CN100369394C

CN100369394C - Method for carrying out frame synchronization timing at receiving and of base station of WiMAX system

Info

Publication number: CN100369394C
Application number: CNB2005101177993A
Authority: CN
Inventors: 吴永东; 郭华志
Original assignee: Beijing Northern Fiberhome Technologies Co Ltd
Current assignee: Wuhan Feng And Zhida Information Technology LLC
Priority date: 2005-11-14
Filing date: 2005-11-14
Publication date: 2008-02-13
Anticipated expiration: 2025-11-14
Also published as: CN1761175A

Abstract

The present invention provides a method of timing synchronization based on OFDM and used for a receiving terminal of a base station of a WiMAX system. The present invention is characterized in that not only the timing synchronization can be performed on repeated 128 sample values of short lead codes based on uplink sporadic pulses, but also timing synchronization correction can be performed on CP cyclical prefixes based on symbols of OFDM frames. According to the present invention, the precision of the timing synchronization can be improved; later influences on frequency synchronization can be reduced; angles of phase deflection can be decreased; the performance of the system can be improved.

Description

Method for frame synchronization timing at receiving end of WiMAX system base station

Technical Field

The invention relates to a method for carrying out frame synchronization timing at a receiving end of a base station of a WiMAX (worldwide interoperability for microwave Access) system in the communication field.

Background

In recent years, a new broadband wireless access technology is introduced by the computer industry and named WiMAX, as the global mobile operators, equipment manufacturers, mobile phone manufacturers and government authorities of all countries invest considerable funds and energy for the construction and operation of 3 rd generation mobile communication networks (3G). WiMAX is an abbreviation for Worldwide Interoperability for Microwave Access, commonly translated into "Worldwide Interoperability for Microwave Access".

WiMAX technology is based on the 802.16 series of standards of IEEE. The earliest IEEE802.16 standard was approved in 12 months of 2001 and is a wireless metropolitan area network standard established for a 10 to 66GHz high band line-of-sight environment. The 802.16 standard includes three standards of IEEE802.16 a, 802.16-2004 and 802.16 e. 802.16a is designed for non-line-of-sight broadband fixed access systems operating in the 2 to 11GHz band, and 802.16-2004 (formerly known as 802.16 d) is an enhancement to 802.16 a. While the 802.16e standard supporting mobility is about to be released, it is a further extension of 802.16a/2004 to the purpose of increasing data mobility in the existing standard.

In the 802.16 series standards, the technical requirements of the air interface between the Base Station BS (Base Station) and the Subscriber Station SS (Subscriber Station) are specified in detail, and in particular, the frame structure requirements of the physical layer, the system design parameters, and the like are specified in detail.

WiMAX systems support a variety of physical layer air interfaces, where support for air interfaces of OFDM modulation techniques has become an indispensable fundamental requirement. OFDM, an abbreviation of Orthogonal Frequency Division Multiplex, means Orthogonal Frequency Division multiplexing in chinese. The OFDM technology is based on orthogonal multiple carriers, which is a kind of multi-carrier spread spectrum technology. The great advantage of OFDM is to combat frequency selective fading or narrow-band interference, which can cause failure of the entire communication link in a single carrier system, but where only a small fraction of the carriers are disturbed and error correction codes can be used for these sub-channels. In the OFDM system, the carriers of each sub-channel are orthogonal to each other, and the frequency spectrums are overlapped with each other, so that the mutual interference among the sub-carriers is reduced, and the frequency spectrum utilization rate is improved. The invention is mainly designed for timing synchronization of a receiving end of a base station of a WiMAX system (IEEE 802.16-2004) based on OFDM.

The WiMAX system base station transceiver based on the OFDM mode is shown in fig. 1: the upper half of the figure is a block diagram of a transmitter and the lower half is a block diagram of a receiver. The figure shows the same block diagram for the IFFT (inverse fourier transform) and FFT (fourier transform) operations, which are performed in very similar steps and can be implemented in the same hardware. The transmitter performs forward error correction coding, interleaving, modulation (constellation mapping), pilot insertion, serial/parallel conversion, IFFT conversion, parallel/serial conversion, cyclic prefix insertion, windowing, digital-to-analog conversion (DAC), and radio frequency transmission to a wireless communication environment (RF TX) on data bits from a MAC (media access control) layer. The receiver is the inverse process of the transmitter and mainly comprises radio frequency reception (RF RX), analog-to-digital conversion (ADC), synchronization, cyclic prefix removal, serial/parallel conversion, FFT demodulation, channel correction, digital demodulation (constellation inverse mapping), deinterleaving, error correction decoding and other links.

