CN101005477A

CN101005477A - Synchronous signal sending and detecting method for orthogonal frequency division complex system

Info

Publication number: CN101005477A
Application number: CN 200710000203
Authority: CN
Inventors: 夏树强; 梁春丽
Original assignee: ZTE Corp
Current assignee: ZTE Corp
Priority date: 2007-01-04
Filing date: 2007-01-04
Publication date: 2007-07-25
Anticipated expiration: 2027-01-04
Also published as: CN100589473C

Abstract

The invention relates to a synchronous signal transmission and detection method of an OFDM system, wherein the transmission method makes the circular convolution peak value of the circular convolution of the coarse synchronous synchronous signal of the receiving end and the local reference signal be located at Different locations; wherein the detection method utilizes the above-mentioned location characteristics to detect the cyclic prefix type while performing synchronous timing. In this method, during the process of obtaining the timing of the synchronization signal, the mobile station simultaneously obtains the cyclic prefix length type of the OFDM symbol where the synchronization signal is located. No additional steps are required to detect the cyclic prefix type. Since the step of obtaining the timing of the synchronization signal by the mobile station is necessary, the mobile station of the present invention can obtain the cyclic prefix type of the synchronization signal without increasing overhead; in addition, compared to the blind detection method for the cyclic prefix type, the present invention provides The method also has the advantages of easy implementation and high detection accuracy.

Description

A Synchronous Signal Transmission and Detection Method for Orthogonal Frequency Division Multiplexing System

技术领域technical field

本发明涉及正交频分复用系统，具体地涉及到带有多种循环前缀的基于正交频分复用系统的同步信号发送与检测方法。The invention relates to an orthogonal frequency division multiplexing system, in particular to a method for sending and detecting synchronous signals based on the orthogonal frequency division multiplexing system with multiple cyclic prefixes.

背景技术Background technique

在正交频分复用系统中，每个OFDM(正交频分复用)符号都包括两部分：循环前缀部分和数据部分。循环前缀部分是数据部分后面部分的拷贝，它具有抵制符号间干扰与载波间干扰的双重作用。一般而言，循环前缀部分数据越长，其抵制符号间干扰与载波间干扰的能力也就越强。但是也有负面影响，循环前缀部分数据越长，系统的开销也就越大。在实际应用中，为了降低循环前缀的这种负面影响，系统通常支持多种循环前缀长度。系统可以根据环境，业务类型的差异灵活选择循环前缀的长度，比如在ETURA系统中，就支持两种循环前缀长度类型以适应不同的需求。In an Orthogonal Frequency Division Multiplexing system, each OFDM (Orthogonal Frequency Division Multiplexing) symbol includes two parts: a cyclic prefix part and a data part. The cyclic prefix part is a copy of the part behind the data part, and it has dual functions of resisting inter-symbol interference and inter-carrier interference. Generally speaking, the longer the data of the cyclic prefix part is, the stronger its ability to resist inter-symbol interference and inter-carrier interference is. However, there are also negative effects. The longer the data in the cyclic prefix part, the greater the overhead of the system. In practical applications, in order to reduce the negative impact of the cyclic prefix, the system usually supports multiple cyclic prefix lengths. The system can flexibly select the length of the cyclic prefix according to the environment and the difference in business type. For example, in the ETURA system, it supports two types of cyclic prefix lengths to meet different requirements.

但是，循环前缀长度的灵活选择给移动台与系统同步带来了不利影响。。因为移动台在同步过程中还需要额外知道同步信号使用的循环前缀长度类型。通常，获取同步信号使用的循环前缀长度类型有两种方式：一种称为盲检测方法，这种方法的基本思路是：移动台尝试所有可能的循环前缀长度类型，根据每种循环前缀长度类型的计算量度结果，确定最为合适的循环前缀长度类型。另外一种方式是规定同步信号只在某一种特定循环前缀长度类型的OFDM符号上发送。这种方法需要标准的支持，并且，同步信号通常只是是在OFDM符号的部分载波上发送，这种强制规定给同步信号与其它信号的复用带来麻烦。However, the flexible selection of the cyclic prefix length has a negative impact on the synchronization between the mobile station and the system. . Because the mobile station also needs to additionally know the type of cyclic prefix length used by the synchronization signal during the synchronization process. Generally, there are two ways to obtain the cyclic prefix length type used by the synchronization signal: one is called the blind detection method, the basic idea of this method is: the mobile station tries all possible cyclic prefix length types, and according to each cyclic prefix length type Determine the most appropriate cyclic prefix length type. Another way is to specify that the synchronization signal is only sent on OFDM symbols of a specific cyclic prefix length type. This method needs the support of the standard, and the synchronization signal is usually only sent on some carriers of the OFDM symbol. This mandatory regulation brings troubles to the multiplexing of the synchronization signal and other signals.

发明内容Contents of the invention

本发明需要解决的技术问题是提供一种正交频分复用系统的同步信号发送与检测方法，在获取同步定时的同时进行循环前缀类型检测，避免额外的循环前缀类型检测开销。The technical problem to be solved by the present invention is to provide a synchronization signal transmission and detection method for an OFDM system, which detects the cyclic prefix type while obtaining synchronization timing, and avoids additional cyclic prefix type detection overhead.

本发明的上述技术问题这样解决，提供一种正交频分复用系统的同步信号发送与检测方法，接收端同步同时进行循环前缀类型检测，包括以下步骤：The above-mentioned technical problem of the present invention is solved in this way, a kind of synchronous signal transmission and detection method of OFDM system is provided, and the receiving end synchronously carries out cyclic prefix type detection simultaneously, comprises the following steps:

1.1)发送端对同步信号的数据部分按循环前缀类型进行循环移位使其在接收端与本地参考信号进行圆周卷积的峰值位于不同位置；1.1) The sending end cyclically shifts the data part of the synchronization signal according to the cyclic prefix type so that the peak value of the circular convolution with the local reference signal at the receiving end is located at a different position;

1.2)接收端在进行同步定时过程中根据所述不同位置区分还原出所述循环前缀类型。1.2) The receiving end distinguishes and restores the cyclic prefix type according to the different positions during the synchronization timing process.

按照本发明提供的发送与检测方法，所述同步定时过程使用本地参考同步信号x^*(0)，x^*(N-1)，x^*(N-2)...x^*(1)，包括以下具体步骤：According to the transmission and detection method provided by the present invention, the synchronization timing process uses local reference synchronization signals x ^* (0), x ^* (N-1), x ^* (N-2)...x ^* (1), Include the following specific steps:

2.1)利用标准同步信号时域重复特性确定同步信号的粗定时数据r(0)...r(N-1)，r(N)...r(2N-1)；2.1) Determine the rough timing data r(0)...r(N-1), r(N)...r(2N-1) of the synchronization signal by using the time-domain repetition characteristic of the standard synchronization signal;

2.2)计算量度c1：c1是r(0)....r(N-1)与本地参考同步信号x^*(0)，x^*(N-1)，x^*(N-2)...x^*(1)的圆周卷积；2.2) Compute the measure c1: c1 is r(0)..r(N-1) with local reference sync signal x ^* (0), x ^* (N-1), x ^* (N-2).. Circumferential convolution of .x ^* (1);

2.3)计算量度c，c＝|c1|²，|x|表示x的模，x为任意表达式；2.3) Calculate the measurement c, c=|c1| ² , |x| represents the modulus of x, and x is an arbitrary expression;

2.4)计算量度c中最大元素对应的索引k，根据k的数值范围和所述圆周卷积峰值规律衍变的对应规则决定同步信号的细定时位置与循环前缀类型。2.4) Calculate the index k corresponding to the largest element in the measure c, and determine the fine timing position and cyclic prefix type of the synchronization signal according to the value range of k and the corresponding rule of the circular convolution peak law evolution.

