CN101807954B

Movatterモバイル変換

Info

Publication number: CN101807954B
Application number: CN 201010129747
Authority: CN
Inventors: 宋健; 戴凌龙; 符剑; 杨知行
Original assignee: Tsinghua University; Boeing Co
Current assignee: Tsinghua University; Boeing Co
Priority date: 2010-03-19
Filing date: 2010-03-19
Publication date: 2013-01-30
Anticipated expiration: 2030-03-19
Also published as: CN101807954A

Abstract

The invention discloses a time domain synchronous frequency division multiple access method for uplink multi-users, which comprises the following steps of: forming signal frames at a sending terminal by taking a superframe as a basic unit, wherein a framing way comprises that a leader sequence is inserted in front of L signal subframes; the signal subframes comprise a time domain data block and a post-guard interval; the leader sequence comprises a training sequence, a pre-guard interval and the post-guard interval; the pre-guard interval is the same as the post-guard interval; and the training sequence consists of a m sequence in which time-domain circulating displacement exists between the adjacent users; and sending the formed signal frames. The time domain synchronous frequency division multiple access method solves the problem that superimposed interference is difficult to eliminate between frame headers and frame bodies among the multi-users when TDS-OFDM is used for uplink multiple access, so that a multi-carrier OFDMA and a single-carrier SC-FDMA can perform the transmission of the uplink multi-users by adopting a uniform structure of the signal frames; and multi-user access is realized by the complexity lower than that of the conventional single-user TDS-OFDM system, and the system performance is higher under the condition of movement and particularly low-speed movement.

Description

Translated fromChinese

上行多用户时域同步频分多址接入方法Uplink multi-user time domain synchronous frequency division multiple access method

技术领域technical field

本发明涉及无线通信中的多址接入技术，具体涉及一种采用单载波或者多载波信号的上行多用户时域同步频分多址接入方法。The invention relates to multiple access technology in wireless communication, in particular to an uplink multi-user time-domain synchronous frequency division multiple access method using single carrier or multi-carrier signals.

背景技术Background technique

在无线通信系统中，多址方式允许多个移动用户同时共享有限的频谱资源。频分多址(FDMA)、时分多址(TDMA)和码分多址(CDMA)是无线通信系统中共享有效带宽的三种主要接入技术。而正交频分复用多址(Orthogonal Frequency Division MultiplexingAccess，OFDMA)技术是一种基于OFDM的多址技术，也称为多用户OFDM技术，该技术最先由Sari和Karam提出并应用于有线电视网络中(CATV)。OFDMA将传输带宽划分成正交的子载波集，将不同的子载波集灵活地分配给不同的用户来达到多用户接入的目的。OFDMA是实现OFDM系统中多用户复用和接入的最有效的方式，因此，近年来也倍受关注和研究，在欧洲数字电视回传信道标准DVB-RCT就采用了OFDMA技术，在IEEE 802.16e标准中，OFDMA作为最核心的物理层技术被应用到WiMAX系统中。OFDMA被广泛视为下一代宽带无线通信的首选技术。In wireless communication systems, multiple access schemes allow multiple mobile users to share limited spectrum resources at the same time. Frequency Division Multiple Access (FDMA), Time Division Multiple Access (TDMA) and Code Division Multiple Access (CDMA) are three main access techniques for sharing effective bandwidth in wireless communication systems. Orthogonal Frequency Division Multiple Access (OFDMA) technology is a multiple access technology based on OFDM, also known as multi-user OFDM technology, which was first proposed by Sari and Karam and applied to cable TV. In the network (CATV). OFDMA divides the transmission bandwidth into orthogonal subcarrier sets, and flexibly allocates different subcarrier sets to different users to achieve the purpose of multi-user access. OFDMA is the most effective way to realize multi-user multiplexing and access in the OFDM system. Therefore, it has also attracted much attention and research in recent years. The OFDMA technology is adopted in the European digital TV return channel standard DVB-RCT, and in IEEE 802.16 In the e standard, OFDMA is applied to the WiMAX system as the core physical layer technology. OFDMA is widely regarded as the technology of choice for next-generation broadband wireless communications.

在诸多的OFDMA系统中，OFDM符号均使用循环前缀(CyclicPrefix)用作IDFT块的保护间隔，以便抵消接收信号中可能存在的多径信号，防止码间串扰，该结构称为循环前缀的OFDM(CP-OFDM)。CP-OFDM目前已经得到了广泛的应用，如DAB、DVB-T、IEEE802.11a、HIPERLAN/2、WLAN、WiMAX等，目前绝大多数的B3G/4G提案都使用了CP-OFDM。申请号为01124144.6的中国发明专利申请“正交频分复用调制系统中保护间隔的填充方法”提出了PN序列作为IDFT块的保护间隔的OFDM帧结构，并以此为基础形成了中国地面数字电视标准DTMB的核心技术TDS-OFDM(Time DomainSynchronous OFDM，TDS-OFDM)。相对于CP-OFDM，由于TDS-OFDM系统中的PN序列除了作为OFDM块的保护间隔以外，在接收端还可以被用做信号帧的帧同步、载波恢复与自动频率跟踪、符号时钟恢复、信道估计等用途，因而不需要像CP-OFDM那样再利用专门的导频或前导训练序列来辅助完成同步及信道估计，因而TDS-OFDM可以提供比CP-OFDM高约10％的频谱效率。此外，已有文献证明，TDS-OFDM可以提供比CP-OFDM更好的系统性能。In many OFDMA systems, OFDM symbols use a cyclic prefix (CyclicPrefix) as the guard interval of the IDFT block in order to offset the multipath signal that may exist in the received signal and prevent intersymbol interference. This structure is called OFDM with a cyclic prefix ( CP-OFDM). CP-OFDM has been widely used, such as DAB, DVB-T, IEEE802.11a, HIPERLAN/2, WLAN, WiMAX, etc. Most of the current B3G/4G proposals use CP-OFDM. The Chinese invention patent application with the application number 01124144.6 "Filling Method of Guard Interval in Orthogonal Frequency Division Multiplexing Modulation System" proposes the PN sequence as the OFDM frame structure of the guard interval of the IDFT block, and forms the Chinese terrestrial digital system based on this The core technology of TV standard DTMB is TDS-OFDM (Time Domain Synchronous OFDM, TDS-OFDM). Compared with CP-OFDM, since the PN sequence in the TDS-OFDM system is not only used as the guard interval of OFDM blocks, it can also be used for frame synchronization of signal frames, carrier recovery and automatic frequency tracking, symbol clock recovery, channel Estimation and other purposes, so it is not necessary to use a special pilot or preamble training sequence to assist in synchronization and channel estimation like CP-OFDM, so TDS-OFDM can provide about 10% higher spectral efficiency than CP-OFDM. In addition, it has been documented that TDS-OFDM can provide better system performance than CP-OFDM.

在以CP-OFDM为基础的OFDMA系统中，不同单一用户的信号帧经过多径信道后，帧头保护间隔和帧体IDFT数据块都会产生如图1(a)中阴影所示的“拖尾”，但由于保护间隔是帧体IDFT数据块的循环前缀，直接去掉CP后的帧体接收序列便具有循环特性，从而将帧体IDFT数据块与信道之间的线性卷积转化为循环卷积，并可以通过简单的离散傅立叶变换(DFT)完成帧体IDFT数据块的信道均衡，进而恢复出发送端的帧体数据。在CP-OFDMA系统中，不同用户发送的频域数据相互正交，则按照单用户OFDM系统中完全一致的方法，基站从接收信号中直接去掉保护间隔，所得到的各用户数据线性叠加在一起的接收信号仍具有循环特性，因此，将此信号作DFT变换到频域，则在频域将各用户相互正交的子载波分离，进而可以恢复出不同用户的发送数据。In the OFDMA system based on CP-OFDM, after the signal frames of different single users pass through the multipath channel, the guard interval of the frame header and the IDFT data block of the frame body will produce "smearing" as shown by the shadow in Figure 1(a). ”, but since the guard interval is the cyclic prefix of the frame body IDFT data block, the frame body receiving sequence after directly removing the CP has a cyclic characteristic, thus converting the linear convolution between the frame body IDFT data block and the channel into a circular convolution , and the channel equalization of the frame body IDFT data block can be completed through a simple discrete Fourier transform (DFT), and then the frame body data at the sending end can be recovered. In the CP-OFDMA system, the frequency domain data sent by different users are orthogonal to each other. According to the completely consistent method in the single-user OFDM system, the base station directly removes the guard interval from the received signal, and the obtained user data are linearly superimposed together. The received signal still has a cyclic characteristic. Therefore, if the signal is transformed into the frequency domain by DFT, the orthogonal subcarriers of each user will be separated in the frequency domain, and then the transmitted data of different users can be recovered.

然而，在以TDS-OFDM为基础的OFDMA系统中，情况就要复杂得多。对于以TDS-OFDM为基础的多址接入系统的某单一用户，由于其帧头保护间隔不是循环前缀，而是不同的PN序列，故TDS-OFDM信号帧经过多径信道后，如图1(b)所示，帧头PN序列产生的“拖尾”与帧体OFDM数据产生的“拖尾”完全不同，在接收端信号中直接截取帧体部分所得的序列因为帧头PN序列的“拖尾”干扰将不再具有循环特性，故不能直接运用DFT变换实现信道均衡。因此，在接收端需要采用不断迭代的方法来消除PN序列对帧体OFDM数据的干扰，以便恢复帧体OFDM信号的循环特性。However, in an OFDMA system based on TDS-OFDM, the situation is much more complicated. For a single user in a multiple access system based on TDS-OFDM, since the frame header guard interval is not a cyclic prefix, but a different PN sequence, the TDS-OFDM signal frame passes through the multipath channel, as shown in Figure 1 As shown in (b), the "smear" generated by the frame header PN sequence is completely different from the "smear" generated by the frame body OFDM data. The sequence obtained by directly intercepting the frame body part from the receiving end signal is due to the " Tailing" interference will no longer have a cyclic characteristic, so DFT transformation cannot be directly used to achieve channel equalization. Therefore, an iterative method is needed at the receiving end to eliminate the interference of the PN sequence on the frame-based OFDM data, so as to restore the cycle characteristics of the frame-based OFDM signal.

