的迭代。具体的迭代次数可以是预先设定好的，也可以是在确定检测结果符合要求以后停止迭代。如何控制迭代次数的具体方式本发明实施例不予限定。
以上实施例，在Turbo检测过程中，滤掉可靠性不满足要求的先验信息
和上述外信息γ，将用于解调的第一反馈信息
置0，表示不把任何先验信息传递给解调模块。从而避免第一反馈信息
中不准确的部分对解调模块的影响，增加每次循环信息的可靠性，在保证Turbo检测效果的情况下可以减少Turbo检测的迭代次数，因此以上方案提升了Turbo检测的鲁棒性，从而可以降低Turbo检测的运算复杂度以及处理时延。
以上实施例可以滤掉可靠性不满足要求的先验信息
和上述外信息γ，本发明实施
例还提供了能够进一步提升Turbo检测效率的方案，具体如下：
若上述先验信息
和上述外信息γ的可靠性满足要求，则通过
分别对上述先验信息
和上述外信息γ进行加权计算，得到上述第一反馈信息
其中，α(n)、β(n)、
和
通过查询预设的调制编码方案MCS获取，n为当前的迭代次数；
然后对上述第一反馈信息
进行速率匹配、编码块级联以及加扰处理，以得到上述第二反馈信息
可选地，本发明实施例还提供了如何确定上述先验信息
和上述外信息γ的可靠性是否满足要求的具体实现方式，需要说明的是，本领域技术人员可以依据其他方式确定上述先验信息
和上述外信息γ的可靠性，因此确定可靠性的方式也并不仅限于本实施例给出的方式，本实施例给出的一个较为方便的判断方式的举例不应理解为对本发明实施例的唯一性限定，具体如下：上述确定上述先验信息
和上述外信息γ的可靠性是否满足要求，包括：
计算上述先验信息
和上述外信息γ之和的绝对值，并将上述绝对值与预设的门限值Τ进行比较；
若上述绝对值小于上述门限值Τ，则上述先验信息
和上述外信息γ的可靠性满足要求；
若上述绝对值大于上述门限值Τ，则上述先验信息
和上述外信息γ的可靠性不满足要求；
若上述绝对值等于上述门限值Τ，则根据预设的规则，确定上述先验信息
和上述外信息γ的可靠性满足要求，或者，上述先验信息和上述外信息γ的可靠性不满足要求；
其中，上述门限值Τ通过查询上述调制编码方案MCS获取。
本实施例中的门限值T可以通过离线仿真的方式确定，在离线仿真过程中，将门限值T与Turbo检测所针对的数据的编码码率及调制阶数关联进行离线仿真。在确定T值以后，将门限值Τ存放到表格中用于查询。
优选地，在本发明实施例中，上述α(n)和β(n)的取值与当前的迭代次数n正相关。
在本发明实施例中，迭代次数越靠后，信息会越可靠因此使用较大的处理因子来增加可靠性高的信息的使用，在保证Turbo检测效果的情况下可以减少Turbo检测的迭代
次数，因此以上方案进一步提升了Turbo检测的鲁棒性，从而可以降低Turbo检测的运算复杂度以及处理时延。
以上实施例使用了两组处理因子，其中α(n)和β(n)可以为线性处理因子，
和
可以为非线性处理因子。通过处理因子为线性的处理因子和非线性的处理因子的划分，可以使反馈信息更加有效。其他的加权计算方式也是可以实现的，本发明实施例所采用的计算方式作为一个优选举例，不应理解为对本发明实施例的唯一性限定。
对于以上处理因子α(n)、β(n)、
和
可以由技术人员预先设定，在本发明实施例中，处理因子α(n)、β(n)、
和
可以通过离线仿真的方式确定，在离线仿真过程中，将处理因子α(n)、β(n)、
和
与Turbo检测所针对的数据的编码码率及调制阶数关联进行离线仿真。在确定处理因子α(n)、β(n)、
和
以后，将处理因子α(n)、β(n)、
和
存放到表格中用于查询。
为了在保证Turbo接收机性能的情况下，减少译码内迭代次数(即每次迭代过程中，对上述先验信息
进行译码的译码次数)D的取值，本发明实施例还提出采用非平衡内迭代次数的方案，即每次译码的内迭代次数不同，用Dn表示。通常情况下，译码次数应该遵循前少后多的原则。当Dn取值为0，表示本次迭代不做译码，中间的加解扰、CB级联、速率匹配等操作也不需要做，解调需要的LLR可以直接从解调模块直接反馈回到解调模块。具体如下：上述对上述先验信息
进行译码的译码次数与当前的迭代次数n正相关。
对于以上译码内迭代次数的取值，可以由技术人员预先设定，在本发明实施例中，内迭代次数可以通过离线仿真的方式确定，在离线仿真过程中，将内迭代次数与Turbo检测所针对的数据的编码码率及调制阶数关联进行离线仿真。在确定内迭代次数以后，将内迭代次数存放到表格中用于查询。
以下实施例将就本发明实施例的具体实现方案进行举例说明，分别就如何减少L和减少D进行分别说明。
从本申请文件具体实施例图2之前的论述中可以看出，Turbo检测技术为了保证其性能，检测的迭代次数要很大，比如L＝5；另外，译码的内迭代次数也很多，比如D＝9。这样的复杂度在UE侧实现是很困难的。本发明实施例将解决L和D的取值过大，复杂度过高的问题。本发明实施例的方案在保证Turbo接收机的性能的前提下，可以大大降低L和D的取值。