按照本发明提供的发送与检测方法，所述步骤2.2)还包括计算量度c2：c2是r(N)....r(2N-1)与本地参考同步信号x^*(0)，x^*(N-1)，x^*(N-2)...x^*(1)的圆周卷积；所述步骤1.3)和1.4)中量度c＝|c1|²+|c2|²。According to the transmission and detection method provided by the present invention, the step 2.2) also includes calculating the measurement c2: c2 is r(N)....r(2N-1) and the local reference synchronization signal x ^* (0), x ^* Circular convolution of (N-1), x ^* (N-2)...x ^* (1); measure c=|c1| ²⁺ |c2| ² in steps 1.3) and 1.4) described.

按照本发明提供的发送与检测方法，所述步骤2.1)包括：According to the sending and detection method provided by the present invention, the step 2.1) includes:

2.1.1)计算量度M(k)： $M (k) = {| Σ_{p = 0}^{N - 1} r^{*} (n + k + p) r (n + k + p + N) |}^{2},$ k＝0，1....2N-1；2.1.1) Calculate the metric M(k): $m (k) = {| Σ_{p = 0}^{N - 1} r^{*} (no + k + p) r (no + k + p + N) |}^{2},$ k=0,1...2N-1;

2.1.2)计算获得M(k)中最大值对应的索引，设为m，则r(n+m)就是同步信号的粗定时位置；2.1.2) Calculate the index corresponding to the maximum value in M(k), set it as m, then r(n+m) is the rough timing position of the synchronization signal;

2.1.3)根据粗定时位置获得同步信号的粗定时数据：r(0)...r(N-1)，r(N)...r(2N-1)。2.1.3) Obtain the rough timing data of the synchronization signal according to the rough timing position: r(0)...r(N-1), r(N)...r(2N-1).

按照本发明提供的发送与检测方法，所述圆周卷积是：According to the sending and detecting method provided by the present invention, the circular convolution is:

$c c 11 = = (\begin{matrix} c c 11 ((00)) \\ c c 11 ((11)) \\ c c 11 ((22)) \\ . . . . . . \\ c c 11 ((N N = = 11)) \end{matrix}) = = (\begin{matrix} {x x}^{* *} ((00)),, & {x x}^{* *} ((11)),, & {x x}^{* *} ((22)) & . . . . . . & {x x}^{* *} ((N N - - 11)) \\ {x x}^{* *} ((N N - - 11)),, & {x x}^{* *} ((00)),, & {x x}^{* *} ((11)) & . . . . . . & {x x}^{* *} ((N N - - 22)) \\ {x x}^{* *} ((N N - - 22)),, & {x x}^{* *} ((N N - - 11)),, & {x x}^{* *} ((00)) & . . . . . . & {x x}^{* *} ((N N - - 33)) \\ . . . . . . \\ {x x}^{* *} ((11)),, & {x x}^{* *} ((22)),, & {x x}^{* *} ((33)) & . . . . . . & {x x}^{* *} ((00)) \end{matrix}) \times \times (\begin{matrix} \begin{matrix} r r ((00)) \\ r r ((11)) \\ r r ((22)) \\ . . . . . . \\ r r ((N N - - 11)) \end{matrix} \end{matrix})$

按照本发明提供的发送与检测方法，所述循环前缀类型是两种，循环前缀长度分别是L1和L2，对应所述循环移位分别是0和左移m+1位，对应所述数据部分分别是理想同步信号x(0)，x(1)....x(2N-1)和左移同步信号x(m+1)....x(2N-1)，x(0)....x(m)；m＝L2+D，0≤D≤N-L1-L2-2，2N是所述数据部分的长度。According to the transmission and detection method provided by the present invention, there are two types of cyclic prefixes, the lengths of cyclic prefixes are L1 and L2 respectively, and the corresponding cyclic shifts are 0 and left-shifted by m+1 bits respectively, corresponding to the data part They are ideal synchronous signal x(0), x(1)....x(2N-1) and left-shifted synchronous signal x(m+1)....x(2N-1), x(0) ....x(m); m=L2+D, 0≤D≤N-L1-L2-2, 2N is the length of the data part.

按照本发明提供的发送与检测方法，D取最优值：According to the transmission and detection method provided by the present invention, D takes the optimal value:

表示对x向上取整；

Indicates that x is rounded up;

_x_表示对x向下取整；_x_ indicates that x is rounded down;

按照本发明提供的发送与检测方法，所述循环前缀类型是两种，对应所述循环移位分别是0和左移m+1位时，所述对应规则是：According to the sending and detecting method provided by the present invention, there are two types of cyclic prefixes, and when the corresponding cyclic shifts are respectively 0 and left shifted by m+1 bits, the corresponding rules are:

k k 判断规则 judgment rule 0≤k≤L1 0≤k≤L1 循环前缀类型1，并且接收信号r(0)相对于同步信号的理想定时位置超前k个样点。 Cyclic prefix type 1, and the received signal r(0) is ahead of the ideal timing position of the synchronization signal by k samples. L1+1≤k≤N-L2-D-2 L1+1≤k≤N-L2-D-2 循环前缀类型2，并且接收信号r(0)相对于同步信号的理想定时位置延后(N-L2-D-1-k)个样点。 Cyclic prefix type 2, and the received signal r(0) is delayed by (N-L2-D-1-k) samples relative to the ideal timing position of the synchronization signal. N-L2-D-1≤k≤N-D-1 N-L2-D-1≤k≤N-D-1 循环前缀类型2，并且接收信号r(0)相对于同步信号的理想定时位置超前(k+L2+D-N)个样点。 Cyclic prefix type 2, and the received signal r(0) is ahead of the ideal timing position of the synchronization signal by (k+L2+D-N) samples. N-D≤k≤N-1 N-D≤k≤N-1 循环前缀类型1，并且接收信号r(0)相对于同步 Cyclic prefix type 1, and receive signal r(0) relative to sync

信号的理想定时位置延后(N-k-1)个样点。 The ideal timing position of the signal is delayed by (N-k-1) samples.

按照本发明提供的发送与检测方法，所述循环前缀类型可以是三种，循环前缀长度分别是L1、L2和L3，对应所述循环移位分别是0、左移m+1位和左移p+1位，m＝N-L1-D-1，p＝N-L1-L2-2D-2；According to the transmission and detection method provided by the present invention, the cyclic prefix types can be three types, the cyclic prefix lengths are L1, L2 and L3 respectively, and the corresponding cyclic shifts are 0, left shift by m+1 bits and left shift respectively p+1 bit, m=N-L1-D-1, p=N-L1-L2-2D-2;

表示对x向上取整；

Indicates that x is rounded up;

_x_表示对x向下取整。_x_ indicates that x is rounded down.