现有技术中有文献在上述迭代干扰消除方法的基础上，提出了一种基于部分判决辅助的迭代干扰消除方法和一种基于训练序列重构的迭代干扰消除方法，以降低迭代干扰消除的复杂度。然而，上述两种方法均存在两个方面的问题：首先，迭代干扰消除方法需要进行多次迭代，算法较为复杂、运算量很大、实现复杂度较高、接收机的功耗较大；其次，只有在接收端能得到理想的信道估计的情况下，才能完全消除PN序列的影响，否则就会存在残余码间干扰，从而严重影响系统性能。可以说，PN序列与OFDM数据块之间的相互干扰是TDS-OFDM系统的主要难点和不足，这一问题在基于TDS-OFDM的多址系统中尤为突出，因为图1(b)中不同用户发送的信号经过不同的信道后，基站接收端必须分离出不同用户的信号，而不同用户的PN序列(即使不同的用户均采用相同的PN序列)经过不同的信道后均会对OFDM数据部分产生不同的干扰，这些干扰叠加在一起，只有当基站估计出所有用户信道冲击响应后才可以按照上述的迭代干扰消除方法将这些叠加的干扰逐一去除，而要得到所有用户的信道估计，必须首先消除叠加在一起的不同用户的数据对PN序列的干扰(假设信道估计仍通过PN序列来获得)，但是数据对PN序列的叠加干扰消除的前提是已知各用户的发送数据，并得到所有用户的信道估计，这在基站正确分离各用户的信号之前是不可能的。因此，在多用户TDS-OFDM中，不同用户的数据与PN序列之间的干扰叠加在一起，使得原本就比较复杂的迭代干扰消除方法根本不可能消除多个用户叠加在一起的干扰，基站接收端也就不可能分离出不同用户的信号。正是由于这个原因，目前基于TDS-OFDM的多址接入系统的文献甚少。In the prior art, on the basis of the iterative interference cancellation method mentioned above, an iterative interference cancellation method based on partial decision assistance and an iterative interference cancellation method based on training sequence reconstruction are proposed to reduce the complexity of iterative interference cancellation Spend. However, both of the above two methods have two problems: first, the iterative interference cancellation method requires multiple iterations, the algorithm is relatively complex, the calculation is large, the implementation complexity is high, and the power consumption of the receiver is large; secondly, , the influence of the PN sequence can be completely eliminated only when the receiving end can obtain an ideal channel estimate, otherwise there will be residual intersymbol interference, which seriously affects the system performance. It can be said that the mutual interference between the PN sequence and the OFDM data block is the main difficulty and deficiency of the TDS-OFDM system, and this problem is particularly prominent in the multiple access system based on TDS-OFDM, because different users After the transmitted signal passes through different channels, the receiving end of the base station must separate the signals of different users, and the PN sequences of different users (even if different users use the same PN sequence) will generate OFDM data after passing through different channels. Different interferences, these interferences are superimposed together, and only after the base station estimates the channel impulse response of all users can these superimposed interferences be removed one by one according to the above-mentioned iterative interference cancellation method, and to obtain the channel estimation of all users, it must be eliminated first The interference of the data of different users superimposed on the PN sequence (assuming that the channel estimation is still obtained through the PN sequence), but the premise of eliminating the superimposed interference of the data on the PN sequence is that the transmitted data of each user is known, and the data of all users are obtained. Channel estimation, which is not possible until the base station correctly separates the signals of each user. Therefore, in multi-user TDS-OFDM, the interference between the data of different users and the PN sequence is superimposed together, making it impossible to eliminate the interference of multiple users superimposed by the originally relatively complicated iterative interference cancellation method, and the base station receives It is impossible to separate the signals of different users at the terminal. It is for this reason that there are very few literatures on multiple access systems based on TDS-OFDM at present.

发明内容Contents of the invention

(一)要解决的技术问题(1) Technical problems to be solved

本发明要解决的技术问题是：如何解决TDS-OFDM用于上行多址接入时多用户间帧头与帧体之间的叠加干扰难以消除的问题；如何使得多载波OFDMA和单载波SC-FDMA可以采用统一的信号帧结构进行上行多用户传输；以及如何以比传统的单用户TDS-OFDM系统更低的复杂度实现多用户接入，并且在移动条件下特别是低速移动时取得更好的系统性能。The technical problem to be solved in the present invention is: how to solve the problem that the overlapping interference between the frame header and the frame body between multiple users is difficult to eliminate when TDS-OFDM is used for uplink multiple access; how to make multi-carrier OFDMA and single-carrier SC- FDMA can use a unified signal frame structure for uplink multi-user transmission; and how to achieve multi-user access with lower complexity than the traditional single-user TDS-OFDM system, and achieve better performance under mobile conditions, especially low-speed mobile system performance.

(二)技术方案(2) Technical solution

为了解决上述技术问题，本发明提供了一种上行多用户时域同步频分多址接入方法，包括以下步骤：In order to solve the above technical problems, the present invention provides an uplink multi-user time domain synchronous frequency division multiple access method, comprising the following steps:

组帧步骤，在发送端以超帧为基本单元组成信号帧，组成所述超帧的方式为：在L个信号子帧之前插入前导序列；其中，所述信号子帧包括时域数据块和后保护间隔，所述前导序列包括训练序列、前保护间隔和所述后保护间隔，所述前保护间隔与后保护间隔相同；In the framing step, the superframe is used as a basic unit to form a signal frame at the sending end, and the superframe is formed in the following manner: a preamble sequence is inserted before L signal subframes; wherein, the signal subframe includes time domain data blocks and a post-guard interval, the preamble sequence including a training sequence, a pre-guard interval and the post-guard interval, the pre-guard interval being the same as the post-guard interval;

信号发送步骤，将按照所述组帧步骤的方式组成的信号帧进行发送。The signal sending step is to send the signal frame formed according to the method of the framing step.

其中，根据无线信道的相干时间大小或者无线信道的变化速度确定所述信号子帧的个数L。Wherein, the number L of the signal subframes is determined according to the coherence time of the wireless channel or the change speed of the wireless channel.

其中，所述无线信道的多普勒扩展带宽越大，则所述信号子帧的个数L的取值越小；所述无线信道的多普勒扩展带宽越小，则所述信号子帧的个数L的取值越大。Wherein, the greater the Doppler spread bandwidth of the wireless channel is, the smaller the value of the number L of the signal subframe is; the smaller the Doppler spread bandwidth of the wireless channel is, the smaller the value of the signal subframe is The larger the value of the number L is.

其中，所述前保护间隔和后保护间隔均为所述训练序列中的最后K个符号，且K大于或等于无线信道的最大时延扩展。Wherein, both the pre-guard interval and the post-guard interval are the last K symbols in the training sequence, and K is greater than or equal to the maximum delay spread of the wireless channel.

其中，所述训练序列是单载波形式的时域m序列，其中，相邻两个用户m、m+1所采用的时域序列为p_m和p_m+1，p_m+1是p_m经过L_s位循环位移后得到的序列；Wherein, the training sequence is a time domain m sequence in the form of a single carrier, wherein the time domain sequences used by two adjacent users m and m+1 are p_m and p_m+1 , and p_m+1 is p_m The sequence obtained after L_s bit cyclic shift;

其中，所述训练序列是多载波形式的频域序列，不同用户的训练序列占用频域上相互正交的子载波，然后通过逆傅立叶变换得到其对应的时域中的训练序列。Wherein, the training sequence is a multi-carrier frequency domain sequence, and the training sequences of different users occupy mutually orthogonal subcarriers in the frequency domain, and then obtain their corresponding training sequences in the time domain through inverse Fourier transform.

其中，所述时域数据块为IDFT时域数据块。Wherein, the time-domain data block is an IDFT time-domain data block.

其中，所述IDFT时域数据块为正交频分复用多址形式的多载波信号，或者单载波频分多址形式的单载波信号。Wherein, the IDFT time-domain data block is a multi-carrier signal in the form of Orthogonal Frequency Division Multiple Access, or a single-carrier signal in the form of Single-Carrier Frequency Division Multiple Access.

其中，若所述训练序列是多载波形式的频域序列，不同用户均按照一一对应的方式分别占用所述训练序列和IDFT时域数据块中对应的相互正交的子载波。Wherein, if the training sequence is a multi-carrier frequency domain sequence, different users occupy the corresponding mutually orthogonal subcarriers in the training sequence and the IDFT time domain data block in a one-to-one correspondence manner.

其中，在所述信号发送步骤之后还包括多用户循环特性重构步骤，所述多用户循环特性重构步骤包括步骤S1：通过所述信号子帧中的后保护间隔与所述前导序列的后保护间隔之间的加减运算对所述IDFT数据块进行时域上的多用户循环特性重构。Wherein, after the signal sending step, a multi-user cyclic characteristic reconstruction step is also included, and the multi-user cyclic characteristic reconstruction step includes step S1: through the post-guard interval in the signal subframe and the preamble sequence The addition and subtraction operations between the guard intervals perform multi-user cycle characteristic reconstruction in the time domain on the IDFT data block.

其中，所述多用户循环特性重构步骤还包括步骤S2：在接收端将时域上经过多用户循环特性重构后的信号变换到频域，并在各用户对应的子载波集合上选择各用户的频域信号，从而在频域上完成所有用户信号的正交分离。Wherein, the step of reconstructing the multi-user cyclic characteristic further includes step S2: at the receiving end, transform the signal after multi-user cyclic characteristic reconstruction in the time domain to the frequency domain, and select each subcarrier set corresponding to each user The user's frequency domain signal, so as to complete the orthogonal separation of all user signals in the frequency domain.

其中，在所述多用户循环特性重构步骤之后还包括信道估计步骤：将接收到的由各用户的训练序列叠加在一起的信号与本地训练序列做循环相关来得到所有用户信道的信道估计结果，然后通过在时域上设置相互正交的用户窗口来选择各用户的信道估计结果。Wherein, after the multi-user cyclic characteristic reconstruction step, a channel estimation step is also included: performing cyclic correlation on the received signals superimposed by the training sequences of each user and the local training sequence to obtain the channel estimation results of all user channels , and then select the channel estimation results of each user by setting mutually orthogonal user windows in the time domain.

