CN100349447C - Data balancing method for meteor trail communication - Google Patents

Abstract

The invention discloses a data equalization method for meteor trail communication. In order to solve the problem that the existing joint channel tracking and maximum likelihood equalization methods are difficult to realize in meteor communication. The technical principle is: based on the initial estimation vector of the channel parameters, the received baseband distortion signal is equalized, and the judgment data is output to restore the original transmitted signal. The specific steps are: according to the initial estimation vector  _o of the channel parameter and the received signal d _k (k=1, 2...), obtain the channel parameter estimation vector (μ _k ); calculate the values of all states of the received signal from this vector Path branch metric increment λ _N (μ _k →μ _k+1 ), and then calculate the path branch metric Г(μ _k+1 ) of each state, and find the state with the smallest branch metric μ _k+1 ^min ; according to The path branch metric Г(μ _k+1 ) and the real-time state deletion threshold Th _k+1 ^max delete the state, and then perform path backtracking on the state μ _k+1 ^min , and output the corresponding decision signal s′ _{(k +1)-δ} . The invention reduces computational complexity and storage space, realizes rapid processing of meteor communication signals and high-quality data reception, and can be used for meteor communication.

Description

Data Equalization Method for Meteor Trail Communication

technical field

The invention belongs to the technical field of communication, and in particular relates to a data equalization method for meteor trail communication, that is, a combined channel tracking and maximum likelihood equalization method ASP based on adaptive state deletion.

Background technique

Meteor trail communication is an effective communication guarantee method with strong anti-interference ability and ability to resist changes in the external environment, and it is a necessary supplement to conventional communication. In recent years, many countries in the world have already possessed their own meteor trail communication equipment and systems. The United States, Japan, etc. have also equipped their national troops with meteor communication systems, and established a meteor trail communication network on this basis. At present, my country is also embarking on research in this area.

Meteor trail communication is a kind of burst communication (MBC), which uses meteor ionization trails to reflect and scatter radio waves to communicate, and is an effective communication guarantee method when it is disturbed by physics or electronics. When conventional communication methods are affected by electronic interference, earthquakes, tsunamis, and nuclear explosions, etc., meteor trail communication can play a role in communication support at this critical moment. Meteor trail communication has the following outstanding features:

1. Good confidentiality and strong anti-interference ability. This is due to the fact that meteor trails are fleeting and have obvious directionality to radio wave reflections, which are not easy to be detected, intercepted and interfered by the enemy. Moreover, meteors will not be destroyed, and communication can be restored in an instant after a nuclear explosion. , electronic interference is also difficult to achieve the purpose of destruction;

2. Long transmission distance and good communication stability. Using ordinary antennas, the communication distance can reach more than 2,000 kilometers, and the communication quality will not be affected by changes in time, space, climate, or interference from the high-altitude ionosphere;

3. The cost of communication transmission and maintenance is low, and the ground transceiver equipment is simple, which greatly reduces the communication cost.

In view of the above characteristics, meteor trail communication can become a backup remote communication means for other communications, and has a special and important position in emergency communication.

Under such background, the present invention is produced in combination with the national "Tenth Five-Year Plan" military pre-research background project "meteor trail communication adaptive data transmission and networking protocol and special antenna technology". Meteor trail communication is a special communication method. The number of meteors entering the atmosphere is inversely proportional to the electron line density, so most of the available trails are less dense. The effective duration of each meteor communication is very short, about tens to hundreds of milliseconds, so the average data throughput rate of meteor communication is very low (generally no more than tens of bits/second). In order to improve the instantaneous data throughput rate, the meteor trail communication system needs to adopt adaptive variable rate technology. In consideration of reducing the complexity of the transceiver, the symbol width of signals of different rates is usually kept the same, that is, the signal bandwidth, and the modulation method of the system is changed, that is, the adaptive coding and modulation technology based on MPSK (M=2, 4...16) is adopted , the modulation mode is from 2PSK, QPSK to 16QAM, and the data transmission rate is adaptively changed from 2kbps to 64kbps, as shown in Figure 1.

Meteor trails are short in duration, and the transmitted data is usually divided into many data frames. In the communication process, the meteor trails used in adjacent frames may be different, so the adaptive coding and modulation is performed in units of data frames. Correspondingly, data transmission and reception processing also take data frames as a unit. The data frame structure of meteor trail communication is shown in Figure 2. The 32-bit frame header in the figure contains the variable rate modulation information corresponding to the frame. The number of data packets contained in each frame is different, and is modulated by adaptive coding. The specific modulation method is determined.

