CN103490860B - A kind of high performance shortwave orthogonal spectrum expansion iterative demodulation and interpretation method - Google Patents

Abstract

A high-performance short-wave orthogonal spread spectrum iterative demodulation and decoding method, including sampling the received signal at a chip rate after passing through a matched filter, correlating the sampled data with the local orthogonal spread spectrum sequence, and using the correlation value to perform non-coherent The demodulated output bit soft value; the bit soft value is used as the input of high-performance coding and decoding, and iterative decoding is performed; the decoding and decision output bits are used to reconstruct the orthogonal spread spectrum modulation signal, and the channel estimation of the path is realized; the path is restored by channel estimation Phase information, to achieve coherent demodulation of orthogonal spread spectrum; according to this process, iterative demodulation and decoding, until the iteration stops, output information bits. This method proposes the iterative demodulation and decoding structure of the orthogonal spread spectrum system; improves the credibility of the demodulation bit soft value, and fully utilizes the performance of channel coding and decoding; at the same time, the iterative process further improves the accuracy of channel estimation, positive Cross-spread demodulation performance and decoding performance, thereby improving the transmission performance of the existing short-wave communication system.

Description

A High Performance HF Orthogonal Spread Spectrum Iterative Demodulation and Decoding Method

technical field

The invention relates to a reliable short-wave data communication transmission technology, in particular to a high-performance orthogonal spread spectrum iterative demodulation and decoding method.

Background technique

The short-wave channel presents the characteristics of multipath fading, and the spread spectrum technology can effectively resist the influence of fading, and has certain anti-interference ability. Compared with direct sequence spread spectrum, multi-ary orthogonal spread spectrum can avoid the limitation of bandwidth on binary numbers, and can support high-ary numbers without additional insertion of cyclic prefix or training sequence, improving transmission efficiency, and according to digital communication theory, When the binary number of multi-ary orthogonal spread spectrum modulation tends to infinity, the performance close to Shannon's limit signal-to-noise ratio can be obtained. Therefore, multi-ary system orthogonal spread spectrum is one of the key technologies to realize the reliable transmission of shortwave data.

In order to improve the transmission performance of the orthogonal spread spectrum system, an iterative decoding and demodulation method of joint channel estimation is proposed. However, since the multi-ary orthogonal spread spectrum system does not have a dedicated pilot channel similar to the mobile communication CDMA system, and the short-wave channel is a typical For time-varying fading channels, the traditional demodulation method often adopts non-coherent demodulation, which reduces the reliability of bit soft values compared with coherent demodulation; in iterative demodulation and decoding, the modulation positive is reconstructed through decision feedback The cross-spread signal is used for channel estimation, and considering the time-varying nature of the channel, the channel estimated by the two orthogonal spread-spectrum symbols before and after is used as the channel estimation in the current symbol period, which improves the channel estimation accuracy and realizes coherent demodulation. Lay a solid foundation. In addition, soft input is one of the basic characteristics of high-performance channel coding and decoding, and hard-decision demodulation greatly limits the decoding performance. Therefore, a bit soft value method for coherent demodulation of orthogonal spread spectrum system is given to further enhance The credibility of the bit soft value is guaranteed, and the coding gain of the high-performance channel coding and decoding technology is further fully utilized.

Contents of the invention

The purpose of the present invention is to further improve the transmission performance of the orthogonal spread spectrum system, design a joint channel estimation iterative demodulation and decoding structure, further improve the channel estimation accuracy and the credibility of the bit soft value, and give full play to the high-performance channel coding and decoding The coding gain of the technology, so as to realize the reliable transmission of short-wave data communication.

Technical scheme of the present invention is:

A high-performance shortwave orthogonal spread spectrum iterative demodulation and decoding method, which uses a joint channel estimation part, orthogonal spread spectrum demodulation part and channel decoding part to perform iterative processing to achieve the purpose of improving shortwave communication transmission performance , including the following steps:

Step 1, the receiving end receives the spread spectrum signal and performs matched filtering, and obtains a sampling sequence of each symbol period of the received signal by sampling;

Step 2, correlating the obtained sampling sequence with the local spread spectrum sequence to obtain a correlation value sequence;

Step 3, using the correlation value sequence to perform orthogonal spread spectrum non-coherent demodulation to obtain the orthogonal spread spectrum non-coherent demodulation bit soft value of each symbol period;

Step 4, using the obtained bit soft value as the input of channel decoding, performing Turbo Product Code TPC or Low Density Parity Check Code LDPC iterative decoding to obtain the decoding decision output bits;