其中，在所述信道估计步骤之后还包括信号恢复步骤，在接收端根据在频域上已经分离出的各用户的接收信号和在时域上分离得到的各用户的信道估计结果，采用单抽头频域均衡方法恢复出发送端所发送的各用户的单载波信号或多载波信号。Wherein, after the channel estimation step, a signal recovery step is also included, at the receiving end, according to the received signal of each user separated in the frequency domain and the channel estimation result of each user separated in the time domain, a single-tap The frequency domain equalization method recovers the single-carrier signal or multi-carrier signal of each user sent by the sending end.

(三)有益效果(3) Beneficial effects

本发明通过设计可用于上行多用户接入的帧结构，解决了TDS-OFDM用于上行多址接入时多用户间帧头与帧体之间的叠加干扰难以消除这一技术难题；同时，由于该帧结构既适用于OFDMA多载波信号，也适用于SC-FDMA单载波信号，从而使得多载波OFDMA和单载波SC-FDMA可以采用统一的信号帧结构进行上行多用户传输；第三，由于本发明设计的新的帧结构，在实现干扰消除和信道估计时不需要迭代运算，因此采用本发明帧结构的上行多用户时域同步频分多址接入(TDS-FDMA)方法还以比传统的单用户TDS-OFDM系统更低的复杂度实现了多用户接入，而且在移动条件下特别是低速移动时能够取得更好的系统性能。By designing a frame structure that can be used for uplink multi-user access, the present invention solves the technical problem that it is difficult to eliminate superimposed interference between frame headers and frame bodies among multi-users when TDS-OFDM is used for uplink multi-access access; at the same time, Since this frame structure is applicable to both OFDMA multi-carrier signals and SC-FDMA single-carrier signals, multi-carrier OFDMA and single-carrier SC-FDMA can use a unified signal frame structure for uplink multi-user transmission; third, because The new frame structure designed by the present invention does not need iterative calculations when realizing interference elimination and channel estimation, so the uplink multi-user time-domain synchronous frequency-division multiple access (TDS-FDMA) method using the frame structure of the present invention is also faster than Traditional single-user TDS-OFDM system realizes multi-user access with lower complexity, and can achieve better system performance under mobile conditions, especially when moving at low speed.

附图说明Description of drawings

图1为基于CP-OFDM的多址接入系统与基于TDS-OFDM的多址接入系统中的帧结构对比(a)与(b)；Fig. 1 is the frame structure comparison (a) and (b) in the multiple access system based on CP-OFDM and the multiple access system based on TDS-OFDM;

图2为本发明实施例的多址接入方法中所设计的帧结构；Fig. 2 is the frame structure designed in the multiple access method of the embodiment of the present invention;

图3为本发明实施例的多址接入方法中接收端所有用户的IDFT数据块的循环特性重构过程的示意图；3 is a schematic diagram of a cyclic characteristic reconstruction process of IDFT data blocks of all users at a receiving end in a multiple access method according to an embodiment of the present invention;

图4为基于本发明实施例的多址接入方法传输OFDMA上行多载波信号和传输SC-FDMA单载波信号的系统结构框图；Fig. 4 is a system structural block diagram of transmitting OFDMA uplink multi-carrier signals and transmitting SC-FDMA single-carrier signals based on the multiple access method of the embodiment of the present invention;

图5为本发明实施例的多址接入方法中的多用户信道估计方法与传统的TDS-OFDM单用户系统中迭代信道估计方法在均方误差(MSE)方面的性能对比结果；5 is a performance comparison result of the multi-user channel estimation method in the multiple access method of the embodiment of the present invention and the iterative channel estimation method in the traditional TDS-OFDM single-user system in terms of mean square error (MSE);

图6为本发明实施例的多址接入方法与传统的TDS-OFDM单用户系统在四种典型多径信道中的误比特率(BER)性能对比结果；Fig. 6 is the bit error rate (BER) performance comparison result of the multiple access method of the embodiment of the present invention and the traditional TDS-OFDM single-user system in four typical multipath channels;

图7为本发明实施例的多址接入方法和传统的TDS-OFDM系统在瑞利衰落信道下的BER性能对比结果。Fig. 7 is a comparison result of BER performance between the multiple access method of the embodiment of the present invention and the traditional TDS-OFDM system under the Rayleigh fading channel.

具体实施方式Detailed ways

下面结合附图和实施例，对本发明的具体实施方式作进一步详细描述。以下实施例用于说明本发明，但不用来限制本发明的范围。The specific implementation manners of the present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

实施例1Example 1

本发明实施例的上行多用户时域同步频分多址接入方法，可支持系统中同时有M个用户发起接入请求并同时处理这M个用户的数据的情况。The uplink multi-user time domain synchronous frequency division multiple access method in the embodiment of the present invention can support the situation that there are M users initiating access requests in the system at the same time and processing the data of the M users at the same time.

在本发明的上行多用户时域同步频分多址接入方法中，各用户的信号帧结构的基本单元是超帧，如图2所示，超帧由前导序列和L个信号子帧构成。In the uplink multi-user TDSFDMA access method of the present invention, the basic unit of the signal frame structure of each user is a superframe, as shown in Figure 2, the superframe is composed of a preamble sequence and L signal subframes .

超帧中的信号子帧个数L可根据无线信道的相干时间大小进行动态调整，通常情况下，用户终端的移动速度越快(或者信道的多普勒扩展越大)，则L的取值越小；反之，用户终端的移动速度越慢(或者信道的多普勒扩展越小)，则L的取值越大。具体而言，超帧中的信号子帧个数L的实施方式可以是：The number L of signal subframes in a superframe can be dynamically adjusted according to the coherence time of the wireless channel. Usually, the faster the mobile speed of the user terminal (or the greater the Doppler spread of the channel), the value of L On the contrary, the slower the moving speed of the user terminal (or the smaller the Doppler spread of the channel), the larger the value of L is. Specifically, the implementation manner of the number L of signal subframes in a superframe may be:

L＝1，即超帧中前导序列后只有一个信号子帧，此时信道估计结果更新的速度最快，适合于信道变化非常快的情况；L=1, that is, there is only one signal subframe after the preamble sequence in the superframe, at this time, the update speed of the channel estimation result is the fastest, which is suitable for the situation where the channel changes very fast;

L＝3，即超帧中前导序列后有3个信号子帧，与LTE中上行帧结构每隔3个符号出现1个导频符号类似，此时信道估计结果更新的速度比较快，适合于信道变化比较快的情况；L=3, that is, there are 3 signal subframes after the preamble sequence in the superframe, which is similar to the uplink frame structure in LTE where 1 pilot symbol appears every 3 symbols. At this time, the update speed of the channel estimation result is relatively fast, which is suitable for When the channel changes rapidly;

L＝12，即超帧中前导序列后有12个信号子帧，此时信道估计结果更新的速度比较慢，适合于信道变化比较慢的情况；或者L=12, that is, there are 12 signal subframes after the preamble in the superframe. At this time, the update speed of the channel estimation result is relatively slow, which is suitable for the case where the channel changes relatively slowly; or

L取其他任意大小正整数，信道估计结果更新的速度与L成反比，当L比较大时，可通过插值，滤波等方式在一定程度上提高信道估计的性能。L takes any other positive integer of any size, and the update speed of the channel estimation result is inversely proportional to L. When L is relatively large, the performance of channel estimation can be improved to a certain extent by means of interpolation and filtering.

超帧中的长度为N_g的前导序列由长度为N_p的m序列构成及其前保护间隔和后保护间隔的，前保护间隔和后保护间隔相同，由K个符号组成，均等同于前导序列中m序列的最后K个符号，即有The preamble sequence of length N_g in the superframe is composed of m sequences of length N_p and its pre-guard interval and post-guard interval. The pre-guard interval and post-guard interval are the same and consist of K symbols, which are equal to the preamble The last K symbols of the m sequence in the sequence, that is, there are

N_g＝N_p+2K. (1)N_g =N_p +2K. (1)

前导序列中的m序列(训练序列的实施方式之一)可用线性反馈移位寄存器(LFSR)来生成，所需m序列的长度N_p取决于上行多用户时域同步频分多址接入系统中可同时支持的用户数M和各用户的所经历信道的最大时延扩展l_max，m序列的长度应当满足N_p≥M·l_max。具体而言，前导序列中的m序列长度的实施方式可以是：N_p＝255；N_p＝511；N_p＝1023；或N_p＝2047。The m-sequence in the preamble sequence (one of the implementations of the training sequence) can be generated by a linear feedback shift register (LFSR), and the length N_p of the required m-sequence depends on the uplink multi-user time-domain synchronous frequency division multiple access system The number of users M that can be supported at the same time and the maximum delay spread l_max of the channel experienced by each user, the length of the m sequence should satisfy N_p ≥ M·l_max . Specifically, the embodiment of the sequence length of m in the leader sequence may be: N_p =255; N_p =511; N_p =1023; or N_p =2047.

前导序列中的前保护间隔和后保护间隔是为了在多径情况下对m序列进行保护，同时用于超帧中IDFT数据块的循环特性重构，前保护间隔和后保护间隔的长度K应大于或等于各用户信道的最大时延扩展l_max，同时小于或等于m序列的长度N_p，即l_max≤K≤N_p。具体而言，前导序列中的前保护间隔和后保护间隔的长度K的实施方式可是：K＝l_max；

其中

表示向上取整；K＝N_p；或者K取满足l_max≤K≤N_p的任意正整数。The pre-guard interval and post-guard interval in the preamble sequence are used to protect the m-sequence under multipath conditions, and are used for the cyclic characteristic reconstruction of the IDFT data block in the superframe. The length K of the pre-guard interval and the post-guard interval should be It is greater than or equal to the maximum delay spread l_max of each user channel, and less than or equal to the length N_p of the m-sequence, that is, l_max ≤ K ≤ N_p . Specifically, the implementation of the length K of the pre-guard interval and the post-guard interval in the preamble sequence may be: K=l_max ;

in

Indicates rounding up; K=N_p ; or K takes any positive integer satisfying l_max ≤ K ≤ N_p .