按照本发明提供的发送与检测方法，所述对应规则是：According to the sending and detection method provided by the present invention, the corresponding rules are:

k k 判断规则 judgment rule 0≤k≤L1 0≤k≤L1 循环前缀类型1，并且接收信号r(0)相对于同步信号的理想定时位置超前k个样点。 Cyclic prefix type 1, and the received signal r(0) is ahead of the ideal timing position of the synchronization signal by k samples. L1+1≤k≤L1+D L1+1≤k≤L1+D 循环前缀类型2，并且接收信号r(0)相对于同步信号的理想定时位置延后(L1+D+1-k)个样点。 Cyclic prefix type 2, and the received signal r(0) is delayed by (L1+D+1-k) samples relative to the ideal timing position of the synchronization signal. L1+D+1≤k≤L1+D+1+L2 L1+D+1≤k≤L1+D+1+L2 循环前缀类型2，并且接收信号r(0)相对于同步信号的理想定时位置超前(k-L1-D-1)个样点。 Cyclic prefix type 2, and the received signal r(0) is ahead of the ideal timing position of the synchronization signal by (k-L1-D-1) samples. L1+L2+D+2≤k≤L1++L2+2D+1 L1+L2+D+2≤k≤L1++L2+2D+1 循环前缀类型3，并且接收信号r(0)相对于同步信号的理想定时位置延后(L1+L2+2D+2-k)个样点。 Cyclic prefix type 3, and the received signal r(0) is delayed by (L1+L2+2D+2-k) samples relative to the ideal timing position of the synchronization signal. L1++L2+L3+2D+3≤k≤L1+L2+L3+2D+2 L1++L2+L3+2D+3≤k≤L1+L2+L3+2D+2 循环前缀类型3，并且接收信号r(0)相对于同步信号的理想定时位置超前(k-L1-L2-2D-2)个样点 Cyclic prefix type 3, and the received signal r(0) is (k-L1-L2-2D-2) samples ahead of the ideal timing position of the sync signal L1+L2+2D+3≤k≤L1+L2+L3+3D+2 L1+L2+2D+3≤k≤L1+L2+L3+3D+2 循环前缀类型1，并且接收信号r(0) Cyclic prefix type 1, and receive signal r(0)

相对于同步信号的理想定时位置延后(L1+L2+L3+3D+3-k)个样点 Delayed by (L1+L2+L3+3D+3-k) samples relative to the ideal timing position of the sync signal

按照本发明提供的发送与检测方法，所述发送端是基站，所述接收端是移动台。According to the sending and detecting method provided by the present invention, the sending end is a base station, and the receiving end is a mobile station.

本发明提供的正交频分复用系统的同步信号发送与检测方法，移动台获得同步信号定时的过程中，同时获得了同步信号所在OFDM符号的循环前缀长度类型。无须额外的步骤对循环前缀类型进行检测。由于，移动台获得同步信号定时的步骤是必须作的，因此本发明移动台不用增加开销就可获得同步信号的循环前缀类型；另外，相比对循环前缀类型进行盲检测的方式，本发明提供的方法还具有容易实现，检测准确率高等优点。In the synchronous signal transmission and detection method of the orthogonal frequency division multiplexing system provided by the present invention, in the process of obtaining the timing of the synchronous signal, the mobile station simultaneously obtains the cyclic prefix length type of the OFDM symbol where the synchronous signal is located. No additional steps are required to detect the cyclic prefix type. Because the step of obtaining the timing of the synchronization signal by the mobile station is necessary, the mobile station of the present invention can obtain the cyclic prefix type of the synchronization signal without increasing overhead; in addition, compared to the blind detection method for the cyclic prefix type, the present invention provides The method also has the advantages of easy implementation and high detection accuracy.

附图说明Description of drawings

下面结合附图和具体实施例进一步对本发明进行详细说明。The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments.

图1为OFDM符号示意图；Fig. 1 is a schematic diagram of OFDM symbols;

图2具有第一种循环前缀的OFDM符号示意图；Fig. 2 has a schematic diagram of OFDM symbols of the first type of cyclic prefix;

图3具有第二种循环前缀的OFDM符号示意图；Fig. 3 has the schematic diagram of the OFDM symbol of the second kind of cyclic prefix;

图4包含对应图2的本发明同步信号的生成示意图；Fig. 4 includes a schematic diagram of the generation of the synchronous signal of the present invention corresponding to Fig. 2;

图5包含对应图3的本发明同步信号的生成示意图；Fig. 5 includes a schematic diagram of generating a synchronous signal of the present invention corresponding to Fig. 3;

图6包含本发明的同步信号检测流程图。Fig. 6 contains a flow chart of synchronization signal detection of the present invention.

具体实施方式Detailed ways

首先，以两种循环前缀类型为例说明本发明基于正交频分复用技术的同步信号发送与检测方法关键步骤：First, two types of cyclic prefixes are taken as examples to illustrate the key steps of the synchronization signal transmission and detection method based on OFDM technology in the present invention:

(一)发送：在该系统中，设FFT的大小为2N，有两种循环前缀，分别为循环前缀类型1和类型2，循环前缀中包括的样点数目分别为：L1，L2。不失一般性，对于循环前缀类型1，设系统发送的同步信号在一个OFDM符号中发送，其数据部分为：x(0)，x(1)....x(2N-1).并且满足如下条件：x(n)＝x(n+N)，n＝0，1...N-1.即，该信号在时域是重复的。设x(0)，x(1)...x(N)与其循环移位的互相关R(u)为：(1) Sending: In this system, the size of the FFT is assumed to be 2N, and there are two types of cyclic prefixes, respectively cyclic prefix type 1 and type 2, and the number of samples included in the cyclic prefix are respectively: L1, L2. Without loss of generality, for cyclic prefix type 1, it is assumed that the synchronization signal sent by the system is sent in one OFDM symbol, and its data part is: x(0), x(1)....x(2N-1). And The following conditions are satisfied: x(n)=x(n+N), n=0, 1...N-1. That is, the signal is repeated in the time domain. Let the cross-correlation R(u) between x(0), x(1)...x(N) and its cyclic shift be:

$R R ((u u)) = = {Σ Σ}_{n no = = 00}^{N N - - 11} x x ((n no)) {x x}^{* *} ((((n no + + u u)) mod mod N N))$

当u≠0时，为一个非常小的数，理想的，When u≠0, For a very small number, ideally,

R(u)互相关应该满足如下条件：R(u) cross-correlation should meet the following conditions:

$R (u) = {\begin{matrix} N, u = 0; \\ 0, u &NotEqual; 0; \end{matrix}$ $R (u) = {\begin{matrix} N, u = 0; \\ 0, u &NotEqual; 0; \end{matrix}$

对于循环前缀类型2，同步信号也在一个OFDM符号中发送，其数据部分为：For cyclic prefix type 2, the synchronization signal is also sent in one OFDM symbol, and its data part is:

x(m+1)....x(2N-1)，x(0)....x(m).x(m+1)....x(2N-1), x(0)....x(m).

其中，m是一个常数，满足如下条件Among them, m is a constant, satisfying the following conditions

m＝L2+D (1)m=L2+D (1)

其中，D也是一个常数，满足如下条件：Among them, D is also a constant, satisfying the following conditions:

0≤D≤N-L1-L2-2 (2)0≤D≤N-L1-L2-2 (2)

D的一个较佳值为：A good value of D is:

表示对x向上取整； (3)

Indicates that x is rounded up; (3)

_x_表示对x向下取整；_x_ indicates that x is rounded down;

(二)接收：在移动台侧，移动台检测同步信号的步骤可以描述为：(2) Receiving: at the mobile station side, the steps for the mobile station to detect the synchronization signal can be described as:

1.粗定时检测：利用同步信号的时域重复特性确定同步信号的粗定时。这个步骤属于现有技术。这一步通常可以把同步信号的定时误差限制在循环前缀范围内。1. Coarse timing detection: use the time domain repetition characteristic of the synchronization signal to determine the coarse timing of the synchronization signal. This step belongs to the prior art. This step usually limits the timing error of the sync signal to within the cyclic prefix range.