本发明实施例将分别针对减少L和减少D提出解决方案，下面分别详述。
一、
为了在Turbo检测的迭代次数L取值较小(例如L＝2或L＝3)的情况下，保证接收机的性能。如图3所示，和图1的差别仅在于：解调模块输入的信息是译码先验信息和外信息经过处理得到，将这保证了在L取值较小的情况下，更加充分的利用每次译码提供的反馈信息，保证接收机的性能。
图3中的信息处理模块包含两类操作，下面分别描述。
操作一：为了避免译码反馈信息中不准确的部分对解调模块的影响，需要识别出不可靠的信息。具体准则为：当满足
时，我们认为对数似然比(Log-Likelihood Ratio,LLR)信息不可靠(即LLR绝对值较小的认为不可靠)，把相应的反馈信息置为0。换句话说，当译码的先验信息与外信息和的绝对值比较小，我们认为LLR其不可靠。把反馈信息置为0，表示不把任何先验信息传递给解调模块。特别的，当译码先验信息和外信息符号相反的时候，上述绝对值通常较小，这说明解调和译码的判断是完全相反时，我们也会认为不可靠。这个特例也表明上述准则是比较合理的。对于不满足(4)式的LLR，我们在这里不做任何处理。
上述门限T和编码的码率及调制阶数有关，可以预先通过离线仿真确定。执行Turbo检测的时候，可以先根据当前的调制编码方案(Modulation Coding Scheme,MCS)查出相应的T，然后再执行操作一。
操作二：对于操作一已经处理的LLR，操作二不再做进一步处理；其他LLR做如下处理：
其中，α(i)和β(i)是线性处理因子，n表示第n次迭代(即每次迭代系数不同，通常情况n越大，线性因子越大，这是因为迭代次数越靠后，信息越可靠)，这样做的目的是控制先验信息和外信息在反馈信息的权重，保证反馈信息的质量，因为先验信息和外信息的作用是不同的。
是
是非线性处理因子，目的是对不同的LLR大小做不同的处理，因为取值较大的LLR信息和取值较小的LLR信息对反馈信息的贡献是不同的。同样的，
是
和当前是第几次迭代有关。
上述四个处理因子的取值和编码的码率及调制阶数有关，可以预先通过离线仿真确定。执行Turbo检测的时候，先根据MCS查出相应的处理因子，然后再执行操作二。
综上，对于减少L的方案，信息处理模块的操作流程如图4所示。
401：获取第一反馈信息
(译码的先验信息)和外信息；
402：根据MCS查表获得(4)式中的门限T和(5)式中的4个处理因子；
403：根据(4)式把不可靠的信息置0；
404：根据(5)式处理未置0的信息。
二、
为了在保证Turbo接收机性能的情况下，减少译码内迭代次数D的取值，本发明实施例提出采用非平衡内迭代次数的方案，即每次译码的内迭代次数不同，用Dn表示。通常情况下，译码次数应该遵循前少后多的原则。当Dn取值为0，表示本次迭代不做译码，中间的加解扰、CB块级联、速率匹配等操作也不需要做，解调需要的LLR可以直接从解调模块直接反馈回到解调模块。
上述Dn的取值和编码的码率及调制阶数有关，可以预先通过离线仿真确定。执行Turbo检测的时候，可以先根据MCS查出相应的Dn，然后再执行内迭代。
例如，MCS＝25时，L＝2下Dn取值为{3，6}，即第一次译码3次内迭代，第二次译码6次内迭代；L＝3下Di取值为{0，3，6}，即第一次不做译码，后两次内迭代次数为3和6。
从上面的实施例可以看出，在保持接收机性能基本不变的前提下，本发明实施例提出的方案可以大大降低UE实现Turbo检测的复杂度。具体而言，相对于本发明实施例的举例而言Turbo检测的迭代次数通常为L＝5，而译码内迭代次数通常为D＝9，本发明实施例的方法至少可以把L和D的取值减少一半，UE Turbo检测的处理功耗以及时延大大减少。
本发明实施例的接收机可以推广到其他通信系统，如GSM(Group Special Mobile communication，全球移动通信系统)、TDS-CDMA(Time Division-Synchronous Code Division Multiple Access，时分-同步码分多址)、WCDMA(Wideband Code Division Multiple Access，宽带码分多址)、CDMA2000(Code Division Multiple Access 2000，码分多址2000)等。另外，本发明实施例的接收机框架不单单可以可消除层间干扰，还可以用来消除小区间干扰、用户间干扰、码间串扰、子载波间干扰等各种类型的干扰。
本发明实施例还提供了一种图拨检测装置，如图5所示，在图5中箭头所示方向为数据流向，包括：
解码单元501，用于接收信号向量y和来自第二处理单元506的第二反馈信息
并获取接收上述信号向量的各个信道的等效频域信道估计矩阵Η；根据上述第二反馈信
息
和上述等效频域信道估计矩阵Η，对上述信号向量y进行解调得到对数似然比λ；