Step 5: decoding and judging the output bits and then performing reconstruction modulation to generate orthogonal spread spectrum modulation symbols for each symbol period, correlating the orthogonal spread spectrum modulation symbols with the receiving sequence of each symbol period to obtain the path response; and using the before and after The path response of each M symbols, estimating the path response of the current symbol;

Step 6: Carry out coherent demodulation of orthogonal spread spectrum according to the phase in the estimated path response result, and obtain the bit soft value of non-coherent demodulation of orthogonal spread spectrum this time; since estimation of the path response of the current symbol needs to use multiple symbols before and after, Then the coherent demodulation cannot be realized for the M symbols before and after one frame of symbols, so use the non-coherent demodulation bit soft values of the 2M symbols at both ends and the coherent demodulated bit soft values of the intermediate symbols as the channel decoding input; Step 4 performs channel decoding, iterates at least once, and outputs a decoding decision result after the iteration stops.

Method of the present invention specifically comprises the following steps:

Step 1: The transmitting end selects a spreading sequence from a total of L orthogonal spreading sequence sets for every K bits, where L=2 ^K , and then performs shaping filtering for transmission; the receiving end performs matching filtering after receiving the signal, And perform sampling and extraction at the chip rate 1/T _c to obtain the sampling sequence R _n of the nth symbol period

R _n =[r(nT), r(nT+T _c ), ..., r(nT+(N-1)T _c )]

Among them: r(nT) represents the first sampling point in the nth symbol period, r(nT+(N-1)T _c ) represents the Nth sampling point in the nth symbol period; N is the length of the spreading code, and the spreading The frequency code symbol period is T=NT _c ;

Step 2: Correlate the obtained sampling sequence R _n with the local spread spectrum sequence W _l to obtain the correlation value z _n,l ,

${\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; l</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}{\mathrm{<span\; class="notranslate">\; W</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}}^{\mathrm{<span\; class="notranslate">\; h</span>}}$

Among them, l represents the local spread spectrum sequence number, 1≤l≤L, Represents the conjugate transpose of the local spreading sequence W _l ;

Sampling sequence R _n is correlated with L local spread spectrum sequences to obtain a correlation value sequence Z _n =[z _n,1 ,z _n,2 ,…,z _n,L ];

Step 3, using Z _n to perform orthogonal spread spectrum non-coherent demodulation: the correlation value is divided into two subsets according to the kth bit data of the corresponding spread spectrum sequence sent as "1" and "0" The soft value of spread spectrum bits obtained by non-coherent demodulation is expressed as:

$\mathrm{<span\; class="notranslate">\; \&Lambda;</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; b</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \&Element;</span>{\mathrm{<span\; class="notranslate">\; S</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}^{<span\; class="notranslate">\; +</span>}}{\mathrm{<span\; class="notranslate">\; max</span>}}{<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; |</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; -</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \&Element;</span>{\mathrm{<span\; class="notranslate">\; S</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}^{<span\; class="notranslate">\; -</span>}}{\mathrm{<span\; class="notranslate">\; max</span>}}{<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; |</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; \&le;</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; \&le;</span>\mathrm{<span\; class="notranslate">\; K</span>}$

Among them, K is the number of bits modulated corresponding to each spread spectrum symbol, b _{n, k} is the kth bit data demodulated in the nth symbol period, represent a subset The maximum value of the modulo square of the correlation value in represent a subset The maximum value of the modulo square of the correlation value in ;

And calculate the bit soft value of demodulation of all symbols in a frame of data by this calculation method, if a frame of data contains F orthogonal spread spectrum symbols, then the total number of bits of a frame of data is F K;

Step 4: Take the soft bit value of one frame of data obtained as the input of channel decoding, perform TPC/LDPC iterative decoding, and output a total of F·K bits of data {b _n,k } from the decoding decision, where b _n,k represent The kth bit data of the nth symbol period;

Step 5: Decision feedback outputs all hard decision bits {b _n,k }, and reconstructs modulation to generate F orthogonal spread spectrum modulation symbols as Among them, for the K-bit data {b _n,1 ,b _n,2 ,…,b _n,K } of the nth symbol period, the orthogonal spread spectrum symbols generated by reconstruction modulation are Will Correlate with the received sequence R _n of the nth symbol period to obtain the path response h _n ;