2.细定时检测与循环前缀长度类型检测：2. Fine timing detection and cyclic prefix length type detection:

设：set up:

a.移动台的本地参考同步信号为：x^*(0)，x^*(N-1)，x^*(N-2)...x^*(1)。a. The local reference synchronization signal of the mobile station is: x ^* (0), x ^* (N-1), x ^* (N-2)...x ^* (1).

b.根据粗定时获得的一个同步信号数据为：r(0)...r(N-1)，r(N)...r(2N-1).移动台细定时检测与循环前缀长度类型检测的步骤可以描述为：b. A synchronization signal data obtained according to coarse timing is: r(0)...r(N-1), r(N)...r(2N-1). Fine timing detection and cyclic prefix length of mobile station The steps of type detection can be described as:

2.1计算量度c1，c1是r(0)....r(N-1)与本地参考同步信号的圆周卷积；2.1 Calculate the measure c1, which is the circular convolution of r(0)..r(N-1) with the local reference synchronization signal;

2.2计算量度c2，c2是r(N)....r(2N-1)与本地参考同步信号的圆周卷积；2.2 Calculate the measure c2, which is the circular convolution of r(N)....r(2N-1) with the local reference synchronization signal;

2.3计算量度c，c＝|c1|²+|c2|²，|x|表示x的模。2.3 Calculate the measure c, c=|c1| ² +|c2| ² , where |x| represents the modulus of x.

2.4计算量度c中最大元素对应的索引：设量度c中最大元素对应的索引为k，0≤k≤N-1。移动台可以根据下表确定同步信号的细定时位置与循环前缀类型：2.4 Calculate the index corresponding to the largest element in the metric c: set the index corresponding to the largest element in the metric c to be k, where 0≤k≤N-1. The mobile station can determine the fine timing position and cyclic prefix type of the synchronization signal according to the following table:

k k 判断规则 Judgment rules 0≤k≤L1 0≤k≤L1 循环前缀类型1，并且接收信号r(0)相对于同步信号的理想定时位置超前k个样点。 Cyclic prefix type 1, and the received signal r(0) is ahead of the ideal timing position of the synchronization signal by k samples. L1+1≤k≤N-L2-D-2 L1+1≤k≤N-L2-D-2 循环前缀类型2，并且接收信号r(0)相对于同步信号的理想定时位置延后(N-L2-D-1-k)个样点。 Cyclic prefix type 2, and the received signal r(0) is delayed by (N-L2-D-1-k) samples relative to the ideal timing position of the synchronization signal. N-L2-D-1≤k≤N-D-1 N-L2-D-1≤k≤N-D-1 循环前缀类型2，并且接收信号r(0)相对于同步信号的理想定时位置超前(k+L2+D-N)个样点。 Cyclic prefix type 2, and the received signal r(0) is ahead of the ideal timing position of the synchronization signal by (k+L2+D-N) samples. N-D≤k≤N-1 N-D≤k≤N-1 循环前缀类型1，并且接收信号r(0)相对于同步信号的理想定时位置延后(N-k-1)个样点。 Cyclic prefix type 1, and the received signal r(0) is delayed by (N-k-1) samples relative to the ideal timing position of the sync signal.

表1Table 1

第二步，结合附图以两种循环前缀类型为例说明本发明原理和详细步骤：The second step is to illustrate the principle and detailed steps of the present invention by taking two types of cyclic prefixes as examples in conjunction with the accompanying drawings:

(一)循环前缀类型(1) Cyclic prefix type

图1为OFDM符号示意图，图1中矩形框中填充左斜线的部分为OFDM符号的循环前缀。矩形框中没有填充内容的为OFDM符号的数据部分，循环前缀部分通常是数据部分后面一段数据的重复。在接收机端，接收机解调OFDM信号时，循环前缀部分的内容通常都会被接收机丢弃，因此，在解调前，接收机必须事先获得OFDM符号的数据部分的位置。Figure 1 is a schematic diagram of an OFDM symbol, and the part filled with left oblique lines in the rectangular box in Figure 1 is the cyclic prefix of the OFDM symbol. The data part of the OFDM symbol is not filled in the rectangular box, and the cyclic prefix part is usually the repetition of a piece of data after the data part. At the receiver side, when the receiver demodulates the OFDM signal, the content of the cyclic prefix part is usually discarded by the receiver. Therefore, before demodulation, the receiver must obtain the position of the data part of the OFDM symbol in advance.

由于环境、业务类型的差异，系统会选择不同的循环前缀类型。图2标识的是一个采用短循环前缀(循环前缀类型1)的子帧，其中，矩形框中填充左斜线的部分为OFDM符号的循环前缀，矩形框中没有填充内容的为OFDM符号的数据部分；图3标识的是一个采用长循环前缀(循环前缀类型2)的子帧，其中，矩形框中填充左斜线的部分为OFDM符号的循环前缀，矩形框中没有填充内容的为OFDM符号的数据部分。如图2和图3所示，由于循环前缀长度的差异，采用长循环前缀的子帧内有6个OFDM符号，采用短循环前缀的子帧内有7个OFDM符号。基于OFDM技术的LTE系统就采用了和图2和图3类似的帧结构。对于超远覆盖或者多播、广播信息传送的应用场景，系统会选用长循环前缀的子帧，此时，每个子帧有6个符号，可减小接收信号符号间干扰；对于一般应用，系统会选用短循环前缀的子帧，此时，每个子帧有7个符号，可提高系统的吞吐量。Due to differences in environments and business types, the system will select different cyclic prefix types. Figure 2 identifies a subframe that uses a short cyclic prefix (cyclic prefix type 1), where the part filled with left slashes in the rectangular box is the cyclic prefix of the OFDM symbol, and the part that is not filled in the rectangular box is the data of the OFDM symbol part; Figure 3 identifies a subframe that uses a long cyclic prefix (cyclic prefix type 2), wherein the part filled with left slashes in the rectangular box is the cyclic prefix of the OFDM symbol, and the part that is not filled in the rectangular box is the OFDM symbol the data portion of the . As shown in Figure 2 and Figure 3, due to the difference in cyclic prefix length, there are 6 OFDM symbols in a subframe using a long cyclic prefix, and 7 OFDM symbols in a subframe using a short cyclic prefix. The LTE system based on OFDM technology adopts a frame structure similar to that shown in Figure 2 and Figure 3 . For the application scenarios of ultra-long-distance coverage or multicast and broadcast information transmission, the system will select subframes with long cyclic prefixes. At this time, each subframe has 6 symbols, which can reduce the interference between symbols of received signals; for general applications, the system A subframe with a short cyclic prefix is selected. At this time, each subframe has 7 symbols, which can improve the throughput of the system.