第一处理单元502，用于对上述对数似然比λ进行解扰，编码块级联及解速率匹配后得到先验信息
译码单元503，用于基于最大后验准则，对上述第一处理单元502得到的先验信息
进行译码，得到外信息γ；
可靠性确定单元504，用于确定上述先验信息
和上述外信息γ的可靠性是否满足要求；
反馈信息处理单元505，用于若上述可靠性确定单元504确定先验信息
和上述外信息γ的可靠性不满足要求，则将第一反馈信息
置0；
上述第二处理单元506，用于对上述反馈信息处理单元505得到的第一反馈信息
进行速率匹配、编码块级联以及加扰处理，以得到上述第二反馈信息
以上实施例，在Turbo检测过程中，滤掉可靠性不满足要求的先验信息
和上述外信息γ，将用于解调的第一反馈信息
置0，表示不把任何先验信息传递给解调模块。从而避免第一反馈信息
中不准确的部分对解调模块的影响，增加每次循环信息的可靠性，在保证Turbo检测效果的情况下可以减少Turbo检测的迭代次数，因此以上方案提升了Turbo检测的鲁棒性，从而可以降低Turbo检测的运算复杂度以及处理时延。
以上实施例可以滤掉可靠性不满足要求的先验信息
和上述外信息γ，本发明实施例还提供了能够进一步提升Turbo检测效率的方案，具体如下：上述反馈信息处理单元505，还用于若上述可靠性确定单元504确定先验信息
和上述外信息γ的可靠性满足要求，则通过
分别对上述先验信息
和上述外信息γ进行加权计算，得到上述第一反馈信息
其中，α(n)、β(n)、
和
通过查询为上述图拨检测装置预设的调制编码方案MCS获取，n为当前的迭代次数。
可选地，本发明实施例还提供了如何确定上述先验信息
和上述外信息γ的可靠性是否满足要求的具体实现方式，需要说明的是，本领域技术人员可以依据其他方式确定上述先验信息
和上述外信息γ的可靠性，因此确定可靠性的方式也并不仅限于本实施例给出的方式，本实施例给出的一个较为方便的判断方式的举例不应理解为对本发明实施例的唯一性限定，具体如下：如图6所示，上述可靠性确定单元504包括：
绝对值计算单元601，用于计算上述先验信息
和上述外信息γ之和的绝对值；
可靠性确定子单元602，用于将上述绝对值与预设的门限值Τ进行比较；
若上述绝对值小于上述门限值Τ，则上述先验信息
和上述外信息γ的可靠性满足
要求；
若上述绝对值大于上述门限值Τ，则上述先验信息
和上述外信息γ的可靠性不满足要求；
若上述绝对值等于上述门限值Τ，则根据预设的规则，确定上述先验信息
和上述外信息γ的可靠性满足要求，或者，上述先验信息
和上述外信息γ的可靠性不满足要求；
其中，上述门限值Τ通过查询上述调制编码方案MCS获取。
本实施例中的门限值T可以通过离线仿真的方式确定，在离线仿真过程中，将门限值T与Turbo检测所针对的数据的编码码率及调制阶数关联进行离线仿真。在确定T值以后，将门限值Τ存放到表格中用于查询。
优选地，上述α(n)、β(n)的取值与当前的迭代次数n正相关。
在本发明实施例中，迭代次数越靠后，信息会越可靠因此使用较大的处理因子来增加可靠性高的信息的使用，在保证Turbo检测效果的情况下可以减少Turbo检测的迭代次数，因此以上方案进一步提升了Turbo检测的鲁棒性，从而可以降低Turbo检测的运算复杂度以及处理时延。
以上实施例使用了两组处理因子，其中α(n)和β(n)可以为线性处理因子，
和
可以为非线性处理因子。通过处理因子为线性的处理因子和非线性的处理因子的划分，可以使反馈信息更加有效。其他的加权计算方式也是可以实现的，本发明实施例所采用的计算方式作为一个优选举例，不应理解为对本发明实施例的唯一性限定。
对于以上处理因子α(n)、β(n)、
和
可以由技术人员预先设定，在本发明实施例中，处理因子α(n)、β(n)、
和
可以通过离线仿真的方式确定，在离线仿真过程中，将处理因子α(n)、β(n)、
和
与Turbo检测所针对的数据的编码码率及调制阶数关联进行离线仿真。在确定处理因子α(n)、β(n)、
和
以后，将处理因子α(n)、β(n)、
和
存放到表格中用于查询。
更具体地，本发明实施例还提供了如何获得对数释然比的具体实现方案，如下：上述解码单元501，用于根据上述等效频域信道估计矩阵Η对上述信号向量y进行最大似然比检测，利用检测得到的结果和上述第二反馈信息
进行软解调，得到上述对数似然比λ。
为了在保证Turbo接收机性能的情况下，减少译码内迭代次数(即每次迭代过程中，对上述先验信息
进行译码的译码次数)D的取值，本发明实施例还提出采用非平衡内
迭代次数的方案，即每次译码的内迭代次数不同，用Dn表示。通常情况下，译码次数应该遵循前少后多的原则。当Dn取值为0，表示本次迭代不做译码，中间的加解扰、CB级联、速率匹配等操作也不需要做，解调需要的LLR可以直接从解调模块直接反馈回到解调模块。具体如下：
上述译码单元503对上述先验信息