${\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}{\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; W</span>}}}_{\mathrm{<span\; class="notranslate">\; no</span>}}^{\mathrm{<span\; class="notranslate">\; h</span>}}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; N</span>}<span\; class="notranslate">\; ;</span>$

Among them, R _n is the sampling sequence of the nth symbol period, Indicates that the hard decision bit reconstructs the orthogonal spread spectrum symbol in the nth symbol period The conjugate transpose of; N is the spreading code length;

Then the path response in the n-2th symbol period is Similarly, the path response of the preceding and following 2M symbol periods can be obtained as h _nM ,...,h _n-1 ,h _n+1 ,...,h _n+M , and the path response of the nth symbol period is estimated by using the 2M path responses for:

${\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; m</span>}}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; \&Center\; Dot;</span><span\; class="notranslate">\; \&Center\; Dot;</span><span\; class="notranslate">\; \&Center\; Dot;</span><span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; \&Center\; Dot;</span><span\; class="notranslate">\; \&Center\; Dot;</span><span\; class="notranslate">\; \&Center\; Dot;</span><span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; m</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; /</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; \&alpha;</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}\mathrm{<span\; class="notranslate">\; exp</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; j</span>}{\mathrm{<span\; class="notranslate">\; \&phi;</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; )</span>$

Among them, M represents the number of symbols before and after the joint channel estimation; α _n is the path gain, and φ _n is the path phase;

Step 6, use the phase information φ _n restored by channel estimation to realize coherent demodulation of orthogonal spread spectrum, first compensate the phase influence of the channel, and then take the real part to express as The coherent demodulation output bit soft value is:

$\mathrm{<span\; class="notranslate">\; \&Lambda;</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; b</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \&Element;</span>{\mathrm{<span\; class="notranslate">\; S</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}^{<span\; class="notranslate">\; +</span>}}{\mathrm{<span\; class="notranslate">\; max</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}^{\mathrm{<span\; class="notranslate">\; R</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \&Element;</span>{\mathrm{<span\; class="notranslate">\; S</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}^{<span\; class="notranslate">\; -</span>}}{\mathrm{<span\; class="notranslate">\; max</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}^{\mathrm{<span\; class="notranslate">\; R</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; \&le;</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; \&le;</span>\mathrm{<span\; class="notranslate">\; K</span>}$

Coherent demodulation of the current symbol requires multiple symbols before and after the channel estimation. It can be seen that the M symbols before and after a frame symbol cannot be coherently demodulated, and the non-coherent demodulation bit soft value of the 2M symbols at both ends is coherent with the middle symbol Demodulate a total of F·K bit soft values as channel decoding input; then turn to the fourth step for channel decoding until the iteration stops, and finally output the decision data bits b _n,k .

In the present invention, according to the delay requirement of the transmission service, an interleaver can be added between channel coding and modulation, effectively improving the system's ability to resist burst errors, and further exerting the performance of channel decoding. The interleaving methods include row-column interleaving and spiral interleaving. , random interleaving, etc.

When the system joins the interleaver step,

In step 4, using the obtained soft data bit values of one frame as the input of channel decoding is specifically implemented as deinterleaving the obtained soft data bit values of one frame of data as input of channel decoding.

In Step 5, the decision feedback outputs all hard decision bits, and the reconstruction modulation generates F orthogonal spread spectrum modulation symbols. The specific implementation is that all hard decision bits output by decision feedback are interleaved before performing reconstruction modulation Generate F orthogonal spread spectrum modulation symbols.

In step 6, the soft values of F·K bits in total are used as input for channel decoding, and the specific realization is that the soft values of F·K bits in total must also be deinterleaved before they can be input as channel decoding.

The implementation method of step 4 of the present invention is as follows: Utilize demodulation output soft value, carry out TPC/LDPC iterative decoding, its decoding method adopts general TPC or LDPC decoding algorithm, after iterative decoding bit soft value is judged, output hard decision Bit b _n,k .

The beneficial effects of the present invention are:

The present invention proposes an iterative decoding and demodulation method for joint channel estimation for orthogonal spread spectrum systems, performs channel estimation by reconstructing and modulating orthogonal spread spectrum signals through decision feedback, and designs an accurate estimation method for the channel response of the current symbol, The method of bit soft value for coherent demodulation of orthogonal spread spectrum system is given, which enhances the reliability of bit soft value and gives full play to the coding gain of high performance channel coding technology. At the same time, joint channel estimation, demodulation and decoding iterations further improve the performance of short-wave communication data transmission.