(二)发送：(2) Send:

图4和图5采用本发明的同步信号的生成示意图。设FFT的大小为2N，有两种循环循环前缀，分别为循环前缀类型1和类型2，循环前缀中包括的样点数目分别为：L1，L2。对于循环前缀类型1，系统发送的同步信号在一个OFDM符号中发送，其数据部分为： x(0)，x(1)....x(2N-1).并且满足如下条件：x(n)＝x(n+N)，n＝0，1...N-1.即，该信号在时域是重复的。数据部分最后L1个数据被复制到循环前缀部分，如图4所示。另外，数据x(0)，x(1)....x(2N-1)满足如下条件：x(n)＝x(n+N)，n＝0，1...N-1.即，该信号在时域是重复的。设x(0)，x(1)...x(N)与其循环移位的互相关R(u)为：Fig. 4 and Fig. 5 are schematic diagrams of synchronous signal generation using the present invention. Assuming that the size of the FFT is 2N, there are two types of cyclic cyclic prefixes, namely cyclic prefix type 1 and type 2, and the numbers of samples included in the cyclic prefix are: L1, L2. For cyclic prefix type 1, the synchronization signal sent by the system is sent in one OFDM symbol, and its data part is: x(0), x(1)....x(2N-1). And the following conditions are met: x( n)=x(n+N), n=0, 1...N-1. That is, the signal is repeated in the time domain. The last L1 data of the data part is copied to the cyclic prefix part, as shown in Figure 4. In addition, the data x(0), x(1)....x(2N-1) satisfy the following conditions: x(n)=x(n+N), n=0, 1...N-1. That is, the signal is repetitive in the time domain. Let the cross-correlation R(u) between x(0), x(1)...x(N) and its cyclic shift be:

当u≠0时，

为一个非常小的数，理想的，When u≠0,

For a very small number, ideally,

R(u)互相关应该满足如下条件： $R (u) = {\begin{matrix} N, u = 0; \\ 0, u &NotEqual; 0; \end{matrix}$ R(u) cross-correlation should meet the following conditions: $R (u) = {\begin{matrix} N, u = 0; \\ 0, u &NotEqual; 0; \end{matrix}$

图5则给出了循环前缀类型2情况下同步信号的生成示意图，对于循环前缀类型2，同步信号也在一个OFDM符号中发送，其数据部分为：Figure 5 shows a schematic diagram of the generation of synchronization signals in the case of cyclic prefix type 2. For cyclic prefix type 2, the synchronization signal is also sent in one OFDM symbol, and its data part is:

x(m+1)....x(2N-1)，x(0)....x(m).x(m+1)....x(2N-1), x(0)....x(m).

这里，m是一个常数，满足如下条件：Here, m is a constant that satisfies the following conditions:

m＝L2+D (1)m=L2+D (1)

0≤D≤N-L1-L2-2 (2)0≤D≤N-L1-L2-2 (2)

D的一个较佳值为：A good value of D is:

表示对x向上取整； (3)

Indicates that x is rounded up; (3)

_x_表示对x向下取整：_x_ indicates that x is rounded down:

从上面的描述可以看出，循环前缀类型2情况下的同步信号的数据部分是循环前缀类型1情况下数据部分的循环移位版本。利用这个特点，当粗同步的同步信号与本地参考信号进行圆周卷积时，对于不同类型的循环前缀，循环卷积的峰值将位于不同的位置，这在接收端移动台利用本地参考信号进行细定时时可以非常清楚的看出来。It can be seen from the above description that the data part of the synchronization signal in the case of cyclic prefix type 2 is a cyclically shifted version of the data part in the case of cyclic prefix type 1 . Taking advantage of this feature, when the coarse synchronization signal is circularly convolved with the local reference signal, the peak value of the circular convolution will be located at different positions for different types of cyclic prefixes, which is fined by the local reference signal at the receiving end. Timing can be seen very clearly.

(三)接收：(3) Receive:

在移动台侧，移动台检测同步信号，具体如图6所示，包括以下步骤：On the mobile station side, the mobile station detects the synchronization signal, specifically as shown in Figure 6, including the following steps:

①粗同步：① Coarse synchronization:

在步骤601，移动台进行同步信号的粗同步。这一步主要利用同步信号的时域重复特性完成。这一步属于现有技术，简述如下，设移动台接收到的信号为r(n)，r(n+1)........。移动台的粗定时包括如下步骤：In step 601, the mobile station performs coarse synchronization of the synchronization signal. This step is mainly accomplished by utilizing the time-domain repetition characteristic of the synchronization signal. This step belongs to the prior art, and is briefly described as follows, assuming that the signal received by the mobile station is r(n), r(n+1)........ The coarse timing of the mobile station includes the following steps:

1.计算如下量度：1. Calculate the following metrics:

$M m ((k k)) = = {| | {Σ Σ}_{p p = = 00}^{N N - - 11} {r r}^{* *} ((n no + + k k + + p p)) r r ((n no + + k k + + p p + + N N)) | |}^{22},, k k = = 0,1 0,1 . . . . . . . . 22 N N - - 11 - - - - - - ((44))$

2.计算获得M(k)中最大值对应的索引，设为m，则r(n+m)就是同步信号的粗定时位置。2. Calculate and obtain the index corresponding to the maximum value in M(k), set it as m, then r(n+m) is the rough timing position of the synchronization signal.

不失一般性，设Without loss of generality, let

a.根据同步信号的粗定时位置获得的同步信号数据为：a. The synchronization signal data obtained according to the rough timing position of the synchronization signal is:

r(0)...r(N-1)，r(N)...r(2N-1).r(0)...r(N-1), r(N)...r(2N-1).

b.移动台的本地参考同步信号序列为：x^*(0)，x^*(N-1)，x^*(N-2)...x^*(1).b. The local reference synchronization signal sequence of the mobile station is: x ^* (0), x ^* (N-1), x ^* (N-2)...x ^* (1).

②细定时检测：同步信号的细定时检测与循环前缀长度类型检测可以用步骤602～605描述：②Fine timing detection: The fine timing detection and cyclic prefix length type detection of the synchronization signal can be described by steps 602-605:

在步骤602，计算量度c1，量度c1等于r(0)....r(N-1)与本地参考同步信号的圆周卷积。设c1＝(c1(0)，c1(1)...c1(N-1))^T，c1的计算公式如下：In step 602, a metric c1 is calculated, which is equal to the circular convolution of r(0)....r(N-1) with the local reference synchronization signal. Let c1=(c1(0), c1(1)...c1(N-1)) ^T , the calculation formula of c1 is as follows:

$cl cl = = (\begin{matrix} c c 11 ((00)) \\ cl cl ((11)) \\ cl cl ((22)) \\ . . . . . . \\ c c 11 ((N N - - 11)) \end{matrix}) = = (\begin{matrix} {x x}^{* *} ((00)),, & {x x}^{* *} ((11)),, & {x x}^{* *} ((22)) & . . . . . . & {x x}^{* *} ((N N - - 11)) \\ {x x}^{* *} ((N N - - 11)),, & {x x}^{* *} ((00)),, & {x x}^{* *} ((11)) & . . . . . . & {x x}^{* *} ((N N - - 22)) \\ {x x}^{* *} ((N N - - 22)),, & {x x}^{* *} ((N N - - 11)),, & {x x}^{* *} ((00)) & . . . . . . & {x x}^{* *} ((N N - - 33)) \\ . . . . . . \\ {x x}^{* *} ((11)),, & {x x}^{* *} ((22)),, & {x x}^{* *} ((33)) & . . . . . . & {x x}^{* *} ((00)) \end{matrix}) \times \times (\begin{matrix} r r ((00)) \\ r r ((11)) \\ r r ((22)) \\ . . . . . . \\ r r ((N N - - 11)) \end{matrix}) - - - - - - ((55))$

为便于理解前面提出的同步信号发送方法，设同步信号的循环前缀类型是循环前缀类型1，根据同步信号的粗定时位置获得的同步信号数据为理想同步信号的一个定时超前版本，为便于讨论，这个不考虑信道的影响，设r(0)...r(N-1)，＝x(N-m)，x(N-m+1)...x(N-1)...x(N-m-1)，m＝0，1.2...L1.观察前面的圆周卷积公式(5)，利用前面描述的同步信号循环移位相关的性质In order to facilitate the understanding of the synchronization signal transmission method proposed above, the cyclic prefix type of the synchronization signal is assumed to be cyclic prefix type 1, and the synchronization signal data obtained according to the coarse timing position of the synchronization signal is a timing advanced version of the ideal synchronization signal. For the convenience of discussion, This does not consider the influence of the channel, let r(0)...r(N-1),=x(N-m), x(N-m+1)...x(N-1)...x( N-m-1), m=0, 1.2...L1. Observe the previous circular convolution formula (5), using the nature of the synchronization signal cyclic shift correlation described above

当u≠0时，

为一个非常小的数，理想的，When u≠0,

For a very small number, ideally,

( $R (u) = {\begin{matrix} N, u = 0; \\ 0, u &NotEqual; 0; \end{matrix}$ )，可以发现，当x(N-m)，x(N-m+1)...x(N-1)....x(N-m-1)与本地参考信号进行圆周卷积后，圆周卷积c1的峰值将位于c1的第m个元素位置处。( $R (u) = {\begin{matrix} N, u = 0; \\ 0, u &NotEqual; 0; \end{matrix}$ ), it can be found that when x(Nm), x(N-m+1)...x(N-1)....x(Nm-1) performs circular convolution with the local reference signal, the circular convolution The peak of the product c1 will be at the mth element position of c1.