对于前导序列的m序列，相邻两个用户m和用户m+1个所采用的m序列p_m，0和p_m+1，0，p_m+1，0是p_m，0经过L_s位循环位移后得到的，即For the m-sequence of the leading sequence, the m-sequence p_m,0 and p m+_1,0 adopted by two adjacent users m and user m+1, p_m+1,0 is p_m,0 after L_s obtained after bit rotation, that is,

${p p}_{m m + + 1,0 1,0} = = {p p}_{m m,, 00}^{{Ξ ξ}_{{L L}_{s the s}}},, - - - - - - ((22))$

其中

表示对序列

进行L_s位循环位移：in

Represents a pair sequence

Carry out L_s bit cyclic shift:

${p p}_{m m}^{{Ξ ξ}_{{L L}_{s the s}}} = = [[{p p}_{m m} (({L L}_{s the s})),, {p p}_{m m} (({L L}_{s the s} + + 11)),, . . . . . .,, {p p}_{m m} (({N N}_{p p} - - 11)),, {p p}_{m m} ((00)),, {p p}_{m m} ((11)),, . . . . . .,, {p p}_{m m} (({L L}_{s the s} - - 11))]] . . - - - - - - ((33))$

m序列循环位移的位数L_s取决取决于上行多用户时域同步频多址接入系统中可同时支持的用户数M、各用户的所经历信道的最大时延扩展l_max、m序列的长度N_p。具体而言，m序列循环位移的位数L_s的实施方式可是：L_s＝l_max，其中l_max为各用户的所经历信道的最大时延扩展；优选地，取

其中表示向上取整。The number of bits L_s of the m-sequence cyclic shift depends on the number of users M that can be supported simultaneously in the uplink multi-user TDSFMA system, the maximum time delay spread l_max of the channels experienced by each user, and the m-sequence length N_p . Specifically, the implementation of the number of bits L_s of the m-sequence cyclic shift can be: L_s =l_max , where l_max is the maximum time delay spread of the channel experienced by each user; preferably, take

in Indicates rounding up.

超帧中的信号子帧由IDFT时域数据块和后保护间隔构成。信号子帧中的IDFT数据块既可以是正交频分复用多址(OFDMA)形式的多载波信号，也可以是单载波频分多址(SC-FDMA)形式的单载波信号。具体而言，信号子帧由IDFT时域数据块的产生实施方式可是：A signal subframe in a superframe consists of IDFT time-domain data blocks and a post-guard interval. The IDFT data block in the signal subframe can be a multi-carrier signal in the form of Orthogonal Frequency Division Multiple Access (OFDMA), or a single-carrier signal in the form of Single-Carrier Frequency Division Multiple Access (SC-FDMA). Specifically, the implementation of the generation of the signal subframe by the IDFT time domain data block can be:

实施方式一：IDFT时域数据块是OFDMA形式的多载波信号，即各用户的频域信号首先经过子载波分配后再进行IDFT变换到时域，其信号产生的具体方法如下：首先，用户m在第i个信号子帧传输的频域信号经过子载波分配后将扩展得到N维的频域矢量

Embodiment 1: The IDFT time domain data block is a multi-carrier signal in the form of OFDMA, that is, the frequency domain signal of each user first passes through subcarrier allocation and then undergoes IDFT transformation to the time domain. The specific method of its signal generation is as follows: first, the user m The frequency domain signal transmitted in the i-th signal subframe After subcarrier allocation, it will be expanded to obtain an N-dimensional frequency domain vector

${X x}_{m m,, i i} ((n no)) = = \{\begin{matrix} {D D.}_{m m,, i i} ((n no)) & n no &Element; &Element; {Γ Γ}_{m m} \\ 00 & n no &NotElement; &NotElement; {Γ Γ}_{m m} \end{matrix},, - - - - - - ((44))$

其中Γ_m表示第m个用户的子载波集合，为保持个用户数据的正交性，各用户的子载波集合{Γ_m}_m＝1^M应当保持正交，即

(i≠j)；然后，将频域信号

经过N点IDFT变换后得到多载波形式的IDFT时域数据块

where Γ_m represents the subcarrier set of the mth user. In order to maintain the orthogonality of user data, the subcarrier set {Γ_m }_{m = 1}^M of each user should be kept orthogonal, that is

(i≠j); Then, the frequency domain signal

After N-point IDFT transformation, the IDFT time-domain data block in the form of multi-carrier is obtained

${x x}_{m m,, i i} ((n no)) = = \frac{11}{\sqrt{N N}} {Σ Σ}_{k k = = 00}^{N N - - 11} {X x}_{m m,, i i} ((k k)) exp exp {{j j \frac{22 πnk πnk}{N N}}} - - - - - - ((55))$

实施方式二：IDFT时域数据块是SC-FDMA形式的单载波信号，各用户的时域信号首先经过DFT变换到频域，然后经过子载波分配后进行IDFT变换回时域，其信号产生的具体方法如下：首先，用户m在第i个信号子帧传输的时域信号

经过L_m点DFT变换后得到其频域信号

Embodiment 2: The IDFT time-domain data block is a single-carrier signal in the form of SC-FDMA. The time-domain signals of each user are first transformed into the frequency domain through DFT, and then IDFT-transformed back to the time domain after subcarrier allocation. The specific method is as follows: First, the time-domain signal transmitted by user m in the i-th signal subframe

After the L_m point DFT transformation, the frequency domain signal is obtained

${D D.}_{m m,, i i} ((n no)) = = \frac{11}{\sqrt{{L L}_{m m}}} {Σ Σ}_{k k = = 00}^{{L L}_{m m} - - 11} {d d}_{m m,, i i} ((k k)) exp exp {{j j \frac{22 πnk πnk}{{L L}_{m m}}}} . . - - - - - - ((66))$

其次，频域信号经过与(4)类似的子载波分配后将扩展得到N维的频域矢量

最后，将频域信号经过N点IDFT变换后得到单载波形式的IDFT时域数据块Second, the frequency domain signal After subcarrier allocation similar to (4), it will be expanded to obtain an N-dimensional frequency domain vector

Finally, the frequency domain signal After N-point IDFT transformation, the IDFT time domain data block in the form of single carrier is obtained

上述IDFT时域数据块的产生的具体实施方式的选择可以通过图4各用户发射机中的单多载波信号选择开关来控制。The selection of the specific implementation manner of generating the above-mentioned IDFT time-domain data block can be controlled by the single-multi-carrier signal selection switch in each user transmitter in FIG. 4 .

对于信号子帧中的IDFT时域数据块，不论其是OFDMA形式的多载波信号，还是SC-FDMA形式的单载波信号，其生成过程都涉及到正交子载波分配。对于子载波分配方案，其具体的实施方式可是：For IDFT time-domain data blocks in a signal subframe, whether it is a multi-carrier signal in the form of OFDMA or a single-carrier signal in the form of SC-FDMA, its generation process involves orthogonal subcarrier allocation. For the subcarrier allocation scheme, its specific implementation method can be:

实施方式一：连续子载波分配，即将整个带宽分配成多个连续的子载波组，每个组里有连续相邻的子载波，给每个用户分配一个或者多个子载波组来传输其信号帧；Embodiment 1: Contiguous subcarrier allocation, that is, the entire bandwidth is allocated into multiple contiguous subcarrier groups, each group has contiguous adjacent subcarriers, and each user is assigned one or more subcarrier groups to transmit its signal frame ;

实施方式二：交织子载波分配，即采用交织的子载波为M个不同用户分配其占用的子载波：具有间隔的一组子载波分配给同一用户，使得每个用户的子载波均匀分布在给定带宽上，不同子信道的载波以规则的方式交织；Embodiment 2: Allocation of interleaved subcarriers, that is, using interleaved subcarriers to allocate subcarriers occupied by M different users: a group of subcarriers with intervals are allocated to the same user, so that the subcarriers of each user are evenly distributed among the given subcarriers. On a fixed bandwidth, carriers of different subchannels are interleaved in a regular manner;

实施方式三：随即子载波分配，即采用伪随机的方式为M个不同用户分配其占用的子载波，具有不等间隔的一组子载波分配给同一用户，使得每个用户的子载波非均匀分布在给定带宽上，不同子信道的载波以伪随机的方式交织。Implementation Mode 3: Random subcarrier allocation, that is, assign subcarriers occupied by M different users in a pseudo-random manner, and a group of subcarriers with unequal intervals are allocated to the same user, so that the subcarriers of each user are non-uniform Distributed over a given bandwidth, carriers of different subchannels are interleaved in a pseudo-random manner.

实施例2Example 2

本实施例给出上行多用户时域同步频分多址接入方法的中的多用户循环特性重构方法。This embodiment provides a multi-user cyclic characteristic reconstruction method in an uplink multi-user time-domain synchronous frequency division multiple access method.