According to the meteor trail communication mechanism and frame structure above, the structure of the meteor trail communication system based on the adaptive coding and modulation mechanism is shown in Figure 3, in which the data processing operation of the receiver is mainly completed by a high-speed digital signal processor (DSP). The control of adaptive variable rate is completed in the response phase of the link layer. It can be seen that the transmission process of the meteor trail communication signal is as follows: the transmitted signal x _k (k=1, 2...) is first encoded into s _k (k=1, 2...) to enhance the anti-interference ability, and then the serial/ And transform it into two-way signals of I-way and Q-way, and turn it into a baseband modulation signal according to the constellation mapping. The modulated signal is wave-formed by a raised cosine filter, then converted into a radio frequency signal by up-conversion, and sent to the channel through the antenna. The transmission signal will be distorted by multipath effect and noise pollution, so the receiving end of Meteor communication should equalize the received distorted signal to restore the original sending signal. The specific receiving process can also be seen from Figure 3 : As shown in the figure, the radio frequency signal received by the antenna is down-converted and filtered to obtain the distorted baseband received signal d _k (k=1, 2...), firstly estimated according to the received training sequence, that is, the transmission frame header Initial estimate vector of channel parameters Then use the initial channel parameter estimation vector The data equalization module is started, and the effective information data after the transmission frame header is equalized to restore the original transmission signal. The decision data output by the equalizer is s' _k (k=1, 2...), and the final output after decoding is x' _k (k=1, 2...). On the basis of accurately estimating channel parameters, if the equalization method used has ideal performance, the decision data s′ _k (k=1, 2…) output by the equalizer should be equal to s _k (k=1, 2…), The corresponding decoding output x' _k (k=1, 2...) is also equal to the original sending signal x _k (k=1, 2...), which can realize the reliable reception of meteor communication data.

It can be seen from the basic principle of Meteor communication that the performance of data equalization directly determines the performance of Meteor communication system. The effective duration of Meteor communication is short, and the data overhead required for channel acquisition, synchronization and frequency error correction has a great impact on the payload data throughput rate. Therefore, it has become a key technology that must be solved in meteor communication to study the data equalization method that has fast processing speed, conforms to the meteor trail channel characteristics, adapts to the adaptive variable speed meteor trail communication mechanism, and has low computing and storage overhead. The present invention is aimed at the data equalization technology under the meteor trail communication condition.

At present, the existing general equalization techniques are divided into two types: linear equalization and nonlinear equalization. Linear equalization mainly refers to a linear equalizer based on a transversal filter structure; nonlinear equalization mainly includes decision feedback equalization (DFE), adaptive equalization and maximum likelihood equalization.

For linear equalization, the linear equalizer based on the transversal filter structure is simple and easy to implement, but it cannot offset serious channel distortion and cannot meet the system performance requirements.

In nonlinear equalization, there are also many problems. For example: ① Decision Feedback Equalization (DFE) achieves a better equalization effect by eliminating the part of the intersymbol interference caused by the decided signal from the current estimated value, but since the update of the equalizer coefficient depends on the adjustment of the adaptive algorithm, so It is impossible to track the rapid changes of the meteor channel, and even the non-convergence of the coefficients occurs. Nonlinear Channel Decision Feedback Equalization Algorithm Based on Wavelet Transform, System Engineering and Electronic Technology, Issue 12, 2002, etc. ② Adaptive equalization uses the inverse simulation principle of the adaptive filter to offset the intersymbol interference generated by the multipath channel. Its convergence speed is slow, and the working premise of the adaptive algorithm is that the statistical characteristics of the channel parameters remain unchanged. In the meteor communication environment, the sudden change of channel characteristics makes the signal non-stationary, and its tracking and convergence capabilities cannot adapt, and cannot meet the performance requirements of the meteor communication system. See Simon Haykin.Adaptive Filter Theory (FourthEdition)[M].Beijing : Publishing House of Electronics Industry, 2002. And Xu Mingyuan, Lin Huafang, Qiu Gongan "Simulation Research of Adaptive Equalization System Based on LMS Algorithm", System Simulation Journal, 2003, Issue 02, etc. ③ Although the maximum likelihood equalization (MLSE) based on the Viterbi algorithm has good receiving performance, it cannot meet the real-time signal processing requirements of Meteor Communication due to the introduction of decision delay and high computational complexity in the front-end channel estimation. See American John G Prokies. "Digital Communication" [M]. Beijing: Publishing House of Electronics Industry, 1998 and Song Liang, Hu Bo, Ling Xieting "A New Adaptive MLSE Receiver Based on Fast Fading Channel", Electronic Journal, 2002, 30(5):723-726. However, on the basis of maximum likelihood equalization, the joint channel tracking and maximum likelihood equalization method (PSP) based on surviving path-by-survival path processing introduces real-time channel tracking into the surviving path processing of the Viterbi algorithm (VA), which overcomes the traditional The decision delay problem of maximum likelihood equalization has theoretically optimal receiving performance, see Italy Riccardo Raheli, Andreas Polydoros, Ching-Kae Tzou.Per-survivor processing: A general approach to MLSE in uncertain environments[J].IEEE Trans on Comm, 1995, 43: 354-364.