Description of drawings

Figure 1 is a schematic diagram of the joint iterative demodulation and decoding structure of the orthogonal spread spectrum system.

Figure 2 is a block diagram of the transmission structure of the orthogonal spread spectrum system.

Fig. 3 is a schematic diagram of the structure of coherent demodulation of orthogonal spread spectrum.

FIG. 4 is a schematic diagram of the performance of iterative demodulation and decoding.

detailed description

The present invention designs a joint channel estimation iterative demodulation and decoding structure, further improves the channel estimation accuracy and the reliability of bit soft values, and fully utilizes the coding gain of high-performance channel coding and decoding technology, thereby enhancing the reliability of short-wave data transmission . Concrete steps of the present invention are as follows:

Step 1: The transmitting end selects a spreading sequence from a total of L sets of orthogonal spreading codes for every K bits, where L=2 ^K , and then performs shaping filtering for transmission; the receiving end performs matching filtering on the received signal, and uses The chip rate is 1/T _c extraction, assuming that the receiver has been synchronized, the sampling sequence of the nth symbol period is

R _n =[r(nT), r(nT+T _c ), ..., r(nT+(N-1)T _c )]

Among them: r(nT) represents the first sampling point in the nth symbol period, r(nT+(N-1)T _c ) represents the Nth sampling point in the nth symbol period; N is the length of the spreading code, and the spreading The frequency code symbol period is T=NT _c ;

Assuming that the transmitted spread spectrum signal is Wu _u , the subscript u represents the sequence number of the spread spectrum; the path channel gain is α _n , and the phase deflection is φ _n , then the sampling sequence of the nth symbol period can be expressed as R _n =α _n exp (jφ _n )W _u +n, n represents complex Gaussian white noise.

Step 2: Correlate the obtained sampling sequence R _n with the local spread spectrum sequence W _l to obtain the correlation value z _n,l ,

${\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; l</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}{\mathrm{<span\; class="notranslate">\; W</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}}^{\mathrm{<span\; class="notranslate">\; h</span>}}$

Among them, l represents the local spread spectrum sequence number, 1≤l≤L, Represents the conjugate transpose of the local spreading sequence W _l ;

Sampling sequence R _n is correlated with L local spread spectrum sequences to obtain a correlation value sequence Z _n =[z _n,1 ,z _n,2 ,…,z _n,L ], and it can be seen that the correlation value sequence is a complex number;

Step 3, due to the time-varying characteristics of the short-wave channel, the accurate channel estimation of the current symbol cannot be obtained according to the synchronization sequence. Therefore, the first orthogonal spread spectrum demodulation directly uses Z _n for non-coherent demodulation; the correlation value is transmitted according to the corresponding spread spectrum The kth bit data of the sequence is "1", "0" is divided into two subsets The soft value of spreading bits for non-coherent demodulation is expressed as:

$\mathrm{<span\; class="notranslate">\; \&Lambda;</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; b</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \&Element;</span>{\mathrm{<span\; class="notranslate">\; S</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}^{<span\; class="notranslate">\; +</span>}}{\mathrm{<span\; class="notranslate">\; max</span>}}{<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; |</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; -</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \&Element;</span>{\mathrm{<span\; class="notranslate">\; S</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}^{<span\; class="notranslate">\; -</span>}}{\mathrm{<span\; class="notranslate">\; max</span>}}{<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; |</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; \&le;</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; \&le;</span>\mathrm{<span\; class="notranslate">\; K</span>}$

Among them: K represents the number of bits demodulated by an orthogonal spread spectrum symbol; b _n,k is the kth bit data demodulated for the nth symbol period, represent a subset The maximum value of the modulo square of the correlation value in represent a subset The maximum value of the modulo square of the correlation value in ;

And calculate the bit soft value of demodulation of all symbols in a frame of data by this calculation method, if a frame of data contains F orthogonal spread spectrum symbols, then the total number of bits of a frame of data is F K;

Step 4: de-interleave the soft bit values of one frame of data obtained, and use it as the input of channel decoding to perform TPC/LDPC iterative decoding, and output a total of F·K bits of data {b _n,k } for the decoding decision, where b _{n, k} represent the kth bit data of the nth symbol period;

Step 5, the decision feedback outputs all the hard decision bits {b _n,k } After interleaving, the reconstruction modulation generates F orthogonal spread spectrum modulation symbols as Among them, for the K-bit data {b _n,1 ,b _n,2 ,…,b _n,K } of the nth symbol period, the orthogonal spread spectrum symbols generated by reconstruction modulation are Will Correlation with the sampling sequence R _n of the nth symbol period to obtain the path response Indicates that the nth symbol period hard-decision bit reconstructed orthogonal spread spectrum symbol The conjugate transpose of ;

When the decision feedback output is correct, the reconstructed orthogonal spread spectrum symbol is the transmitted spread spectrum symbol, regardless of the influence of noise, then h _n =α _n exp(jφ _n ), it can be seen that the decision reconstruction can obtain the path response.