图3中，(a)为M个用户的接收信号示意图，(b)为第m个用户的循环特性重构示意图，(c)为M个用户的联合循环特性重构示意图。如图3所示，对于上行多用户时域同步频分多址接入系统中的全部M个用户，M个用户的信号将在接收端叠加在一起。前导序列对应的接收信号

和信号子帧i对应的接收信号

可分别表示为：In Fig. 3, (a) is a schematic diagram of received signals of M users, (b) is a schematic diagram of cyclic characteristic reconstruction of the mth user, and (c) is a schematic diagram of joint cyclic characteristic reconstruction of M users. As shown in FIG. 3 , for all M users in the uplink multi-user TDSFDMA access system, the signals of the M users will be superimposed at the receiving end. The received signal corresponding to the preamble sequence

Received signal corresponding to signal subframe i

Can be expressed as:

$\{\begin{matrix} {r r}_{total total,, 00} ((n no)) = = {Σ Σ}_{m m = = 11}^{M m} {r r}_{m m,, 00} ((n no)) & 00 \leq \leq n no \leq \leq {N N}_{p p} + + 22 K K - - 11 \\ {r r}_{total total,, i i} ((n no)) = = {Σ Σ}_{m m = = 11}^{M m} {r r}_{m m,, i i} ((n no)) & 00 \leq \leq n no \leq \leq N N + + K K - - 11 \end{matrix} - - - - - - ((77))$

为了重构信号子帧i中IDFT数据块的循环特性，将该子帧IDFT数据块对应的接收信号{r_total，i(n)}_n＝0^N-1首先加上本帧IDFT数据块后保护间隔所对应的接收信号{r_total，i(n)}_n＝N^N+K-1，然后减去前导序列中后保护间隔所对应的接收序列

从而得到一个新的序列

{y^{'}}_{total, i} = {{y^{'}}_{total, i} (n)}_{n = 0}^{N - 1} :

In order to reconstruct the cyclic characteristics of the IDFT data block in the signal subframe i, the received signal {r_{total, i} (n)}_n=0^N-1 corresponding to the IDFT data block of the subframe is firstly added to the IDFT data block of this frame The received signal corresponding to the guard interval {r_{total, i} (n)}_n=N^N+K-1 , and then subtract the received sequence corresponding to the rear guard interval in the preamble sequence

resulting in a new sequence

{the y}^{'}_{total, i} = {{the y}^{'}_{total, i} (no)}_{no = 0}^{N - 1} :

${y the y}^{' '}_{total total,, i i} ((n no)) = = \{\begin{matrix} {r r}_{total total,, i i} ((n no)) + + {r r}_{total total,, i i} ((n no + + N N)) - - {r r}_{total total,, o o} ((n no + + {N N}_{p p} + + K K)) & 00 \leq \leq n no \leq \leq K K - - 11 \\ {r r}_{total total,, i i} ((n no)) & K K \leq \leq n no \leq \leq N N - - 11 \end{matrix} - - - - - - ((88))$

将式(8)代入式(7)，可得：Substituting formula (8) into formula (7), we can get:

${y the y}^{' '}_{total total,, i i} ((n no)) = = \{\begin{matrix} {Σ Σ}_{m m = = 11}^{M m} {r r}_{m m,, i i} ((n no)) + + {r r}_{m m,, i i} ((n no + + N N)) - - {r r}_{m m,, o o} ((n no + + {N N}_{p p} + + K K)) & 00 \leq \leq n no \leq \leq K K - - 11 \\ {Σ Σ}_{m m = = 11}^{M m} {r r}_{m m,, i i} ((n no)) & K K \leq \leq n no \leq \leq N N - - 11 \end{matrix} - - - - - - ((99))$

假设信道在一个超帧中基本不变，则前导序列中的m序列、后保护间隔以及各信号子帧中的后保护间隔经过多径信道后均产生相同的拖尾，在图3(a)中用相同的阴影来表示。因此，我们可以得到：Assuming that the channel is basically unchanged in a superframe, the m-sequence in the preamble, the post-guard interval, and the post-guard interval in each signal subframe all produce the same smear after passing through the multipath channel, as shown in Figure 3(a) are represented by the same shading. Therefore, we can get:

$\{\begin{matrix} {r r}_{m m,, i i} ((n no + + N N)) - - {r r}_{m m,, o o} ((n no + + {N N}_{p p} + + K K)) = = {y the y}_{m m,, i i} ((n no + + N N)) - - {g g}_{m m,, o o} ((n no + + {N N}_{p p})) & 00 \leq \leq n no < < {l l}_{m m} - - 11 \\ {r r}_{m m,, i i} ((n no)) = = {y the y}_{m m,, i i} ((n no)) + + {g g}_{m m,, o o} ((n no + + {N N}_{p p})) & 00 \leq \leq n no < < {l l}_{m m} - - 11 \end{matrix} - - - - - - ((1010))$

其中，

和

分别是前导序列中的m序列

和信号子帧i中的IDFT数据块

经过信道

后的响应，l_m为用户m所经历的信道h_m，i的最大时延扩展。将(10)代入(9)，可得：in,

and

are the m sequences in the leading sequence

and the IDFT data block in signal subframe i

through the channel

After the response, l_m is the maximum delay spread of channel h_m,i experienced by user m. Substituting (10) into (9), we can get:

${y the y}^{' '}_{total total,, i i} ((n no)) = = {Σ Σ}_{m m = = 11}^{M m} {y the y}^{' '}_{m m,, i i} ((n no)),, 00 \leq \leq n no \leq \leq N N - - 11 - - - - - - ((1111))$

其中in

${y the y}^{' '}_{m m,, i i} ((n no)) = = \{\begin{matrix} {y the y}_{m m,, i i} ((n no + + N N)) + + {y the y}_{m m,, i i} ((n no)) & 00 \leq \leq n no < < {l l}_{m m} - - 11 \\ {y the y}_{m m,, i i} ((n no)) & {l l}_{m m} - - 11 \leq \leq n no \leq \leq N N - - 11 \end{matrix} - - - - - - ((1212))$

得到序列

的过程可由图3(b)所示，从(12)及图3(b)可以看出，新序列y′_m，i的形式与CP-OFDM系统中的IDFT接收信号完全一致。因此，新序列y′_m，i已完成了对用户m在第i子帧的IDFT数据块的循环特性重构。按照完全相同的方式，图3(b)也给出了超帧中其他子帧的IDFT数据块的循环特性重构过程。get sequence

The process of can be shown in Fig. 3(b). It can be seen from (12) and Fig. 3(b) that the form of the new sequence y′_m,i is completely consistent with the IDFT received signal in the CP-OFDM system. Therefore, the new sequence y′_m,i has completed the reconstruction of the cyclic characteristics of the IDFT data block of the i-th subframe of the user m. In exactly the same way, Fig. 3(b) also shows the reconstruction process of the cyclic characteristics of the IDFT data blocks of other subframes in the superframe.

得到新序列y′_total，i的过程可由图3(c)所示。显然，由于y′_total，i是已完成循环特性重构的序列集合{y′_m，i}_m＝1^M的线性组合，因此式(11)以及图3(c)中的y′_total，i也具有循环特性，这与图1所示的CP-OFDMA系统中接收信号的形式完全一致。至此，利用本发明提出的帧结构，通过接收端简单的加减运算便完成了上行多用户时域同步频分多址接入系统中接收IDFT数据块的联合循环特性重构。The process of obtaining the new sequence_y'total,i can be shown in Figure 3(c). Apparently, since y′_{total, i} is a linear combination of the sequence set {y′_{m, i} }_m=1^M that has completed the cyclic characteristic reconstruction, so y′_{total in formula (11) and Fig. 3(c), i} also has a cyclic characteristic, which is completely consistent with the form of the received signal in the CP-OFDMA system shown in Figure 1. So far, using the frame structure proposed by the present invention, the combined cycle characteristic reconstruction of receiving IDFT data blocks in the uplink multi-user TDSFDMA access system is completed through simple addition and subtraction operations at the receiving end.

对上述加减运算处理后得到的多用户时域线性叠加的具有循环特性的信号y′_total，i作DFT变换得到其频域信号

然后按照与发送端一一对应的子载波分配方式，在频域选取属于各自用户的信号：After the above addition and subtraction operations, the multi-user time-domain linearly superimposed signal_y'total,i with cyclic characteristics is obtained by DFT transformation to obtain its frequency-domain signal

Then select the signals belonging to the respective users in the frequency domain according to the one-to-one subcarrier allocation method corresponding to the sending end:

${Y Y}^{' '}_{m m,, i i} ((n no)) = = \{\begin{matrix} {Y Y}^{' '}_{total total,, i i} ((n no)) & n no &Element; &Element; {Γ Γ}_{m m} \\ 00 & n no &NotElement; &NotElement; {Γ Γ}_{m m} \end{matrix},, 11 \leq \leq m m \leq \leq M m - - - - - - ((1313))$

其中，n为子载波的编号，m代表不同的用户。式中Y′_m，i(n)为经过循环前缀重构得到的信号y′_m，i(n)的频域表示，其矢量表示为 ${Y^{'}}_{m, i} = [{Y^{'}}_{m, i} (n)] |_{n &Element; Γ_{m}} .$ Wherein, n is a subcarrier number, and m represents a different user. where Y′_{m, i} (n) is the frequency domain representation of the signal y′_{m, i} (n) obtained through cyclic prefix reconstruction, and its vector representation is ${Y^{'}}_{m, i} = [{Y^{'}}_{m, i} (no)] |_{no &Element; Γ_{m}} .$

由于不同用户占用的子载波之间是相互正交的(

(i≠j))，因此，在时域线性叠加的多用户信号在频域被正交分离开来，从而实现了多址信号的频域正交分离。Since the subcarriers occupied by different users are mutually orthogonal (

(i≠j)), therefore, the multi-user signals linearly superimposed in the time domain are separated orthogonally in the frequency domain, thereby realizing the orthogonal separation of the frequency domain of the multiple access signals.

可见，在上行多用户时域同步频分多址接入系统的接收端，通过一次简单的加减运算就可以同时重构出所有用户在时域线性叠加的帧体IDFT数据块的循环特性，这不但避免了多址接入系统中为分别重构各用户接收数据的循环特性可能带来的高复杂度的运算及其误差，而且该多址接入系统中的循环特性重构方法比传统TDS-OFDM系统中单用户接收数据的迭代循环重构方法还要简单得多。具有循环特性的信号经过DFT变换后，即可在频域将各个用户的信号正交分离。It can be seen that at the receiving end of the uplink multi-user Time Domain Synchronous Frequency Division Multiple Access system, the cyclic characteristics of the frame body IDFT data blocks linearly superimposed by all users in the time domain can be reconstructed at the same time through a simple addition and subtraction operation. This not only avoids the highly complex calculations and errors that may be caused by reconstructing the cyclic characteristics of the data received by each user in the multiple access system, but also the cyclic characteristic reconstruction method in the multiple access system is better than the traditional method. The iterative cyclic reconstruction method of single user receiving data in TDS-OFDM system is much simpler. After the signal with cyclic characteristic undergoes DFT transformation, the signals of each user can be separated orthogonally in the frequency domain.