From the whole WiMAX transceiver block diagram, it can be seen that, in the design of the WiMAX system based on OFDM, the requirement for synchronization is quite high in order to ensure the orthogonality between the subcarriers. Once out of synchronization, orthogonality between subcarriers is affected, thereby severely affecting system performance. Therefore, the superiority and inferiority of the synchronization algorithm can lead to the performance of the whole system. A high-performance and practical synchronization algorithm is designed, and becomes a key link of the whole system design.

The synchronization algorithm includes time domain synchronization and frequency domain synchronization. Wherein, the frame synchronization (time domain synchronization) is the precondition and the foundation for the whole WiMAX system to correctly receive the data. The physical layer of the 802.16-2004 protocol based on the OFDM mode supports frame-based transmission. One frame includes one downlink subframe and one uplink subframe. The downlink subframe consists of one downlink physical layer PDU and the uplink subframe consists of contention slots for initial search and bandwidth request purposes and one or more uplink physical layer bursts transmitted by different SSs. The Burst of each uplink SS contains a short preamble.

The downlink physical layer PDU is started by a long preamble for physical layer synchronization. An uplink physical layer PDU (protocol data unit) is started by a short preamble for synchronization of uplink bursts.

An OFDM symbol is a useful symbol time T from the time domain _b And length T of cyclic prefix CP _g Sum T _s . The CP is a copy of the length of the end of the useful symbol time to collect multipath and maintain orthogonality of the subcarriers. The length of the CP should be larger than the maximum multipath delay length, and the OFDM symbol is as shown in fig. 2.

According to the specification of an IEEE 802.16-2004 protocol, the value of an uplink short lead code in a frequency domain sequence of the WiMAX system based on OFDM is determined by the following formula:

among them, \ 58286, the factor is related to 3dB gain. P _ALL The frequency domain sequence is specified uniformly by the protocol. Thus, the preamble received by the base station to the SS (subscriber station) appears in the time domain as two 128 sample repetitions, preceded by a cyclic prefix of the preamble, as shown in fig. 3.

According to the design of practical system, the base station receives bursts of physical layer of the up-going SS (subscriber station), it must first detect the short preamble in front of the Burst of each SS for frame timing synchronization, and once the preamble is detected, the base station can receive the data Burst.

The WiMAX system based on OFDM has very strict requirements on the orthogonality between subcarriers, and whether the orthogonality is strict or not directly affects the performance of the system. The requirement of orthogonality is accomplished by means of a synchronization module. In the WiMAX system, synchronization is generally performed according to the sequence of timing synchronization (symbol, frame, and sample synchronization) and frequency offset synchronization (there may be an algorithm design in which timing and frequency offset synchronization are performed simultaneously). The accuracy of the timing synchronization directly affects the accuracy of the following frequency domain synchronization. In the design of an actual system, considering that an OFDM system using a cyclic prefix has relatively loose requirements on symbol timing synchronization, and the system has no requirement on timing frequency offset (a WiMAX system is sensitive to frequency offset, which not only affects the deflection of signal phases, but also destroys the orthogonality among subcarriers, which affects the amplitude and signal-to-noise ratio of signals, and requires coarse synchronization before FFT transformation, and fine synchronization must be performed after FFT transformation), the orthogonality among subcarriers will not be destroyed as long as the starting time of a symbol falls within the cyclic prefix, and ICI will not be caused, so long as phase compensation is performed. However, in multipath environment, in order to obtain the best system performance, the best symbol timing needs to be determined, and although the starting point of the symbol timing can be arbitrarily selected in the guard interval, it is obvious that any change of the symbol timing increases the sensitivity of the OFDM system to the delay spread, so the delay spread that the system can tolerate is lower than its design value. To minimize this negative effect, it is desirable to minimize the error in symbol timing synchronization.