Misjudgment will lead to wrong estimation of path response. In order to improve the accuracy of path response estimation, multiple symbols before and after are used to estimate path response, and the appropriate number of symbols is selected according to the time-varying characteristics of the channel. In addition, when estimating the path response of the current symbol, to prevent the influence of the wrong decision in the previous iteration, instead of using the current path response estimate h _n , use the path response h _nM ,...,h _n-1 , h _n estimated by M symbol periods before and after ₊₁ ,…,h _n+M , estimate the path response of the nth symbol period as:

${\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; m</span>}}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; \&Center\; Dot;</span><span\; class="notranslate">\; \&Center\; Dot;</span><span\; class="notranslate">\; \&Center\; Dot;</span><span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; \&Center\; Dot;</span><span\; class="notranslate">\; \&Center\; Dot;</span><span\; class="notranslate">\; \&Center\; Dot;</span><span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; m</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; /</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; \&alpha;</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}\mathrm{<span\; class="notranslate">\; exp</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; j</span>}{\mathrm{<span\; class="notranslate">\; \&phi;</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; )</span>$

Among them: M represents the number of symbols before and after the joint channel estimation; α _n is the path gain, and φ _n is the path phase information.

Step 6: use channel estimation to restore phase information φ _n to realize coherent demodulation of orthogonal spread spectrum, first compensate the phase influence of the channel, and then take the real part to express as The coherent demodulation output bit soft value is:

$\mathrm{<span\; class="notranslate">\; \&Lambda;</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; b</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \&Element;</span>{\mathrm{<span\; class="notranslate">\; S</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}^{<span\; class="notranslate">\; +</span>}}{\mathrm{<span\; class="notranslate">\; max</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}^{\mathrm{<span\; class="notranslate">\; R</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \&Element;</span>{\mathrm{<span\; class="notranslate">\; S</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}^{<span\; class="notranslate">\; -</span>}}{\mathrm{<span\; class="notranslate">\; max</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}^{\mathrm{<span\; class="notranslate">\; R</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; \&le;</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; \&le;</span>\mathrm{<span\; class="notranslate">\; K</span>}$

Coherent demodulation of the current symbol needs to use multiple symbol path responses before and after. It can be seen that the M symbols at the front and rear ends of a frame symbol cannot be coherently demodulated, and the non-coherent demodulation bit soft value of the 2M symbols at both ends and the middle symbol The coherently demodulated bit soft value is deinterleaved and used as channel decoding input; then turn to the fourth step for channel decoding until the iteration stops, and finally output the decision data bits b _n,k .

Embodiment of the invention

Taking the shortwave 24kHz signal bandwidth as an example, the chip width T _c =1/19200 seconds, the orthogonal spread spectrum code adopts the Hadamard sequence with a length of 256, and each spread spectrum modulation symbol transmits log ₂ (256)=8 bits; channel coding The code is a Turbo product code (32,26)×(32,26), and other high-performance channel codecs can also be used in practical applications; the entire frame signal in this embodiment is a TPC(32,26)×(32,26) array, so when the interleaver is not used, the specific steps are as follows:

Step 1, the sending end has a total of 676 bits of information, after performing TPC encoding to obtain 1024 bits of data, every 8 bits are modulated to generate 1 orthogonal spread spectrum symbol, then a total of 1024 bits are modulated to generate 128 orthogonal spread spectrum symbols, and then shaping filtering is performed Sending; at the receiving end, the received signal is matched and filtered, and extracted at the chip rate 1/T _c , assuming that the receiver has been synchronized, the sampling sequence of the nth symbol period is R _n =[r(nT),r(nT +T _c ),...,r(nT+255 T _c )], the spreading code symbol period is T=256 T _c ;

Step 2, correlate the sampling sequence with the local spread spectrum sequence, and get The subscript l represents the local spread spectrum sequence number, and the superscript H represents the conjugate transpose. There are 256 local spread spectrum sequences in total, and the correlation value sequence obtained is Z _n =[z _n,1 ,z _n,2 ,…, z _n,256 ];

Step 3, using Z _n to perform non-coherent demodulation by orthogonal spread spectrum. The correlation value is divided into two subsets with a size of 128 according to the kth bit data of the corresponding spreading sequence being sent as "1" and "0". The table gives an example of set division where the second bit is "1" or "0".