实施例3Example 3

本实施例给出采用与实施例2相同方法在接收端在频域上完成了所有用户信号的正交分离后，通过信道估计恢复出发端原始数据的过程。This embodiment presents the process of recovering the original data at the originating end through channel estimation after the receiving end completes the orthogonal separation of all user signals in the frequency domain using the same method as inEmbodiment 2.

经过循环特性重构并在频域上正交分离的用户信号Y′_m，i可表示为：The user signal Y′_m,i which is reconstructed by cyclic characteristics and separated orthogonally in the frequency domain can be expressed as:

Y′_m，i(n)＝H_m，i(n)·X_m，i(n)+W_m，i(n) 0≤n≤N-1 (14)Y'_{m, i} (n) = H_{m, i} (n) X_{m, i} (n) + W_{m, i} (n) 0≤n≤N-1 (14)

其中W_m，i(n)为噪声，

为第m个用户在第i帧经过的信道h_m，i的DFT变换。where W_m,i (n) is the noise,

is the DFT transformation of the channel h_m,i passed by the mth user in the ith frame.

可见，在频域得到的可正交分离的多址信号是不同用户的发送信号经过各自信道后得到的信号，因此，要恢复出各用户的发送信号，还必须得到不同用户各自的信道估计，然后通过频域信道均衡后就可以恢复发端的原始数据。It can be seen that the orthogonally separable multiple access signals obtained in the frequency domain are the signals obtained after the transmitted signals of different users pass through their respective channels. Therefore, to recover the transmitted signals of each user, the channel estimates of different users must also be obtained. Then the original data at the sending end can be recovered after channel equalization in the frequency domain.

各用户的信道估计可通过本地m序列与前导序列中接收到的m序列做循环相关来得到。由于帧结构的设计中相邻用户m和m+1在前导序列中所采用的m序列p_m+1，0和p_m，0满足式(2)中所示的L_s位循环位移关系，那么The channel estimation of each user can be obtained by doing circular correlation between the local m-sequence and the m-sequence received in the preamble. Since the m-sequence p_m+1,0 and p_m,0 adopted by adjacent users m and m+1 in the preamble sequence in the design of the frame structure satisfy the L_s- bit cyclic displacement relationship shown in formula (2), So

${p p}_{m m,, 00} = = {p p}_{1,0 1,0}^{{Ξ ξ}_{((m m - - 11)) {L L}_{s the s}}},, 11 \leq \leq m m \leq \leq M m - - - - - - ((1515))$

由于m序列具有非常良好的自相关性，即有：Since the m-sequence has very good autocorrelation, that is:

${p p}_{m m,, 00} &CircleTimes; &CircleTimes; {p p}_{m m,, 00} \approx \approx {N N}_{p p} \cdot \cdot δ δ ((n no)),, 11 \leq \leq m m \leq \leq M m - - - - - - ((1616))$

其中

表示循环相关。那么，p_j，0和p_k，0之间的互相关函数则为：in

Indicates circular correlation. Then, the cross-correlation function between p_j,0 and p_k,0 is:

${p p}_{j j,, 00} &CircleTimes; &CircleTimes; {p p}_{k k,, 00} \approx \approx {N N}_{p p} \cdot \cdot δ δ [[n no - - ((k k - - j j)) \cdot \cdot {L L}_{s the s}]],, 11 \leq \leq k k,, j j \leq \leq M m - - - - - - ((1717))$

由于各用户的前导序列中均包含了长度为K的m序列的前保护间隔，当

时，接收端接收到的M个m序列的叠加序列

本身就满足循环特性，q_m可表示为：Since the preamble sequence of each user contains the pre-guard interval of the m-sequence of length K, when

When , the superposition sequence of M m-sequences received by the receiving end

itself satisfies the cyclic characteristics, q_m can be expressed as:

${q q}_{m m} = = {Σ Σ}_{m m = = 11}^{M m} {p p}_{m m,, 00} &CircleTimes; &CircleTimes; {h h}_{m m,, i i} + + {v v}_{m m} - - - - - - ((1818))$

其中矢量v_m表示用户m的接收m序列中的高斯噪声。where the vector v_m represents the Gaussian noise in user m's received m sequence.

将接收到叠加序列q_m与本地m序列之一p_1，0作循环相关可得：The circular correlation between the received superposition sequence q_m and one of the local m sequences p_1,0 can be obtained:

${q q}_{m m} &CircleTimes; &CircleTimes; {p p}_{1,0 1,0} \approx \approx (({Σ Σ}_{m m = = 11}^{M m} {p p}_{m m,, 00} &CircleTimes; &CircleTimes; {h h}_{m m,, i i})) &CircleTimes; &CircleTimes; {p p}_{1,0 1,0} \approx \approx {N N}_{p p} \cdot &Center Dot; {Σ Σ}_{m m = = 11}^{M m} {h h}_{m m,, i i} \cdot &Center Dot; δ δ [[n no - - ((m m - - 11)) \cdot &Center Dot; {L L}_{s the s}]] - - - - - - ((1919))$

由式(19)可知，h_m，i在时域上被搬移了(m-1)·L_s，若l_max≤L_s且M·L_s≤N_p，那么被搬移后的h_m，i(1≤m≤M)与在时域上互不重叠，也即M个用户的信道h_m，i(1≤m≤M)在时域上被正交分离，从而可得到用户m的信道估计

From equation (19), it can be seen that h_{m, i} has been shifted by (m-1)·L_s in the time domain, if l_max ≤L_s and M·L_s ≤N_p , then the shifted h_{m, i} (1≤m≤M) does not overlap with each other in the time domain, that is, the channels h_{m of M users, i} (1≤m≤M) are separated orthogonally in the time domain, so that the user m’s channel estimation

可见，原本在时域上混叠在一起无法分离的多用户信道，经过上述本地m序列与前导序列中接收到的m序列做一次循环相关后，便可在时域上完成正交分离。It can be seen that the multi-user channels that are aliased together in the time domain and cannot be separated can be separated orthogonally in the time domain after performing a circular correlation between the local m-sequence and the m-sequence received in the preamble sequence.

值得注意的是，用于上述多用户联合信道估计的本地m序列可以是M个用户中任何一个用户的前导序列所采用的m序列p_u，0(1≤u≤M)，这是因为It is worth noting that the local m-sequence used for the above multi-user joint channel estimation can be the m-sequence p_u,0 (1≤u≤M) used by the preamble sequence of any one of the M users, because

${q q}_{m m} &CircleTimes; &CircleTimes; {p p}_{u u,, 00} \approx \approx (({Σ Σ}_{m m = = 11}^{M m} {p p}_{m m,, 00} &CircleTimes; &CircleTimes; {h h}_{m m,, i i})) &CircleTimes; &CircleTimes; {p p}_{u u,, 00} \approx \approx {N N}_{p p} \cdot &Center Dot; {Σ Σ}_{m m = = 11}^{M m} {h h}_{m m,, i i} \cdot \cdot δ δ [[n no - - ((m m - - u u)) \cdot &Center Dot; {L L}_{s the s}]] - - - - - - ((2020))$

将时域上经过正交分离得到的各用户的信道估计

做DFT变换到频域得到

然后通过简单的单抽头频域均衡器去均衡经过正交分离得到的各用户接收信号，则可恢复出发送端各用户的发射信号：The channel estimation of each user obtained by orthogonal separation in the time domain

Do DFT transform to frequency domain to get

Then use a simple single-tap frequency domain equalizer to equalize the received signals of each user obtained through orthogonal separation, and then restore the transmitted signal of each user at the sending end:

${\overset{^^}{X x}}_{m m,, i i} = = \frac{{Y Y}^{' '}_{m m,, i i}}{{\overset{^^}{H h}}_{m m,, i i}} - - - - - - ((21 twenty one))$

其中即为第m个用户在的第i个信号子帧中发送的频域信号X_m，i的估计。对频域均衡后的频域信号

进行判决，其具体实施方式可以是：若发送端IDFT数据块采用的是OFDMA形式的多载波信号，则对直接进行判决；若发送端IDFT数据块采用的是SC-FDMA形式的单载波信号，则需要对

做IDFT变换后得到时域信号

然后再进行判决。in That is, the estimate of the frequency domain signal X_m,i sent by the mth user in the i'th signal subframe. The frequency domain signal after frequency domain equalization

Make a decision, and its specific implementation method can be: if what the sending end IDFT data block adopts is the multi-carrier signal of OFDMA form, then to Make a decision directly; if the IDFT data block at the sending end uses a single-carrier signal in the form of SC-FDMA, you need to

After IDFT transformation, the time domain signal is obtained

Then come the judgment.

上述判决方式的具体实施方式的选择可以通过图4接收机中的单多载波信号选择开关来控制。The selection of the specific implementation manner of the above decision manner can be controlled by the single-multi-carrier signal selection switch in the receiver in FIG. 4 .

至此，我们已分离并恢复出上行多用户时域同步频分多址接入系统中发端各用户的发送数据。图4给出了基于发明上行多用户时域同步频分多址接入系统的传输OFDMA上行多载波信号和传输SC-FDMA上行单载波信号的系统框图。So far, we have separated and recovered the transmission data of each user at the originating end in the uplink multi-user TDSFDMA access system. Fig. 4 shows a system block diagram of OFDMA uplink multi-carrier signal transmission and SC-FDMA uplink single-carrier signal transmission based on the inventive uplink multi-user time domain synchronous frequency division multiple access system.