再看另外一种假设，设同步信号的循环前缀类型是循环前缀类型1，根据同步信号的粗定时位置获得的同步信号数据为理想同步信号的一个定时延后版本，设r(0)...r(N-1)，＝x(p)，x(p+1)...x(N-1)，x(0)....x(p-1).p＝1，2...P.(P为最大可能的定时延后数目，需要指出的时，用于发送的同步信号具有时域重复特性，移动台根据粗定时获得的同步信号，绝大部分情况是理想同步信号的一个定时超前版本，由于前面描述的粗同步信号属于成熟算法，这里不再对其进行深入描述)观察前面的圆周卷积公式(5)，可以发现，当x(N-p)，x(N-p+1)...x(N-1)....x(N-p-1)与本地参考信号进行圆周卷积后，圆周卷积c1的峰值将位于c1的第(N-p)个元素位置处。Let’s look at another hypothesis. Let the cyclic prefix type of the synchronization signal be cyclic prefix type 1, and the synchronization signal data obtained according to the rough timing position of the synchronization signal is a timing delayed version of the ideal synchronization signal. Let r(0).. .r(N-1),=x(p),x(p+1)...x(N-1),x(0)....x(p-1).p=1,2 ...P.(P is the maximum possible number of timing delays. When it needs to be pointed out, the synchronization signal used for transmission has the characteristics of time domain repetition. The synchronization signal obtained by the mobile station according to the rough timing is ideal synchronization in most cases A timing advanced version of the signal, since the coarse synchronization signal described above belongs to a mature algorithm, it will not be described in depth here) Observing the previous circular convolution formula (5), it can be found that when x(N-p), x(N -p+1)...x(N-1)....x(N-p-1) After circular convolution with the local reference signal, the peak value of the circular convolution c1 will be located at the (N-p)th element of c1 location.

再看第三种假设，设同步信号的循环前缀类型是循环前缀类型2，设发送的同步信号的数据部分为：x(L2+D+1)....x(N-1)，x(0)...x(L2+D)，x(L2+D+1)....x(N-1)，x(0)...x(L2+D)。根据同步信号的粗定时位置获得的同步信号数据为理想同步信号的一个定时超前版本，设Let's look at the third hypothesis again. Let the cyclic prefix type of the synchronization signal be cyclic prefix type 2, and set the data part of the synchronization signal sent as: x(L2+D+1)....x(N-1), x (0)...x(L2+D), x(L2+D+1)....x(N-1), x(0)...x(L2+D). The synchronous signal data obtained according to the coarse timing position of the synchronous signal is a timing advanced version of the ideal synchronous signal, set

r(0)...r(N-1)，＝x(L2+D+1-m)，x(L2+D+1-m+1)...x(N-1)，x(0)....x(m-1).M＝0，1....L2。观察前面的圆周卷积公式(5)，可以发现，当x(L2+D+1-m)，x(L2+D+1-m+1)...x(N-1)，x(0)....x(m-1)与本地参考信号进行圆周卷积后，圆周卷积c1的峰值将位于c1的第(N-L2-D-1+m)个元素位置处。r(0)...r(N-1),=x(L2+D+1-m), x(L2+D+1-m+1)...x(N-1), x( 0)....x(m-1).M=0, 1....L2. Observing the previous circular convolution formula (5), it can be found that when x(L2+D+1-m), x(L2+D+1-m+1)...x(N-1), x( 0)....x(m-1) performs circular convolution with the local reference signal, and the peak value of the circular convolution c1 will be located at the (N-L2-D-1+m)th element position of c1.

再看第四种假设，设同步信号的循环前缀类型是循环前缀类型2，设发送的同步信号的数据部分为：x(L2+D+1)....x(N-1)，x(0)...x(L2+D)，x(L2+D+1)....x(N-1)，x(0)...x(L2+D)。根据同步信号的粗定时位置获得的同步信号数据为理想同步信号的一个定时延后版本，设Look at the fourth hypothesis again. Let the cyclic prefix type of the synchronization signal be cyclic prefix type 2, and let the data part of the synchronization signal sent be: x(L2+D+1)....x(N-1), x (0)...x(L2+D), x(L2+D+1)....x(N-1), x(0)...x(L2+D). The synchronous signal data obtained according to the coarse timing position of the synchronous signal is a timing delayed version of the ideal synchronous signal, set

r(0)...r(N-1)，＝x(L2+D+1+p)，x(L2+D+1+p+1)...x(N-1)，x(0)....x(L2+D+p).观察前面的圆周卷积公式(5)，可以发现，当x(L2+D+1+p)，x(L2+D+1+p+1)...x(N-1)，x(0)....x(L2+D+p)与本地参考信号进行圆周卷积后，圆周卷积c1的峰值将位于c1的第(N-L2-D-1-p)个元素位置处。r(0)...r(N-1),=x(L2+D+1+p), x(L2+D+1+p+1)...x(N-1), x( 0)....x(L2+D+p). Observing the previous circular convolution formula (5), it can be found that when x(L2+D+1+p), x(L2+D+1+p +1)...x(N-1), x(0)....x(L2+D+p) performs circular convolution with the local reference signal, the peak value of circular convolution c1 will be located at the first (N-L2-D-1-p) element positions.

从上面的分析可以看出，通过合理的设计循环前缀类型2的同步信号相对于循环前缀类型1的延时，移动台可以在对同步信号进行细定时的同时获得循环前缀长度类型。如前所述，移动台根据粗定时获得的同步信号，绝大部分情况是理想同步信号的一个定时超前版本，因此，对于这种定时超前的情况需要重点考虑。当定时超前时，对于循环前缀类型1，圆周卷积c1的峰值索引范围为[0，L1]。为了避免混淆，对于循环前缀类型2，圆周卷积c1的峰值索引期望范围应该是[L1+1，N-1]，为了达到这个目的，循环前缀类型2相对于循环前缀类型1的循环移位数目应该满足如下关系：From the above analysis, it can be seen that by reasonably designing the delay of the synchronization signal of the CP type 2 relative to the CP type 1, the mobile station can obtain the CP length type while fine-tuning the synchronization signal. As mentioned above, the synchronization signal obtained by the mobile station according to the rough timing is, in most cases, a timing-advanced version of the ideal synchronization signal. Therefore, it is important to consider this timing-advance situation. When the timing is advanced, for cyclic prefix type 1, the peak index range of the circular convolution c1 is [0, L1]. To avoid confusion, for cyclic prefix type 2, the expected range of the peak index of circular convolution c1 should be [L1+1, N-1], to achieve this, the cyclic shift of cyclic prefix type 2 relative to cyclic prefix type 1 The number should satisfy the following relationship:

L2≤m≤N-L1-2 (6)L2≤m≤N-L1-2 (6)

不失去一般性，设m＝L2+D，D则需要满足前面提到的公式(2)，即：Without loss of generality, let m=L2+D, and D needs to satisfy the aforementioned formula (2), namely:

0≤D≤N-L1-L2-20≤D≤N-L1-L2-2

此时，对于循环前缀类型2，圆周卷积c1的峰值索引范围是[N-L2-D-1，N-D-1]。At this time, for cyclic prefix type 2, the peak index range of circular convolution c1 is [N-L2-D-1, N-D-1].