The timing deviation of the timing synchronization error may be positive or negative, and a positive timing indicates an advance of the optimum timing, and a negative timing indicates a retard of the optimum timing. Through analysis of the OFDM symbols with the cyclic prefixes, the fact that if the timing deviation is positive, the symbol timing is advanced and falls into the cyclic prefixes is found, an OFDM window for FFT integration only contains the current symbol and cyclic shift sample values thereof, and inter-symbol crosstalk cannot be caused. If the timing deviation is negative, the symbol timing lags, and the OFDM window for FFT integration contains most samples of the current symbol and part of samples in the cyclic prefix of the next OFDM symbol, which may cause crosstalk between OFDM symbols and seriously affect the system performance.

Disclosure of Invention

In order to solve the above problems, the present invention provides a timing synchronization method for receiving end of WiMAX system base station based on OFDM, which comprises the following steps:

a. delaying a received signal after analog-to-digital conversion by a first fixed sample value, sending one path of signal to an energy window with the length of the first fixed sample value to solve energy, and taking the square of the energy; after the other path of signals is subjected to conjugation processing, the other path of signals and the original signals pass through a correlation window with the length being a first fixed sample value, then the square of a modulus value is obtained, the obtained square of the modulus value is divided by the square of an energy signal, and the initial synchronization timing point n1 of the first fixed sample value is preliminarily determined by searching the maximum value of the two paths of signals after division;

b. delaying the received signal after analog-to-digital conversion by a second fixed sample value, sending one path of the delayed signal to an energy window with the length of the second fixed sample value to solve energy, and taking the square of the energy; after the other path of signals is subjected to conjugation processing, the other path of signals and the original signals pass through a related window with the length of a second fixed sample value, then the square of a modulus value is obtained, the obtained square of the modulus value is divided by the square of an energy signal, and the initial synchronization timing point n2 of the second fixed sample value is preliminarily determined by searching the maximum value of the division of the two paths of signals; wherein the second fixed sample length is twice the length of the first fixed sample length.

c. Comparing (N1-N2) the length of the cyclic prefix CP, wherein the length of the CP is equal to the second fixed sample value, setting the synchronization timing point N1 to a final symbol timing synchronization instant N0 if the comparison result is 0, setting the final symbol timing synchronization instant to N0= (N1- "N/2") if the comparison result is positive N samples, "N/2" means taking the smallest integer equal to or greater than N/2, and setting N1 to the final symbol timing synchronization instant N0 if the comparison result is negative;

according to the invention, the first fixed sample is 128 samples and the second fixed sample is 256 samples.

According to the invention, the length of the energy window in step a is 128 samples, and the length of the correlation window is 128 samples.

According to the invention, the energy window in step b is CP length samples and the correlation window is CP length samples.

The synchronization method has the following advantages:

one of the advantages of the algorithm is that the traditional algorithm is emphasized to carry out timing synchronization according to the repeatability of 128 samples, and the CP cyclic prefix in front of the same OFDM symbol is used for further correction of the timing synchronization, so that the accuracy of the timing synchronization is improved, the influence of the frequency synchronization in the rear is reduced, the phase deflection angle is reduced, and the system performance is improved.

The invention has the second advantage that the condition that the timing deviation is negative is avoided as much as possible, and the serious reduction of the system performance caused by the symbol timing lag is eliminated.

In addition, the invention determines symbol timing by the ratio of the output signal energy of the correlator to the delayed signal energy, thus carrying out normalization processing on the energy of the output signal of the correlator, effectively avoiding overlarge output peak value of the correlator, reducing the influence on the actual system performance and improving the precision of synchronous timing.

Drawings

FIG. 1 is a block diagram of a WiMAX transceiver based on an OFDM mode;

FIG. 2 is a schematic diagram of an OFDM symbol;

FIG. 3 is a diagram of a frame short preamble time domain structure;

FIG. 4 is a general flow of WiMAX base station receive end synchronization;

fig. 5 is a schematic diagram of a WiMAX base station frame synchronization algorithm based on OFDM.