进行译码的译码次数与当前的迭代次数n正相关。
本发明实施例还提供了另一种Turbo检测装置，如图7所示，包括：接收器701、发射器702、处理器703、存储器704、以及译码器705；
其中，处理器703用于以迭代的方式执行如下操作：接收信号向量y和第二反馈信息
并获取接收上述信号向量的各个信道的等效频域信道估计矩阵Η；根据上述第二反馈信息
和上述等效频域信道估计矩阵Η，对上述信号向量y进行解调得到对数似然比λ；对上述对数似然比λ进行解扰，编码块级联及解速率匹配后得到先验信息
将先验信息
发送给上述译码器705；接收译码器705返回的外信息γ；确定上述先验信息
和上述外信息γ的可靠性是否满足要求，若上述先验信息
和上述外信息γ的可靠性不满足要求，则将第一反馈信息
置0；对上述第一反馈信息
进行速率匹配、编码块级联以及加扰处理，以得到上述第二反馈信息
译码器705，用于接收来自处理器703的先验信息
基于最大后验准则，对上述先验信息
进行译码，得到外信息γ；并发送给处理器703。
译码器705的译码过程中还会得到检测结果，本发明实施例的方案不涉及该检测结果，故未作更详细的说明。
以上实施例，在Turbo检测过程中，滤掉可靠性不满足要求的先验信息
和上述外信息γ，将用于解调的第一反馈信息
置0，表示不把任何先验信息传递给解调模块。从而避免第一反馈信息
中不准确的部分对解调模块的影响，增加每次循环信息的可靠性，在保证Turbo检测效果的情况下可以减少Turbo检测的迭代次数，因此以上方案提升了Turbo检测的鲁棒性，从而可以降低Turbo检测的运算复杂度以及处理时延。
更具体地，上述处理器703用于根据第二反馈信息
和上述等效频域信道估计矩阵Η，对上述信号向量y进行解调得到对数似然比λ，包括：根据上述等效频域信道估计矩阵Η对上述信号向量y进行最大似然比检测，利用检测得到的结果和上述第二反馈信息
进行软解调，得到上述对数似然比λ。
以上实施例可以滤掉可靠性不满足要求的先验信息
和上述外信息γ，本发明实施
例还提供了能够进一步提升Turbo检测效率的方案，具体如下：处理器703在以迭代的方式执行操作的过程中，若上述先验信息
和上述外信息γ的可靠性满足要求，则通过
分别对上述先验信息
和上述外信息γ进行加权计算，得到上述第一反馈信息
其中，α(n)、β(n)、
和
通过查询为上述Turbo检测装置预设的调制编码方案MCS获取，n为当前的迭代次数；
然后对上述第一反馈信息
进行速率匹配、编码块级联以及加扰处理，以得到上述第二反馈信息
可选地，本发明实施例还提供了如何确定上述先验信息
和上述外信息γ的可靠性是否满足要求的具体实现方式，需要说明的是，本领域技术人员可以依据其他方式确定上述先验信息
和上述外信息γ的可靠性，因此确定可靠性的方式也并不仅限于本实施例给出的方式，本实施例给出的一个较为方便的判断方式的举例不应理解为对本发明实施例的唯一性限定，具体如下：上述处理器703用于确定上述先验信息
和上述外信息γ的可靠性是否满足要求，包括：计算上述先验信息
和上述外信息γ之和的绝对值，并将上述绝对值与预设的门限值Τ进行比较；
若上述绝对值小于上述门限值Τ，则上述先验信息
和上述外信息γ的可靠性满足要求；
若上述绝对值大于上述门限值Τ，则上述先验信息
和上述外信息γ的可靠性不满足要求；
若上述绝对值等于上述门限值Τ，则根据预设的规则，确定上述先验信息
和上述外信息γ的可靠性满足要求，或者，上述先验信息
和上述外信息γ的可靠性不满足要求；
其中，上述门限值Τ通过查询上述调制编码方案MCS获取。
本实施例中的门限值T可以通过离线仿真的方式确定，在离线仿真过程中，将门限值T与Turbo检测所针对的数据的编码码率及调制阶数关联进行离线仿真。在确定T值以后，将门限值Τ存放到表格中用于查询。
优选地，在本发明实施例中，上述α(n)和β(n)的取值与当前的迭代次数n正相关。
在本发明实施例中，迭代次数越靠后，信息会越可靠因此使用较大的处理因子来增加可靠性高的信息的使用，在保证Turbo检测效果的情况下可以减少Turbo检测的迭代次数，因此以上方案进一步提升了Turbo检测的鲁棒性，从而可以降低Turbo检测的运算复杂度以及处理时延。
以上实施例使用了两组处理因子，其中α(n)和β(n)可以为线性处理因子，
和
可以为非线性处理因子。通过处理因子为线性的处理因子和非线性的处理因子的划分，可以使反馈信息更加有效。其他的加权计算方式也是可以实现的，本发明实施例所采用的计算方式作为一个优选举例，不应理解为对本发明实施例的唯一性限定。
对于以上处理因子α(n)、β(n)、
和
可以由技术人员预先设定，在本发明实施例中，处理因子α(n)、β(n)、
和
可以通过离线仿真的方式确定，在离线仿真过程中，将处理因子α(n)、β(n)、
和
与Turbo检测所针对的数据的编码码率及调制阶数关联进行离线仿真。在确定处理因子α(n)、β(n)、