此时还有N-L1-L2-2个位置可以用来对定时延迟的情况进行判定，当D取较佳值(公式3)，即：At this time, there are still N-L1-L2-2 positions that can be used to judge the timing delay. When D takes a better value (formula 3), that is:

表示x向上取整；

Indicates that x is rounded up;

_x_表不对x向下取整；_x_ indicates that x is not rounded down;

移动台可以判定的最大定时延迟的范围对于循环前缀类型1/2是相同的。The range of maximum timing delays that a mobile station can decide is the same for cyclic prefix type 1/2.

当然，根据数字信号处理原理，r(N-1)....r(0)与本地参考同步信号的圆周卷积可以利用傅立叶变换高效实现。Of course, according to the principle of digital signal processing, the circular convolution of r(N-1)..r(0) with the local reference synchronization signal can be realized efficiently by Fourier transform.

在步骤603，计算量度c2，r(N)....r(2N-1)与本地参考同步信号的圆周卷积。计算方法与步骤402完全相同。In step 603, the circular convolution of the measures c2,r(N)....r(2N-1) with the local reference synchronization signal is calculated. The calculation method is exactly the same as step 402.

需要之处的是，这一步并不是必须步骤，比如，移动台所处的信道的信噪比很高，这一步就可以省略。What is needed is that this step is not a necessary step, for example, if the signal-to-noise ratio of the channel where the mobile station is located is very high, this step can be omitted.

在步骤604，计算量度c，c＝|c1|²+|c2|²，|x|表示x的模。当步骤403省略时，c＝|c1|² In step 604, a measure c is calculated, c=|c1| ² +|c2| ² , where |x| represents the modulus of x. When step 403 is omitted, c=|c1| ²

在步骤605，计算量度c中最大元素对应的索引：设量度c中最大元素对应的索引为k，0≤k≤N-1。由于循环前缀类型2发送的同步信号是循环前缀类型1的循环移位版本，量度c中最大元素对应的索引可以认为是循环前缀类型、接收信号r(0)相对于同步信号的理想定时位置的同步偏差二元函数，移动台可以根据下表确定同步信号的细定时位置与循环前缀类型：In step 605, calculate the index corresponding to the largest element in the metric c: set the index corresponding to the largest element in the metric c to be k, where 0≤k≤N-1. Since the synchronization signal sent by cyclic prefix type 2 is a cyclically shifted version of cyclic prefix type 1, the index corresponding to the largest element in the metric c can be considered as the ideal timing position of the cyclic prefix type and the received signal r(0) relative to the synchronization signal Synchronization deviation binary function, the mobile station can determine the fine timing position and cyclic prefix type of the synchronization signal according to the following table:

表1Table 1

最后，本发明不限制于两种循环前缀类型，可以任意多种循环前缀类型，下面以三种循环前缀类型的最佳实施例为例来进一步说明本发明：Finally, the present invention is not limited to two types of cyclic prefixes, and can be any number of cyclic prefix types. The following best examples of three types of cyclic prefixes are used as examples to further illustrate the present invention:

(一)循环移位(1) Cyclic shift

设FFT的大小为2N，有三种循环前缀，分别为循环前缀类型1，2和3，循环前缀中包括的样点数目分别为：L1，L2，L3。不失一般性，Assuming that the size of the FFT is 2N, there are three types of cyclic prefixes, namely cyclic prefix types 1, 2 and 3, and the numbers of samples included in the cyclic prefixes are: L1, L2, and L3, respectively. Without loss of generality,

①对于循环前缀类型1，设系统发送的同步信号在一个OFDM符号中发送，其数据部分为：x(0)，x(1)....x(2N-1).并且满足如下条件：x(n)＝x(n+N)，n＝0，1...N-1.即，该信号在时域是重复的。设x(0)，x(1)...x(N)与其循环移位的互相关R(u)为：①For cyclic prefix type 1, suppose the synchronization signal sent by the system is sent in one OFDM symbol, and its data part is: x(0), x(1)....x(2N-1). And the following conditions are met: x(n)=x(n+N), n=0, 1...N-1. That is, the signal is repeated in the time domain. Let the cross-correlation R(u) between x(0), x(1)...x(N) and its cyclic shift be:

②对于循环前缀类型2，同步信号也在一个OFDM符号中发送，其数据部分为：② For cyclic prefix type 2, the synchronization signal is also sent in one OFDM symbol, and its data part is:

x(m+1)....x(2N-1)，x(0)....x(m).x(m+1)....x(2N-1), x(0)....x(m).

其中，m的一个较佳值为：Among them, a better value of m is:

m＝N-L1-D-1.m=N-L1-D-1.

表示对x向上取整；

Indicates that x is rounded up;

_x_表示对x向下取整；_x_ indicates that x is rounded down;

(1’) (1')

③对于循环前缀类型3，同步信号也在一个OFDM符号中发送，其数据部分为：③ For cyclic prefix type 3, the synchronization signal is also sent in one OFDM symbol, and its data part is:

x(p+1)....x(2N-1)，x(0)....x(p).x(p+1)....x(2N-1), x(0)....x(p).

其中，p的一个较佳值为：Among them, a better value of p is:

p＝N-L1-L2-2D-2.p=N-L1-L2-2D-2.

表示对x向上取整；

Indicates that x is rounded up;

_x_表示对x向下取整；_x_ indicates that x is rounded down;

(1”) (1")

(二)对应规则(2) Corresponding rules

k k 判断规则 Judgment rules 0≤k≤L1 0≤k≤L1 循环前缀类型1，并且接收信号r(0)相对于同步信号的理想定时位置超前k个样点。 The cyclic prefix type is 1, and the received signal r(0) is k samples ahead of the ideal timing position of the synchronization signal. L1+1≤k≤L1+D L1+1≤k≤L1+D 循环前缀类型2，并且接收信号r(0)相对于同步信号的理想定时位置延后(L1+D+1-k)个样点。 Cyclic prefix type 2, and the received signal r(0) is delayed by (L1+D+1-k) samples relative to the ideal timing position of the synchronization signal. L1+D+1≤k≤L1+D+1+L2 L1+D+1≤k≤L1+D+1+L2 循环前缀类型2，并且接收信号r(0)相对于同步信号的理想定时位置超前(k-L1-D-1)个样点。 Cyclic prefix type 2, and the received signal r(0) is (k-L1-D-1) samples ahead of the ideal timing position of the synchronization signal. L1+L2+D+2≤k≤L1++L2+2D+1 L1+L2+D+2≤k≤L1++L2+2D+1 循环前缀类型3，并且接收信号r(0)相对于同步信号的理想定时位置延后(L1+L2+2D+2-k)个样点。 Cyclic prefix type 3, and the received signal r(0) is delayed by (L1+L2+2D+2-k) samples relative to the ideal timing position of the synchronization signal. L1++L2+L3+2D+3≤k≤L1+L2+L3+2D+2 L1++L2+L3+2D+3≤k≤L1+L2+L3+2D+2 循环前缀类型3，并且接收信号r(0)相对于同步信号的理想定时位置超前(k-L1-L2-2D-2)个样点 Cyclic prefix type 3, and the received signal r(0) is ahead of the ideal timing position of the synchronization signal by (k-L1-L2-2D-2) samples L1+L2+2D+3≤k≤L1+L2+L3+3D+2 L1+L2+2D+3≤k≤L1+L2+L3+3D+2 循环前缀类型1，并且接收信号r(0)相对于同步信号的理想定时位置延后(L1+L2+L3+3D+3-k)个样点 Cyclic prefix type 1, and the received signal r(0) is delayed by (L1+L2+L3+3D+3-k) samples relative to the ideal timing position of the synchronization signal