The soft value of spreading bits for non-coherent demodulation is expressed as:

$\mathrm{<span\; class="notranslate">\; \&Lambda;</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; b</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \&Element;</span>{\mathrm{<span\; class="notranslate">\; S</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}^{<span\; class="notranslate">\; +</span>}}{\mathrm{<span\; class="notranslate">\; max</span>}}{<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; |</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; -</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \&Element;</span>{\mathrm{<span\; class="notranslate">\; S</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}^{<span\; class="notranslate">\; -</span>}}{\mathrm{<span\; class="notranslate">\; max</span>}}{<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; |</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; \&le;</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; \&le;</span>\mathrm{<span\; class="notranslate">\; 8</span>}$

b _n,k is the kth bit data demodulated in the nth symbol period.

Step 4, every 128 orthogonal spread spectrum symbols are demodulated and output 128*8=1024 soft values, which are used as the input of channel decoding, and TPC(32,26)×(32,26) iterative decoding is performed, considering the operation amount, TPC internal iterative decoding 8 times, decoding judgment output 1024-bit data;

Step 5: Use the decision feedback to output 1024-bit hard decision information, and reconstruct the modulation to generate 128 orthogonal spread spectrum modulation symbols. The symbols are counted from the first one, and a frame of symbols is recorded as Then the corresponding receiving sequence is {R ₁ ,R ₂ ,…,R ₁₂₈ }, and the path response corresponding to each symbol is calculated, 1≤n≤128; in order to improve the accuracy of channel estimation and consider the time-varying characteristics of the channel, the path response of the current symbol period is estimated by using the path responses corresponding to the two symbols before and after:

${\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 4,3</span>}<span\; class="notranslate">\; \&le;</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; \&le;</span>\mathrm{<span\; class="notranslate">\; 126</span>}$

It can be seen from the above formula that an accurate channel response estimate for 124 symbols is obtained, and symbols 1, 2, 127, and 128 do not give accurate estimates. Using the estimated channel response, the phase information φ _n is recovered:

${\mathrm{<span\; class="notranslate">\; \&phi;</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; the\; angle</span>}<span\; class="notranslate">\; (</span>{\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; 3</span>}<span\; class="notranslate">\; \&le;</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; \&le;</span>\mathrm{<span\; class="notranslate">\; 126</span>}$

In the formula, angle(·) is a function to take complex phase.

Step 6: use channel estimation to restore phase information φ _n to realize coherent demodulation of orthogonal spread spectrum, first compensate the phase influence of the channel, and then take the real part to express as 3≤n≤126, 1≤l≤256, the coherent demodulation output bit soft value is:

$\mathrm{<span\; class="notranslate">\; \&Lambda;</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; b</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \&Element;</span>{\mathrm{<span\; class="notranslate">\; S</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}^{<span\; class="notranslate">\; +</span>}}{\mathrm{<span\; class="notranslate">\; max</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}^{\mathrm{<span\; class="notranslate">\; R</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \&Element;</span>{\mathrm{<span\; class="notranslate">\; S</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}^{<span\; class="notranslate">\; -</span>}}{\mathrm{<span\; class="notranslate">\; max</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}^{\mathrm{<span\; class="notranslate">\; R</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; \&le;</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; \&le;</span>\mathrm{<span\; class="notranslate">\; 8</span>}$

Then one frame of symbol coherent demodulation outputs 124*8=992 bit soft values, and symbols 1, 2, 127, 128 use 4*8=32 bit soft values of non-coherent demodulation as channel decoding input; Enter the fourth step for channel decoding until the iteration stops, and finally output the decision data bits.

Figure 4 shows the performance curve of the joint iterative demodulation and decoding method under the condition of additive Gaussian white noise. The internal iteration of TPC decoding is 8 times, and the overall iteration of joint channel estimation, orthogonal spread spectrum demodulation and channel decoding is 2 It can be seen that the iterative demodulation and decoding method proposed by the present invention can effectively improve the transmission performance of the system. When the bit error rate is 1E-5, compared with the traditional non-coherent demodulation reception, it can obtain a signal of about 0.4dB. Noise Ratio Gain.