和
以后，将处理因子α(n)、β(n)、
和
存放到表格中用于查询。
为了在保证Turbo接收机性能的情况下，减少译码内迭代次数(即每次迭代过程中，对上述先验信息
进行译码的译码次数)D的取值，本发明实施例还提出采用非平衡内迭代次数的方案，即每次译码的内迭代次数不同，用Dn表示。通常情况下，译码次数应该遵循前少后多的原则。当Dn取值为0，表示本次迭代不做译码，中间的加解扰、CB级联、速率匹配等操作也不需要做，解调需要的LLR可以直接从解调模块直接反馈回到解调模块。具体如下：上述处理器703对上述先验信息
进行译码的译码次数与当前的迭代次数n正相关。
对于以上译码内迭代次数的取值，可以由技术人员预先设定，在本发明实施例中，内迭代次数可以通过离线仿真的方式确定，在离线仿真过程中，将内迭代次数与Turbo检测所针对的数据的编码码率及调制阶数关联进行离线仿真。在确定内迭代次数以后，将内迭代次数存放到表格中用于查询。
值得注意的是，上述Turbo检测装置结构示意图实施例中，所包括的各个单元只是按照功能逻辑进行划分的，但并不局限于上述的划分，只要能够实现相应的功能即可；另外，各功能单元的具体名称也只是为了便于相互区分，并不用于限制本发明的保护范围。
另外，本领域普通技术人员可以理解实现上述各方法实施例中的全部或部分步骤是可以通过程序来指令相关的硬件完成，相应的程序可以存储于一种计算机可读存储介质中，上述提到的存储介质可以是只读存储器，磁盘或光盘等。
以上仅为本发明较佳的具体实施方式，但本发明的保护范围并不局限于此，任何熟悉本技术领域的技术人员在本发明实施例揭露的技术范围内，可轻易想到的变化或替
换，都应涵盖在本发明的保护范围之内。因此，本发明的保护范围应该以权利要求的保护范围为准。
Claims (12)
- 一种图拨检测方法，其特征在于，包括：以迭代的方式执行如下操作：接收信号向量y和第二反馈信息并获取接收所述信号向量的各个信道的等效频域信道估计矩阵Η；
根据所述第二反馈信息和所述等效频域信道估计矩阵Η，对所述信号向量y进行解调得到对数似然比λ；
对所述对数似然比λ进行解扰，编码块级联及解速率匹配后得到先验信息
基于最大后验准则，对所述先验信息进行译码，得到外信息γ；
确定所述先验信息和所述外信息γ的可靠性是否满足要求，若所述先验信息
和所述外信息γ的可靠性不满足要求，则将第一反馈信息
置0；
对所述第一反馈信息进行速率匹配、编码块级联以及加扰处理，以得到所述第二反馈信息
- 根据权利要求1所述方法，其特征在于，所述方法还包括：算，得到所述第一反馈信息
其中，α(n)、β(n)、
和
通过查询预设的调制编码方案MCS获取，n为当前的迭代次数；
然后对所述第一反馈信息进行速率匹配、编码块级联以及加扰处理，以得到所述第二反馈信息
- 根据权利要求2所述方法，其特征在于，所述确定所述先验信息和所述外信息γ的可靠性是否满足要求，包括：
计算所述先验信息和所述外信息γ之和的绝对值，并将所述绝对值与预设的门限值Τ进行比较；
若所述绝对值小于所述门限值Τ，则所述先验信息和所述外信息γ的可靠性满足要求；
若所述绝对值大于所述门限值Τ，则所述先验信息和所述外信息γ的可靠性不满足要求；
若所述绝对值等于所述门限值Τ，则根据预设的规则，确定所述先验信息和所述 外信息γ的可靠性满足要求，或者，所述先验信息
和所述外信息γ的可靠性不满足要求；
其中，所述门限值Τ通过查询所述调制编码方案MCS获取。 - 根据权利要求2所述方法，其特征在于，所述α(n)和β(n)的取值与当前的迭代次数n正相关。
- 根据权利要求1至4任一所述方法，其特征在于，所述根据所述第二反馈信息和所述等效频域信道估计矩阵Η，对所述信号向量y进行解调得到对数似然比λ，包括：
根据所述等效频域信道估计矩阵Η对所述信号向量y进行最大似然比检测，利用检测得到的结果和所述第二反馈信息进行软解调，得到所述对数似然比λ。
- 根据权利要求1至4任意一项所述方法，其特征在于，所述对所述先验信息进行译码的译码次数与当前的迭代次数n正相关。
- 一种图拨检测装置，其特征在于，包括：解码单元，用于接收信号向量y和来自第二处理单元的第二反馈信息并获取接收所述信号向量的各个信道的等效频域信道估计矩阵Η；根据所述第二反馈信息
和所述等效频域信道估计矩阵Η，对所述信号向量y进行解调得到对数似然比λ；
第一处理单元，用于对所述对数似然比λ进行解扰，编码块级联及解速率匹配后得到先验信息
译码单元，用于基于最大后验准则，对所述第一处理单元得到的先验信息进行译码，得到外信息γ；
可靠性确定单元，用于确定所述先验信息和所述外信息γ的可靠性是否满足要求；
反馈信息处理单元，用于若所述可靠性确定单元确定先验信息和所述外信息γ的可靠性不满足要求，则将第一反馈信息
置0；