Detailed Description

In an actual WiMAX system, the synchronization procedure of the base station is generally: after performing analog-to-digital conversion (ADC) on a received signal, a base station performs frame-symbol coarse synchronization to determine OFDM symbols and initial sampling points of frames; then, frequency offset coarse synchronization and correction are carried out on the basis of timing coarse synchronization; then FFT transform is carried out, then frequency offset, sample value fine synchronization and correction are carried out in a frequency domain, and finally fine synchronization of accurate symbol sample value timing is carried out on data symbols after frequency offset correction and channel estimation. The invention mainly aims at the rough frame timing synchronization design of the receiving end of the WiMAX system base station based on OFDM, and provides a preferred method for carrying out the frame timing synchronization based on the short preamble. The design schematic is shown in fig. 5:

the delay D1 is 128 samples, the delay D2 is 256 lengths, the length of the correlation window L1 is 128, the length of the correlation window L2 is CP length, the setting is carried out according to the system requirement, the energy window is used for calculating the instantaneous energy of the delay signal, and the output M (M) is expressed as:

wherein

Wherein

The starting point n1 of the first 128 sample of the preamble code can be determined by searching the maximum value of the output M1 (M), the starting point n2 of the cyclic prefix of the preamble code can be determined by searching the maximum value of the output M2 (M), the comparator is used for comparing the difference value of n1 and n2, and the timing synchronization point n0 of the preamble code is reasonably set according to the relation between the size of the difference value and the length of the CP. In actual system design, modules such as a correlation window L1 (128 samples), L2 (CP length set by a base station), an energy window P1 (128 samples), P2 (CP length set by the base station), and the like, which are involved in the schematic diagram, may perform time division processing by using the same hardware. Since the CP occurs at the beginning of each OFDM symbol and the 128-repeated samples occur only in the Burst preamble of each SS, a plurality of cyclic prefix correlator peaks may occur in an uplink subframe, where the starting point n2 of the determined CP should be the timing synchronization point determined by the first correlation peak occurring before the synchronization timing point n1.

The following is a detailed description with reference to the schematic diagram shown in fig. five.

Step (one), the WiMAX base station performs D/A conversion to the received signal sent by the SS to obtain a signal sequence r (m), the signal delays D1 sample values through a register, namely delays 128 sample values, one path of the signal is subjected to conjugate conversion and then is subjected to correlation operation with the original signal, and a window of a correlator is subjected to 128 operation to obtain a correlated signal c1 (m). And the other path of signal is subjected to energy summation, an energy window P1 takes a value of 128 to obtain a signal P1 (M), c1 (M) is subjected to modulus value square taking and then is divided by the square of the energy signal P1 (M) to obtain a ratio sequence M1 (M), and the moment when the maximum value is obtained by searching M1 (M) is preliminarily judged to be the starting moment n1 of a 128 sample value. Since the two 128 samples before and after the preamble symbol have the same value, the maximum value can be obtained after the correlation, and the correlation of the 128 samples may also be affected by noise with a certain error. The correlation of the other samples is small and the output after passing through the correlator is some random noise sequence.

And (II) repeating the process of the step (I), delaying the received signal r (M) by D2 sample values, namely delaying 256 sample values, taking the window value of the correlator as the CP length of the cyclic prefix set by the system, taking the energy window P2 as the CP length of the cyclic prefix, obtaining the maximum value by searching and outputting the time of M2 (M), and preliminarily judging that the timing synchronization point determined by the first correlation peak appearing in front of the timing synchronization point of n1 is the starting time n2 of the preamble cyclic prefix. Although the sample sequence in the cyclic prefix is influenced by the multipath delay, the cyclic prefix is the repetition of the last CP length sample of the OFDM symbol, and despite the influence of the multipath delay, the correlation is still higher than that of other random noises, and the time when the c2 (m) module value output by the correlator reaches the maximum value still has a relatively high probability of being the prefix starting time n2. Because the processing procedures of the step (ii) and the step (i) are basically the same, in a specific implementation procedure, the same hardware may be used for time division processing.

And thirdly, subtracting the 128 sample value repetition time n1 which is preliminarily judged from the prefix initial time n2, comparing the result with the CP length, and reasonably setting the initial time of the preamble symbol according to a certain error principle according to different comparison results so as to finish synchronous timing. Here we use the following timing scheme. Although 256-point IFFT transformation is performed on a received OFDM symbol to modulate 256 subcarriers, some samples may be affected by random noise in a channel with frequency selective fading, but the probability that the determined timing synchronization point occurs at the point where the output M (M) has the maximum value is much higher than the probability at other points, and the probability around the optimal timing point is significantly reduced.