Claims (4)

1. A high-performance shortwave orthogonal spread spectrum iterative demodulation and decoding method is characterized in that the method adopts a joint channel estimation part, an orthogonal spread spectrum demodulation part and a channel decoding part to carry out iterative processing to achieve the improvement of shortwave The purpose of communication transmission performance, including the following steps: Step 1, the receiving end receives the spread spectrum signal and performs matched filtering, and obtains a sampling sequence of each symbol period of the received signal by sampling; Step 2, correlating the obtained sampling sequence with the local spread spectrum sequence to obtain a correlation value sequence; Step 3, using the correlation value sequence to perform orthogonal spread spectrum non-coherent demodulation to obtain the orthogonal spread spectrum non-coherent demodulation bit soft value of each symbol period; Step 4, using the obtained bit soft value as the input of channel decoding, performing Turbo Product Code TPC or Low Density Parity Check Code LDPC iterative decoding to obtain the decoding decision output bits; Step 5: decoding and judging the output bits and then performing reconstruction modulation to generate orthogonal spread spectrum modulation symbols for each symbol period, correlating the orthogonal spread spectrum modulation symbols with the receiving sequence of each symbol period to obtain the path response; and using the before and after The path response of each M symbols, estimating the path response of the current symbol; Step 6: Carry out coherent demodulation of orthogonal spread spectrum according to the phase in the estimated path response result, and obtain the bit soft value of non-coherent demodulation of orthogonal spread spectrum this time; since estimation of the path response of the current symbol needs to use multiple symbols before and after, Then the coherent demodulation cannot be realized for the M symbols before and after one frame of symbols, so use the non-coherent demodulation bit soft values of the 2M symbols at both ends and the coherent demodulated bit soft values of the intermediate symbols as the channel decoding input; Step 4: Perform channel decoding, iterate at least once, and output the decoding judgment result after the iteration stops; The method specifically includes the following steps: Step 1: The transmitting end selects a spreading sequence from a total of L orthogonal spreading sequence sets for every K bits, where L=2 ^K , and then performs shaping filtering for transmission; the receiving end performs matching filtering after receiving the signal, And carry out sampling extraction with chip rate 1/T _c , obtain the sampling sequence R _n of the nth symbol period, R _n adopts the following formula to express; R _n =[r(nT),r(nT+T _c ),...,r(nT+(N-1)T _c )] Among them: r(nT) represents the first sampling point in the nth symbol period, r(nT+(N-1)T _c ) represents the Nth sampling point in the nth symbol period; N is the length of the spreading code, and the spreading The frequency code symbol period is T=NT _c ; Step 2: Correlate the obtained sampling sequence R _n with the local spread spectrum sequence W _l to obtain correlation values z _n , _l , z _n,l = R _n W _l ^H Among them, l represents the local spread spectrum sequence number, 1≤l≤L, W _l ^H represents the conjugate transposition of the local spread spectrum sequence W _l ; Sampling sequence R _n is correlated with L local spread spectrum sequences to obtain correlation value sequence Z _n =[z _n,1 ,z _n,2 ,…,z _n,L ]; Step 3, using Z _n to perform orthogonal spread spectrum non-coherent demodulation: the correlation value is divided into two subsets according to the kth bit data of the corresponding spread spectrum sequence sent as "1" and "0" The soft value of spread spectrum bits obtained by non-coherent demodulation is expressed as: $\mathrm{<span\; class="notranslate">\; \&Lambda;</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; b</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \&Element;</span>{\mathrm{<span\; class="notranslate">\; S</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}^{<span\; class="notranslate">\; +</span>}}{\mathrm{<span\; class="notranslate">\; m</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; x</span>}}{<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; |</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; -</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \&Element;</span>{\mathrm{<span\; class="notranslate">\; S</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}^{<span\; class="notranslate">\; -</span>}}{\mathrm{<span\; class="notranslate">\; m</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; x</span>}}{<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; |</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; \&le;</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; \&le;</span>\mathrm{<span\; class="notranslate">\; K</span>}$ Among them, K is the number of bits modulated corresponding to each spread spectrum symbol, b _{n, k} is the kth bit data demodulated in the nth symbol period, represent a subset The maximum value of the modulo square of the correlation value in represent a subset The maximum value of the modulo square of the correlation value in ; And calculate the bit soft value of demodulation of all symbols in a frame of data by this calculation method, if a frame of data contains F orthogonal spread spectrum symbols, then the total number of bits of a frame of data is F K; Step 4: Take the soft bit value of one frame of data obtained as the input