所述第二处理单元，用于对所述反馈信息处理单元得到的第一反馈信息进行速率匹配、编码块级联以及加扰处理，以得到所述第二反馈信息
- 根据权利要求7所述装置，其特征在于，所述反馈信息处理单元，还用于若所述可靠性确定单元确定先验信息和所述外信息γ的可靠性满足要求，则通过
分别对所述先验信息
和所述外信息γ进行加权计算，得到所述第一反馈信息
其中，α(n)、β(n)、
和
通过查询为所述图拨检测装置预设的调制编码方案MCS获取，n为当前的迭代次 数。
- 根据权利要求8所述装置，其特征在于，所述可靠性确定单元包括：绝对值计算单元，用于计算所述先验信息和所述外信息γ之和的绝对值；
可靠性确定子单元，用于将所述绝对值与预设的门限值Τ进行比较；若所述绝对值小于所述门限值Τ，则所述先验信息和所述外信息γ的可靠性满足要求；
若所述绝对值大于所述门限值Τ，则所述先验信息和所述外信息γ的可靠性不满足要求；
若所述绝对值等于所述门限值Τ，则根据预设的规则，确定所述先验信息和所述外信息γ的可靠性满足要求，或者，所述先验信息
和所述外信息γ的可靠性不满足要求；
其中，所述门限值Τ通过查询所述调制编码方案MCS获取。 - 根据权利要求8所述装置，其特征在于，所述α(n)和β(n)的取值与当前的迭代次数n正相关。
- 根据权利要求7～10任意一项所述装置，其特征在于，所述解码单元，用于根据所述等效频域信道估计矩阵Η对所述信号向量y进行最大似然比检测，利用检测得到的结果和所述第二反馈信息进行软解调，得到所述对数似然比λ。
- 根据权利要求7～10任意一项所述装置，其特征在于，所述译码单元对所述先验信息进行译码的译码次数与当前的迭代次数n正相关。
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410364464.0 | 2014-07-28 | ||
| CN201410364464.0A CN105337685B (zh) | 2014-07-28 | 2014-07-28 | 一种图拨检测方法及其装置 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2016015592A1 true WO2016015592A1 (zh) | 2016-02-04 |
Family
ID=55216760
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/CN2015/084932 WO2016015592A1 (zh) | 2014-07-28 | 2015-07-23 | 一种图拨检测方法及其装置 |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN105337685B (zh) |
| WO (1) | WO2016015592A1 (zh) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TWI599183B (zh) * | 2016-05-02 | 2017-09-11 | 瑞昱半導體股份有限公司 | 最大可能性偵測器與偵測方法 |
| CN111726195A (zh) * | 2019-03-21 | 2020-09-29 | 华为技术有限公司 | 一种数据传输方法及通信装置 |
| CN116723069A (zh) * | 2023-08-08 | 2023-09-08 | 华侨大学 | 一种多模块迭代Turbo均衡方法、装置、设备、服务器及介质 |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102185682A (zh) * | 2011-06-23 | 2011-09-14 | 西安电子科技大学 | 联合Turbo码与网络编码的中继传输及对应译码方法 |
| CN102571108A (zh) * | 2012-02-24 | 2012-07-11 | 清华大学深圳研究生院 | 一种应用于Turbo乘积码的自适应迭代译码方法 |
| US20130007555A1 (en) * | 2009-12-23 | 2013-01-03 | Mindspeed Technologies, Inc. | Methods, systems, and apparatus for tail termination of turbo decoding |
| CN103873073A (zh) * | 2014-03-20 | 2014-06-18 | 北京遥测技术研究所 | 一种基于并行与加窗结构的Turbo码高速译码实现方法 |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8467466B2 (en) * | 2005-11-18 | 2013-06-18 | Qualcomm Incorporated | Reduced complexity detection and decoding for a receiver in a communication system |