If the value is 0 after comparison, it means that the timing synchronization judged by 128 samples is basically consistent with the timing synchronization judged by the correlator with the CP length, and according to the correlation principle of probability theory, the probability of the maximum value of output M (M) appearing in the judgment of two timing synchronization points is much higher than the probability of the maximum value appearing in other points, so we can set the final symbol timing synchronization time n0= n1;

if after the comparison, the value is positive N samples, the judged synchronization time N1 lags behind N2. According to the relevant principle of probability theory, in the case of neglecting the occurrence of small probability events, the events occurring in this case are mainly possible as follows: firstly, n1 is accurate in timing, and n2 is advanced in timing; secondly, n2 is accurate in timing, and n1 is delayed in timing; for other possible occurrences, it may be left out of consideration because of the relatively low probability of occurrence. Then, the final symbol timing synchronization time N0= (N1- "N/2"), where "N/2" represents the smallest integer equal to or greater than N/2. Of course, for the sake of safety, the final symbol timing synchronization time N0= (N1-N) may be set, and the purpose of this setting is mainly to reduce the probability that the timing synchronization time lags the optimal timing synchronization time, and to avoid the serious influence of the inter-symbol crosstalk caused thereby on the system.

If the value is negative, the judged synchronization time n1 is advanced by n2, and according to the analysis of the probability theory and considering that the cyclic prefix is easily influenced by the multipath delay, the precision of the symbol timing synchronization point n2 determined based on the correlation peak determined by the cyclic prefix is lower than that of the symbol timing synchronization point n1 determined based on the correlation peak of 128 cyclic sample values. Therefore, we set the final symbol timing synchronization time n0= n1, and this setting is still for the purpose of reducing the probability that the timing synchronization time lags the optimal timing synchronization time, and avoiding the serious influence of the inter-symbol crosstalk caused thereby on the system.

According to the finally determined timing synchronization time n0, the coarse synchronization of the frame and symbol timing is initially completed, and the subsequent frequency offset estimation and correction can be continued. For further accurate timing synchronization correction, the fine frame timing synchronization algorithm after the synchronization process of the receiving end of the WiMAX system can be used for correction.

Various modifications and alterations of this invention will become apparent to those skilled in the art without departing from the spirit and scope of this invention. It is therefore intended that the present invention cover the modifications and variations of this invention provided they come within the scope of the appended claims and their equivalents.

Claims

1. A timing synchronization method for receiving end of WiMAX system base station based on OFDM includes following steps:

a. delaying a first fixed sample value of a received signal after analog-to-digital conversion, sending one path of the delayed signal to a first energy window with the length of the first fixed sample value for energy summation to obtain a first energy signal, and solving the square of the first energy signal; after conjugate processing, the other path of signals and the original signals pass through a first correlation window with the length of a first fixed sample value, then the module value square is calculated, the module value square of the signals after passing through the first correlation window is divided by the square of a first energy signal to obtain a first ratio sequence, and the starting synchronization timing point n1 of the first fixed sample value is preliminarily determined by searching the moment when the first ratio sequence obtains the maximum value;

b. delaying a received signal subjected to analog-to-digital conversion by a second fixed sample value, sending one path of the delayed signal to a second energy window with the length of the second fixed sample value for energy summation to obtain a second energy signal, and solving the square of the second energy signal; the other path of the signals after conjugation processing and the original signals pass through a second correlation window with the length of a second fixed sample value, then the module value square is calculated, the module value square of the signals after passing through the second correlation window is divided by the module value square of a second energy signal to obtain a second ratio sequence, and the starting synchronization timing point n2 of the second fixed sample value is preliminarily determined by searching the moment when the second ratio sequence obtains the maximum value;

c. comparing the lengths of (N1-N2) and a Cyclic Prefix (CP), wherein the length of the CP is equal to the second fixed sample, if the comparison result is 0, then setting sync timing point (N1) to the final symbol timing sync instant (N0- "N/2") if the comparison result is positive N samples, setting the final symbol timing sync instant (N0- "(N1-" N/2 ")," "N/2" means taking the smallest integer equal to or greater than N/2, and if the comparison result is negative, setting N1 to the final symbol timing sync instant (N0);

wherein the second fixed sample length is twice the length of the first fixed sample length.

2. The timing synchronization method of claim 1, wherein said first fixed sample is 128 samples and said second fixed sample is 256 samples.