of channel decoding, perform TPC/LDPC iterative decoding, and output a total of F·K bits of data {b _n,k } from the decoding decision, where b _n,k represent The kth bit data of the nth symbol period; Step 5: Decision feedback outputs all hard decision bits {b _n,k }, and reconstructs modulation to generate F orthogonal spread spectrum modulation symbols as Among them, for the K-bit data {b _n,1 ,b _n,2 ,…,b _n,K } of the nth symbol period, the orthogonal spread spectrum symbols generated by reconstruction modulation are Will Correlate with the received sequence R _n of the nth symbol period to obtain the path response h _n ; ${\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}{\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; W</span>}}}_{\mathrm{<span\; class="notranslate">\; no</span>}}^{\mathrm{<span\; class="notranslate">\; h</span>}}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; N</span>}<span\; class="notranslate">\; ;</span>$ Among them, R _n is the sampling sequence of the nth symbol period, Indicates that the hard decision bit reconstructs the orthogonal spread spectrum symbol in the nth symbol period The conjugate transpose of; N is the spreading code length; Then the path response in the n-2th symbol period is Similarly, the path response of the preceding and following 2M symbol periods can be obtained as h _nM ,...,h _n-1 ,h _n+1 ,...,h _n+M , and the path response of the nth symbol period is estimated by using the 2M path responses for: ${\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; m</span>}}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; ...</span><span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; ...</span><span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; m</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; /</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; \&alpha;</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}\mathrm{<span\; class="notranslate">\; exp</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; j\&phi;</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; )</span>$ Among them, M represents the number of symbols before and after the joint channel estimation; α _n is the path gain, and φ _n is the path phase; Step 6, use the phase information φ _n restored by channel estimation to realize coherent demodulation of orthogonal spread spectrum, first compensate the phase influence of the channel, and then take the real part to express as The coherent demodulation output bit soft value is: $\mathrm{<span\; class="notranslate">\; \&Lambda;</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; b</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \&Element;</span>{\mathrm{<span\; class="notranslate">\; S</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}^{<span\; class="notranslate">\; +</span>}}{\mathrm{<span\; class="notranslate">\; m</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}^{\mathrm{<span\; class="notranslate">\; R</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \&Element;</span>{\mathrm{<span\; class="notranslate">\; S</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}^{<span\; class="notranslate">\; -</span>}}{\mathrm{<span\; class="notranslate">\; m</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}^{\mathrm{<span\; class="notranslate">\; R</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; \&le;</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; \&le;</span>\mathrm{<span\; class="notranslate">\; K</span>}$ Coherent demodulation of the current symbol requires multiple symbols before and after the channel estimation. It can be seen that the M symbols before and after a frame symbol cannot be coherently demodulated, and the non-coherent demodulation bit soft value of the 2M symbols at both ends is coherent with the middle symbol Demodulate a total of F·K bit soft values as channel decoding input; then turn to the fourth step for channel decoding until the iteration stops, and finally output the decision data bits b _n,k . 2. a kind of high-performance short-wave orthogonal spread spectrum iterative demodulation and decoding method according to claim 1, is characterized in that, in step 1, at first channel coding is carried out to sending information bit, afterward coded data is carried out Cross spread spectrum modulation to obtain a set of orthogonal spread spectrum sequences. 3. a kind of high-performance short-wave orthogonal spread spectrum iterative demodulation and decoding method according to claim 2 is characterized in that, in the step 1, transmit information bit is carried out between channel coding and orthogonal spread spectrum modulation , also including the step of interleaving the coded data; After performing the interleaving step at the sender, In step 4, after deinterleaving the obtained soft data bit value of one frame, it is used as an input for channel decoding; At the same time, in step 5, after all the hard decision bits output by the decision feedback are interleaved, they are reconstructed and modulated to generate F orthogonal spread spectrum modulation symbols; In addition, in Step 6, the F·K bit soft value is deinterleaved and input as channel decoding. 4. a kind of high-performance short-wave orthogonal spread spectrum iterative demodulation and decoding method according to claim 1, is characterized in that, described step 4 realization method is as follows: Utilize demodulation output soft value, carry out TPC/LDPC Iterative decoding, the decoding method adopts the general TPC or LDPC decoding algorithm, after the iterative decoding, the soft value of the bit is judged, and the hard decision bit b _n,k is output.