| CN101026434A (zh) * | 2006-02-24 | 2007-08-29 | 中国科学院上海微系统与信息技术研究所 | 一种低复杂度的迭代检测译码方法及装置 |
| CN101409693B (zh) * | 2008-11-06 | 2012-12-26 | 北京韦加航通科技有限责任公司 | 采用Turbo码的单载波迭代均衡实现方法及系统 |
| CN101420291A (zh) * | 2008-12-08 | 2009-04-29 | 北京邮电大学 | 中继系统中网络与信道编码的联合译码方法 |
-
2014
- 2014-07-28 CN CN201410364464.0A patent/CN105337685B/zh active Active
-
2015
- 2015-07-23 WO PCT/CN2015/084932 patent/WO2016015592A1/zh active Application Filing
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130007555A1 (en) * | 2009-12-23 | 2013-01-03 | Mindspeed Technologies, Inc. | Methods, systems, and apparatus for tail termination of turbo decoding |
| CN102185682A (zh) * | 2011-06-23 | 2011-09-14 | 西安电子科技大学 | 联合Turbo码与网络编码的中继传输及对应译码方法 |
| CN102571108A (zh) * | 2012-02-24 | 2012-07-11 | 清华大学深圳研究生院 | 一种应用于Turbo乘积码的自适应迭代译码方法 |
| CN103873073A (zh) * | 2014-03-20 | 2014-06-18 | 北京遥测技术研究所 | 一种基于并行与加窗结构的Turbo码高速译码实现方法 |
Non-Patent Citations (1)
| Title |
|---|
| LI, JUNQIANG ET AL.: "Turbo Space-time Multiuser Detection for Coded Multi-Carrier CDMA Systems", ACTA ELECTRONICA SINICA, vol. 31, no. 4, 30 April 2003 (2003-04-30), pages 489 - 491 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TWI599183B (zh) * | 2016-05-02 | 2017-09-11 | 瑞昱半導體股份有限公司 | 最大可能性偵測器與偵測方法 |
| CN111726195A (zh) * | 2019-03-21 | 2020-09-29 | 华为技术有限公司 | 一种数据传输方法及通信装置 |
| CN111726195B (zh) * | 2019-03-21 | 2021-12-31 | 华为技术有限公司 | 一种数据传输方法及通信装置 |
| CN116723069A (zh) * | 2023-08-08 | 2023-09-08 | 华侨大学 | 一种多模块迭代Turbo均衡方法、装置、设备、服务器及介质 |
| CN116723069B (zh) * | 2023-08-08 | 2023-12-05 | 华侨大学 | 一种多模块迭代Turbo均衡方法、装置、设备、服务器及介质 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105337685B (zh) | 2018-12-07 |
| CN105337685A (zh) | 2016-02-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Fan et al. | Faster-than-Nyquist signaling: An overview | |
| Wu et al. | Iterative multiuser receiver in sparse code multiple access systems | |