为了分析本发明所提出的上行多用户时域同步频分多址接入方法的复杂度，表1给出了本发明的多址接入方法与直接迭代法(参见文献[1]J.Wang，Z.Yang，C.Pan，J.Song，and L.Yang，“Iterativepadding substruction of the PN sequence for the TDS-OFDM overbroadcasting channels，”IEEE Trans.Consumer Electron.，vol.51，no.4，pp.1148-1152，Nov.2005)、基于部分判决反馈的迭代干扰消除方法(参见文献[2]Shigang Tang，Kewu Peng，Ke Gong，et al.，″NovelDecision-Aided Channel Estimation for TDS-OFDM Systems，″in Proc.IEEE International Conference on Communications(ICC′08)，May.2008，vol.1，pp.946-950)、基于训练序列重构的方法(参见文献[3]FangYang，Jintao Wang，Jun Wang，et al.，“Channel Estimation for theChinese DTTB System Based on a Novel Iterative PN SequenceReconstruction，″in Proc.IEEE International Conference onCommunications(ICC′08)，May.2008，pp.285-289)等文献中记载的方法在实现干扰消除和信道估计时所需要的计算复杂度对比。表中的J表示迭代次数。In order to analyze the complexity of the uplink multi-user time domain synchronous frequency division multiple access method proposed by the present invention, table 1 has provided the multiple access method and the direct iteration method of the present invention (referring to document [1] J.Wang , Z.Yang, C.Pan, J.Song, and L.Yang, "Iterative padding substruction of the PN sequence for the TDS-OFDM overbroadcasting channels," IEEE Trans.Consumer Electron., vol.51, no.4, pp .1148-1152, Nov.2005), an iterative interference cancellation method based on partial decision feedback (see literature [2] Shigang Tang, Kewu Peng, Ke Gong, et al., "NovelDecision-Aided Channel Estimation for TDS-OFDM Systems, ″in Proc.IEEE International Conference on Communications (ICC′08), May.2008, vol.1, pp.946-950), methods based on training sequence reconstruction (see literature [3] FangYang, Jintao Wang, Jun Wang , et al., "Channel Estimation for the Chinese DTTB System Based on a Novel Iterative PN Sequence Reconstruction," in Proc.IEEE International Conference on Communications (ICC′08), May.2008, pp.285-289) etc. Comparing the computational complexity required to implement interference cancellation and channel estimation. J in the table indicates the number of iterations.

表1Table 1

运算operation 文献[1]Literature [1] 文献[2]Literature [2] 文献[3]Literature [3] 本发明 this invention 256点IFFT/FFT256 points IFFT/FFT 00 00 00 33 2048点IFFT/FFT2048 points IFFT/FFT 4(J+1)4(J+1) 2(J+1)2(J+1) 3(J+1)3(J+1) 00 3780点IFFT/FFT3780 points IFFT/FFT 2 2 1 1 2 2 1 1 4200点IFFT/FFT4200 points IFFT/FFT 00 5(J+1)5(J+1) 00 00 8192点IFFT/FFT8192 points IFFT/FFT 3(J+1)3(J+1) 00 00 00

从表1中可以看出，当迭代次数J＝1时，基于部分判决反馈的迭代干扰消除方法的复杂度是直接迭代方法的68％，基于训练序列重构的方法的复杂度是直接迭代法的24％，而本发明基于新的超帧结构，由于不需要迭代，而且联合循环重构和联合信道估计的方法都非常简单，因此其复杂度仅为直接迭代法的6％。当迭代次数J增大时，本发明所述方法的相对复杂度则更低。It can be seen from Table 1 that when the number of iterations J=1, the complexity of the iterative interference elimination method based on partial decision feedback is 68% of that of the direct iterative method, and the complexity of the method based on training sequence reconstruction is 68% of that of the direct iterative method 24% of , and the present invention is based on a new superframe structure, since no iteration is required, and the methods of combined loop reconstruction and joint channel estimation are very simple, so its complexity is only 6% of that of the direct iterative method. When the number of iterations J increases, the relative complexity of the method of the present invention is lower.

基于上述描述及具体实施方式，对本发明所提出的上行多用户时域同步频分多址接入方法，以系统同时支持4个用户(M＝4)为例，系统的主要参数如表2所示，本实施例对该系统的可行性和性能进行了计算机仿真，仿真中所用的信道为表3所示的4种典型无线多径信道Brazil A和Brazil D(参见文献：“Digital Television Systems-BrazilianTests-Final Report，”ANATEL SP，May 2000)，以及ITU推荐的信道模型Indoor B and Vehicular A(参见文献：Recommendation ITU-RM.1225，“Guideline for Evaluation of Radio Transmission Technologyfor IMT-2000，”1997)。上行多用户系统中的用户1、用户2、用户3和用户4分别经过多径信道Brazil A、Indoor B、Vehicular A和Brazil D。Based on the above description and specific implementation, for the uplink multi-user time-domain synchronous frequency division multiple access method proposed by the present invention, taking the system supporting 4 users (M=4) at the same time as an example, the main parameters of the system are shown in Table 2 Shown, this embodiment has carried out computer simulation to the feasibility and performance of this system, and the channel used in the simulation is 4 kinds of typical wireless multi-path channels Brazil A and Brazil D shown in table 3 (referring to literature: "Digital Television Systems- BrazilianTests-Final Report," ANATEL SP, May 2000), and the channel model Indoor B and Vehicular A recommended by ITU (see literature: Recommendation ITU-RM.1225, "Guideline for Evaluation of Radio Transmission Technology for IMT-2000," 1997) .User 1,User 2,User 3, andUser 4 in the uplink multi-user system pass through multipath channels Brazil A, Indoor B, Vehicular A, and Brazil D respectively.

表2Table 2

用户数目MNumber of users M 44 系统带宽system bandwidth 8MHz8MHz 可用OFDM子载波数总数NThe total number of available OFDM subcarriers N 37803780 子载波间隔subcarrier spacing 2kHz2kHz 子载波分配方案Subcarrier Allocation Scheme 交织分配interleaving allocation 每个用户分配到的子载波数The number of subcarriers allocated to each user 945945 调制方式 Modulation QPSKQPSK 符号率symbol rate 7.56MHz7.56MHz m序列长度N_pm sequence length N_p 255255 超帧中的子帧个数LThe number of subframes L in asuperframe 44 前导序列中前保护间隔、后保护间隔长度KThe length of the pre-guard interval and post-guard interval in the preamble sequence K 6464

表3table 3

本实施例给出了本发明提出的上行多用户时域同步频分多址接入系统中的多用户信道估计方法与传统TDS-OFDM系统中迭代信道估计方法(参考文献[1])在均方误差(MSE)方面的性能对比仿真结果。在图5给出仿真结果中，Iter表示迭代次数。文献[1]已指出，当Iter＝2时，信道估计的MSE可基本达到迭代法所能取得的性能上界，但图8中的仿真结果表明，此时迭代法信道估计的性能仍然没有本发明中利用前导序列进行多用户联合信道估计的方法性能好。这是因为在传统的TDS-OFDM系统中，IDFT数据块对PN的干扰将会影响迭代信道估计算法的性能，虽然可以通过迭代的干扰消除来减小法来这钟干扰，但是干扰不能完全消除。相反，在本发明中用于信道估计的m序列由于在帧结构设计中保护了前保护间隔，因此不存在IDFT数据块对m序列的干扰，因此即使信道估计是针对上行多用户系统中的所有用户同时进行的，也可以得到更精确的信道估计结果。This embodiment gives the multi-user channel estimation method in the uplink multi-user time domain synchronous frequency division multiple access system proposed by the present invention and the iterative channel estimation method in the traditional TDS-OFDM system (Reference [1]) The performance in terms of square error (MSE) is compared with the simulation results. In the simulation results shown in Figure 5, Iter represents the number of iterations. Literature [1] has pointed out that when Iter=2, the MSE of channel estimation can basically reach the upper bound of the performance obtained by the iterative method, but the simulation results in Figure 8 show that the performance of the iterative channel estimation is still not as good as this In the invention, the method for performing multi-user joint channel estimation by using the preamble sequence has good performance. This is because in the traditional TDS-OFDM system, the interference of the IDFT data block to the PN will affect the performance of the iterative channel estimation algorithm. Although the interference can be reduced by iterative interference elimination, the interference cannot be completely eliminated. . On the contrary, the m-sequence used for channel estimation in the present invention protects the front guard interval in the frame structure design, so there is no interference of the IDFT data block to the m-sequence, so even if the channel estimation is for all uplink multi-user systems At the same time, the user can also obtain more accurate channel estimation results.

图6给出了提出的上行多用户时域同步频分多址接入系统与传统的TDS-OFDM单用户系统在四种典型多径信道中的误比特率(BER)性能对比结果。仿真采用的子载波分配方式为交织分配，多用户系统中的4个用户分别占用可用子载波总数的1/4，传统的TDS-OFDM单用户系统中只有一个用户并占用所有可用子载波。仿真结果表明，本发明的上行多用户系统中每个用户所能达到的BER性能与传统TDS-OFDM单用户系统中一个用户所能达到的BER性能非常接近，且略好于传统的单用户系统。这是因为，一方面，基于本发明前导序列中m序列的信道估计可以获得更精确的信道估计结果；另一方面，与传统TDS-OFDM单用户系统不同的是，本发明在重构IDFT数据块的循环特性时不需要任何信道信息(CSI)，因而可以获得更精确的重构结果。Figure 6 shows the bit error rate (BER) performance comparison between the proposed uplink multi-user TDSFA system and the traditional TDS-OFDM single-user system in four typical multipath channels. The subcarrier allocation method used in the simulation is interleaved allocation. In the multi-user system, four users occupy 1/4 of the total number of available subcarriers respectively. In the traditional TDS-OFDM single-user system, there is only one user and occupies all available subcarriers. Simulation results show that the BER performance that each user can achieve in the uplink multi-user system of the present invention is very close to the BER performance that one user can achieve in the traditional TDS-OFDM single-user system, and slightly better than the traditional single-user system . This is because, on the one hand, more accurate channel estimation results can be obtained based on the channel estimation of the m-sequence in the preamble sequence of the present invention; on the other hand, unlike the traditional TDS-OFDM single-user system, the present invention can reconstruct The cyclic properties of the blocks do not require any channel information (CSI), thus obtaining more accurate reconstruction results.