。

表2Table 2

Claims

1, a kind of synchronous signal transmission and detection method of OFDM system, it is characterized in that, receiving terminal synchronously carries out cyclic prefix type detection simultaneously, comprises the following steps:

1.1) The sending end cyclically shifts the data part of the synchronization signal according to the cyclic prefix type so that the peak value of the circular convolution with the local reference signal at the receiving end is located at a different position;

1.2) The receiving end distinguishes and restores the cyclic prefix type according to the different positions during the synchronization timing process.

2. The sending and detecting method according to claim 1, wherein the synchronization timing process uses local reference synchronization signals x ^* (0), x ^* (N-1), x ^* (N-2).. .x ^* (1), including the following specific steps:

2.1) Utilize the time-domain repetition characteristics of the standard synchronization signal to determine the rough timing data of the synchronization signal

r(0)...r(N-1), r(N)...r(2N-1);

2.2) Compute measure c1: c1 is r(0)....r(N-1) with local reference sync signal

Circular convolution of x ^* (0), x ^* (N-1), x ^* (N-2)...x ^* (1);

2.3) Calculate the measurement c, c=|c1| ² , |x| represents the modulus of x, and x is an arbitrary expression;

2.4) Calculate the index k corresponding to the largest element in the measure c, according to the value range of k and the

The corresponding rule of the evolution of the circular convolution peak rule determines the fine timing position and cyclic prefix type of the synchronization signal.

3. The sending and detecting method according to claim 2, characterized in that, said step 2.2) also includes calculating the measure c2: c2 is r(N)....r(2N-1) and the local reference synchronization signal x Circular convolution of ^* (0), x ^* (N-1), x ^* (N-2)...x ^* (1); measure c=|c1| ²⁺ in steps 1.3) and 1.4) |c2| ² .

4. The sending and detecting method according to claim 2, characterized in that the step 2.1) includes:

2.1.1) Calculate the metric M(k):

m (k) = {| Σ_{p = 0}^{N - 1} r^{*} (no + k + p) r (no + k + p + N) |}^{2}, k = 0,1 . . . . 2 N - 1;

2.1.2) Calculate the index corresponding to the maximum value in M(k), set it as m, then r(n+m) is the rough timing position of the synchronization signal;

2.1.3) Obtain the rough timing data of the synchronization signal according to the rough timing position: r(0)...r(N-1), r(N)...r(2N-1).

5. The sending and detecting method according to claim 2, wherein the circular convolution is:

c c 11 = = \begin{matrix}  \end{matrix} [\begin{matrix} c c 11 ((00)) \\ c c 11 ((11)) \\ c c 11 ((22)) \\ \dots \dots \\ c c 11 ((N N - - 11)) \end{matrix}] = = [\begin{matrix} {x x}^{* *} ((00)),, & {x x}^{* *} ((11)),, & {x x}^{* *} ((22)) & \dots \dots & {x x}^{* *} ((N N - - 11)) \\ {x x}^{* *} ((N N - - 11)),, & {x x}^{* *} ((00)),, & {x x}^{* *} ((11)) & \dots \dots & {x x}^{* *} ((N N - - 22)) \\ {x x}^{* *} ((N N - - 22)),, & {x x}^{* *} ((N N - - 11)),, & {x x}^{* *} ((00)) & \dots \dots & {x x}^{* *} ((N N - - 33)) \\ \dots \dots \\ {x x}^{* *} ((11)),, & {x x}^{* *} ((22)),, & {x x}^{* *} ((33)) & \dots \dots & {x x}^{* *} ((00)) \end{matrix}] \times \times [\begin{matrix} r r ((00)) \\ r r ((11)) \\ r r ((22)) \\ \dots \dots \\ r r ((N N - - 11)) \end{matrix}] . .

6. The sending and detecting method according to claim 1, wherein there are two types of cyclic prefixes, the lengths of cyclic prefixes are L1 and L2 respectively, and the corresponding cyclic shifts are 0 and left shifted by m+1 respectively Bits, corresponding to the data part are ideal synchronous signals x(0), x(1)....x(2N-1) and left-shifted synchronous signals x(m+1)....x(2N- 1), x(0)....x(m); m=L2+D, 0≤D≤N-L1-L2-2, 2N is the length of the data part.

7. The sending and detecting method according to claim 5, wherein D takes an optimal value:

Indicates that x is rounded up;

_x_ indicates that x is rounded down.

8. The sending and detecting method according to claim 2, 4 or 6, characterized in that the corresponding rules are:

K judgment rule 0≤k≤L1 Cyclic prefix type 1, and the received signal r(0) is ahead of the ideal timing position of the synchronization signal by k samples. L1+1≤k≤N-L2-D-2 Cyclic prefix type 2, and the received signal r(0) is delayed by (N-L2-D-1-k) samples relative to the ideal timing position of the sync signal N-L2-D-1≤k≤N-D-1 Cyclic prefix type 2, and the received signal r(0) is synchronized relative to

The ideal timing position of the signal is (k+L2+D-N) samples ahead. N-D≤k≤N-1 Cyclic prefix type 1, and the received signal r(0) is delayed by (N-k-1) samples relative to the ideal timing position of the synchronization signal.

9. The sending and detecting method according to claim 1, wherein the cyclic prefix type can be three types, the cyclic prefix lengths are L1, L2 and L3 respectively, and the corresponding cyclic shifts are 0 and left shift respectively m+1 bit and left shift p+1 bit, m=N-L1-D-1, p=N-L1-L2-2D-2;

in,

Indicates that x is rounded up;

_x_ indicates that x is rounded down.

10. The sending and detecting method according to claim 2 or 9, characterized in that the corresponding rules are:

k judgment rule 0≤k≤L1 Cyclic prefix type 1, and the received signal r(0) is ahead of the ideal timing position of the synchronization signal by k samples. L1+1≤k≤L1+D Cyclic prefix type 2, and the received signal r(0) is delayed by (L1+D+1-k) samples relative to the ideal timing position of the synchronization signal. L1+D+1≤k≤L1+D+1+L2 Cyclic prefix type 2, and the received signal r(0) is ahead of the ideal timing position of the synchronization signal by (k-L1-D-1) samples. L1+L2+D+2≤k≤L1++L2+2D+1 Cyclic prefix type 3, and the received signal r(0) is delayed by (L1+L2+2D+2-k) samples relative to the ideal timing position of the synchronization signal.

L1++L2+L3+2D+3≤k≤L1+L2+L3+2D+2 Cyclic prefix type 3, and the received signal r(0) is ahead of the ideal timing position of the synchronization signal by (k-L1-L2-2D-2) samples L1+L2+2D+3≤k≤L1+L2+L3+3D+2 Cyclic prefix type 1, and the received signal r(0) is delayed by (L1+L2+L3+3D+3-k) samples relative to the ideal timing position of the synchronization signal