| EP2262130A2 (en) | Reduced complexity detection and decoding for a receiver in a communication system | |
| US9319182B2 (en) | Near maximum likelihood spatial multiplexing receiver | |
| KR20120054086A (ko) | 간섭 제거된 스트림들에 대한 채널 및 잡음 적응을 갖는 단일 하드웨어 엘리먼트의 사용에 의한 ｓｉｃ-ｍｉｍｏ 디코딩의 하드웨어 단순화 | |
| WO2015184967A1 (zh) | 抑制同频干扰的方法和装置 | |
| WO2016034051A1 (zh) | 一种干扰抑制方法和装置 | |
| US9083499B1 (en) | Decision feedback solution for channels with low signal to noise ratio | |
| US9118446B2 (en) | Methods and user equipment in a wireless communication network | |
| WO2016015592A1 (zh) | 一种图拨检测方法及其装置 | |
| WO2022242573A1 (zh) | 数据传输方法、装置、通信设备及存储介质 | |
| TW201110601A (en) | Methods and systems using norm approximation for maximum likelihood MIMO decoding | |
| Şahin et al. | Spectrally efficient iterative MU-MIMO receiver for SC-FDMA based on EP | |
| Sarieddeen et al. | Likelihood-based modulation classification for MU-MIMO systems | |
| US8467439B2 (en) | Adaptively switching equalization operations in a node of a wireless network | |
| Seifert et al. | Soft-PIC frequency-domain equalization in iterative MIMO receivers for the LTE-A uplink | |
| JP2019501582A (ja) | 高次ｑａｍ信号を復調するための方法およびシステム | |
| Yamada et al. | Performance of time and frequency compression of faster-than-nyquist signaling in frequency-selective fading channels | |
| Radosavljevic et al. | Probabilistically bounded soft sphere detection for MIMO-OFDM receivers: algorithm and system architecture | |
| TWI593257B (zh) | 處理序列估測的裝置及方法 | |
| Gomaa et al. | Efficient soft-input soft-output detection of dual-layer MIMO systems | |
| Kashoob et al. | Data-aided iterative reweighted LMMSE channel estimation for MIMO OFDM | |
| Kinjo | A low complexity soft demapping method for expanded M-QAM constellations | |
| Choi et al. | Low complexity soft-input soft-output group detection for massive MIMO systems | |
| WO2017049464A1 (zh) | 一种信号解调方法及终端 |
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: 15827507 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: 15827507 Country of ref document: EP Kind code of ref document: A1 |