图7给出了本发明提出的上行多用户TDS-FDMA系统和传统的TDS-OFDM系统在瑞利衰落信道下的BER性能对比，采用的信道模型为VehicularA，信道的最大多普勒扩展分别为10Hz、30Hz和100Hz，分别对应终端用户的低速、中速和高速移动场景。从仿真结果中可以看出，对于BER为5×10^-3，当最大多普勒扩展分别为10Hz时，传统单用户TDS-OFDM系统中所需的SNR约为25dB，而多用户TDS-FDMA系统所需的SNR则约为22dB，因此SNR提升了约3dB；当最大多普勒扩展分别为30Hz时，SNR提升了约4dB；当最大多普勒扩展分别为100Hz时，SNR的提升空间大幅减小，二者的BER性能非常接近。可以看出，在低速特别是中速时变信道下本发明提出的上行多用户TDS-FDMA系统可以比传统单用户TDS-OFDM系统获得更好的BER性能，这是因为，本发明帧结构在多用户IDFT数据的联合循环特性重构过程中不需要任何信道信息，避免了PN序列与数据部分的迭代干扰消除，而传统的单用户TDS-OFDM系统在循环特性的迭代重构过程中需要不断利用信道信息以逐步消除PN序列与数据部分的干扰，在时变条件下，通过信道估计得到的信道信息存在一定的误差，而这个误差在迭代过程中可能不断累加，从而导致了系统性能的恶化；在高速移动情况下，由于信道的相干时间缩短，而本发明的信道估计是通过前导序列获得的，因此更新速度相对减慢，因此由于快速时变引起的信道估计误差就变大，从而损失了循环重构不需要信道信息带来的增益。这个问题可以通过动态调整帧结构中信号子帧的个数L来解决，然而较小的L就意味这较大的传输有效性损失，不过这是在高速移动条件下为了获得传输的可靠性需要付出的代价。Fig. 7 has provided the BER performance comparison of the uplink multi-user TDS-FDMA system proposed by the present invention and the traditional TDS-OFDM system under the Rayleigh fading channel, the channel model adopted is VehicularA, and the maximum Doppler spread of the channel is respectively 10Hz, 30Hz and 100Hz correspond to low-speed, medium-speed and high-speed mobile scenarios of end users respectively. It can be seen from the simulation results that for a BER of 5×10^-3 , when the maximum Doppler spread is 10 Hz respectively, the required SNR in the traditional single-user TDS-OFDM system is about 25dB, while the multi-user TDS-FDMA The SNR required by the system is about 22dB, so the SNR is increased by about 3dB; when the maximum Doppler spread is 30Hz, the SNR is increased by about 4dB; when the maximum Doppler spread is 100Hz, the SNR can be greatly improved The BER performance of the two is very close. It can be seen that the uplink multi-user TDS-FDMA system proposed by the present invention can obtain better BER performance than the traditional single-user TDS-OFDM system under low-speed, especially medium-speed time-varying channels, because the frame structure of the present invention is in The combined cycle characteristic reconstruction process of multi-user IDFT data does not require any channel information, which avoids the iterative interference cancellation of the PN sequence and data part, while the traditional single-user TDS-OFDM system needs to continuously Use channel information to gradually eliminate the interference between the PN sequence and the data part. Under time-varying conditions, there is a certain error in the channel information obtained through channel estimation, and this error may continue to accumulate during the iterative process, resulting in the deterioration of system performance. ; In the case of high-speed mobile, because the coherence time of the channel is shortened, and the channel estimation of the present invention is obtained through the preamble sequence, the update speed is relatively slowed down, so the channel estimation error caused by the fast time-varying becomes larger, thereby losing The gain brought by the cyclic reconstruction does not require channel information. This problem can be solved by dynamically adjusting the number L of signal subframes in the frame structure. However, a smaller L means a greater loss of transmission effectiveness, but this is required in order to obtain transmission reliability under high-speed mobile conditions. the price paid.

上述理论分析和仿真结果都表明，本发明的上行多用户时域同步频分多址接入方法通过帧结构中IDFT数据块的后保护间隔和前导序列的后保护间隔之间一次简单的加减运算便可完成所有用户接收DFT数据块的联合循环特性重构，通过本地m序列与前导序列中m序列的一次循环相关便可完成所有用户的信道估计，整个方案简单可行，而且以比传统的单用户TDS-OFDM系统更低的复杂度实现了多用户接入，并且在低速和中速移动条件下取得了更好的系统性能。同时，本发明还为多载波OFDMA信号和单载波SC-FDMA信号提供了一种统一而灵活的上行多址帧结构。The above theoretical analysis and simulation results all show that the uplink multi-user time-domain synchronous frequency division multiple access method of the present invention uses a simple addition and subtraction between the post-guard interval of the IDFT data block in the frame structure and the post-guard interval of the preamble sequence The combined cyclic characteristic reconstruction of all users receiving DFT data blocks can be completed through the operation, and the channel estimation of all users can be completed by a cyclic correlation between the local m sequence and the m sequence in the preamble sequence. The whole scheme is simple and feasible, and it is more efficient than the traditional The lower complexity of the single-user TDS-OFDM system enables multi-user access, and achieves better system performance under low-speed and medium-speed mobile conditions. At the same time, the present invention also provides a unified and flexible uplink multiple access frame structure for multi-carrier OFDMA signals and single-carrier SC-FDMA signals.

以上所述仅是本发明的优选实施方式，应当指出，对于本技术领域的普通技术人员来说，在不脱离本发明技术原理的前提下，还可以做出若干改进和变型，这些改进和变型也应视为本发明的保护范围。The above is only a preferred embodiment of the present invention, it should be pointed out that for those of ordinary skill in the art, without departing from the technical principle of the present invention, some improvements and modifications can also be made, these improvements and modifications It should also be regarded as the protection scope of the present invention.

Claims

Translated fromChinese

1.一种上行多用户时域同步频分多址接入方法，其特征在于，包括以下步骤：1. an uplink multi-user time-domain synchronous frequency division multiple access method, is characterized in that, comprises the following steps:

组帧步骤，在发送端以超帧为基本单元组成信号帧，组成所述超帧的方式为：在L个信号子帧之前插入前导序列；其中，所述信号子帧包括时域数据块和后保护间隔，所述前导序列包括训练序列、前保护间隔和所述后保护间隔，所述前保护间隔与后保护间隔相同；所述前保护间隔和后保护间隔均为所述训练序列中的最后K个符号，且K大于或等于无线信道的最大时延扩展；In the framing step, the superframe is used as a basic unit to form a signal frame at the sending end, and the superframe is formed in the following manner: a preamble sequence is inserted before L signal subframes; wherein, the signal subframe includes time domain data blocks and After the guard interval, the leading sequence includes a training sequence, a front guard interval and the rear guard interval, the former guard interval is the same as the rear guard interval; the former guard interval and the rear guard interval are both in the training sequence The last K symbols, and K is greater than or equal to the maximum delay spread of the wireless channel;

2.如权利要求1所述的上行多用户时域同步频分多址接入方法，其特征在于，根据无线信道的相干时间大小或者无线信道的变化速度确定所述信号子帧的个数L。2. the uplink multi-user time domain synchronous frequency division multiple access method as claimed in claim 1, is characterized in that, according to the coherence time size of wireless channel or the rate of change of wireless channel, determine the number L of said signal subframes .

3.如权利要求2所述的上行多用户时域同步频分多址接入方法，其特征在于，所述无线信道的多普勒扩展带宽越大，则所述信号子帧的个数L的取值越小；所述无线信道的多普勒扩展带宽越小，则所述信号子帧的个数L的取值越大。3. the uplink multi-user time domain synchronous frequency division multiple access method as claimed in claim 2, is characterized in that, the Doppler spread bandwidth of described wireless channel is bigger, then the number L of described signal subframes The smaller the value of ; the smaller the Doppler spread bandwidth of the wireless channel, the larger the value of the number L of the signal subframes.

4.如权利要求1所述的上行多用户时域同步频分多址接入方法，其特征在于，所述训练序列是单载波形式的时域m序列，其中，相邻两个用户m、m+1所采用的时域序列为p_m和p_m+1，p_m+1是p_m经过L_s位循环位移后得到的序列。4. the uplink multi-user time domain synchronous frequency division multiple access method as claimed in claim 1, is characterized in that, described training sequence is the time domain m sequence of single carrier form, wherein, adjacent two users m, The time-domain sequences used by m+1 are_pm and pm₊₁ , and pm₊₁ is a sequence obtained after cyclic shifting of_pm by L_s bits.

5.如权利要求1所述的上行多用户时域同步频分多址接入方法，其特征在于，所述训练序列是多载波形式的频域序列，不同用户的训练序列占用频域上相互正交的子载波，然后通过逆傅立叶变换得到其对应的时域中的训练序列。5. the uplink multi-user time domain synchronous frequency division multiple access method as claimed in claim 1, is characterized in that, described training sequence is the frequency domain sequence of multi-carrier form, and the training sequence of different users occupies mutually on the frequency domain Orthogonal subcarriers are then inverse Fourier transformed to obtain their corresponding training sequences in the time domain.

6.如权利要求1～5之任一项所述的上行多用户时域同步频分多址接入方法，其特征在于，所述时域数据块为IDFT时域数据块。6. The uplink multi-user Time Domain Synchronous Frequency Division Multiple Access method according to any one of claims 1 to 5, wherein the time domain data block is an IDFT time domain data block.

7.如权利要求6所述的上行多用户时域同步频分多址接入方法，其特征在于，所述IDFT时域数据块为正交频分复用多址形式的多载波信号，或者单载波频分多址形式的单载波信号。7. the uplink multi-user time domain synchronous frequency division multiple access method as claimed in claim 6, is characterized in that, described IDFT time domain data block is the multi-carrier signal of OFDM multiple access form, or A single carrier signal in the form of single carrier frequency division multiple access.

8.如权利要求5所述的上行多用户时域同步频分多址接入方法，其特征在于，若所述训练序列是多载波形式的频域序列，不同用户均按照一一对应的方式分别占用所述训练序列和IDFT时域数据块中对应的相互正交的子载波。8. The uplink multi-user time-domain synchronous frequency-division multiple access method as claimed in claim 5, wherein if the training sequence is a frequency-domain sequence in the form of a multi-carrier, different users are in a one-to-one correspondence manner Respectively occupy the training sequence and the corresponding mutually orthogonal subcarriers in the IDFT time domain data block.