However, because the PSP method for channel-by-survival path processing requires M ^L calculation and storage space when performing channel tracking and data detection, where M is the modulation number and L is the equivalent channel length, it is difficult to implement in engineering. In response to this situation, the existing technologies at home and abroad mainly adopt the following three methods for simplification:

The first type of method is that Canadian J.Omidi and Rollins proposed to simplify the structure of the adaptive module method used for channel estimation in the joint channel tracking and maximum likelihood equalization method to reduce the overall complexity. See Canadian J.Omidi, P.G.Gulak, and S.Pasupathy, Parallel Structures for Joint Channel Estimation and Data Detection Over Fading Channels[J], IEEE Transactions on Selected Areas of Communications, 1998, 16(5): 1616-1629. and Canada Rollins, M.E., J.Simmons, S. Simplified per-survivor Kalman processing in fast frequency-selective fading channels [J]. IEEE Trans on Comm., 1997, 45(5): 544-553.

The second type of method is that G Castellini in Italy proposed to reduce the number of channel tracking at each moment. In the case of little performance loss, only perform channel tracking of the surviving path once for the state with the smallest branch metric, and use the estimated channel parameters As the common channel parameter of all states at this moment, that is, the MSP method, see Italian G Castellini, F Conti, E Del, Re, L Pierucci.A continuously adaptive MLSE receiver for mobile communications: algorithm and performance[J].IEEE Trans on Comm, 1997, 45(1):80-89.), or use the PSP method and the MSP method in combination on this basis to reduce the amount of computation, see Korea Jung Suk Joo, SeungChul Hong, Yong Hoon Lee.Adaptive MLSE receiver: hybrid of per-survivorprocessing and tentative decision MLSE [J]. Electronics Letters, 2000, 36(7): 678-680.);

The third type of method is based on the combination of channel tracking and maximum likelihood detection combined with decision feedback equalization according to Zhenhong Li of Finland, see Zhenhong Li, Piirainen, O., Mammela, A.A new reduced-complexity adaptive PSP-MLSE receiver for GSM/EDGE systems[J]. IEEE International Symposium on Personal, Indoor and Mobile Radio Communications, 2001(1): 124-128.), using decision feedback equalization to deal with partial multipath, equivalently reducing the multipath processing of the joint detection method Path channel length.

All the above methods start from the local module or outside of the joint channel tracking and maximum likelihood equalization method (PSP), and try to reduce the calculation amount of the method by simplifying the local algorithm or performing external processing. However, the structure of the PSP method has not changed, and still requires a large amount of computing and storage space, which is still difficult to implement for meteor communication systems that require fast data processing and are suitable for time-varying channel characteristics.

Technical content of the invention

As one of the key technologies of meteor communication, the data equalization method used is closely related to the characteristics of the meteor trail channel. As mentioned above, the PSP joint channel tracking and equalization method based on the survival path processing overcomes the channel estimation delay and error propagation problems of the traditional equalization algorithm, and combines the real-time tracking of the channel parameter vector with data equalization, which has the best Receive performance. However, its algorithmic complexity and huge storage space for data operations make it difficult to implement in the actual Meteor system. Therefore, the purpose of the present invention will solve these problems exactly, set out from meteor channel characteristic and communication mechanism, according to the train of thought of dynamic processing, propose a kind of data equalization method that is used for meteor trail communication, namely based on self-adaptive state A joint channel tracking and maximum likelihood equalization (ASP) method based on puncturing. On the basis of effectively reducing the computational complexity and storage space, realize the rapid processing of meteor communication signals and high-quality data reception, thereby improving the effective data pass rate of meteor communication, and opening up a new field of key technology for signal processing of a new generation of meteor trail communication system .

对接收到的基带畸变信号d_k(k＝1，2…)进行均衡处理，输出判决数据s′_k(k＝1，2…)以恢复出原始的发送信号。The technical scheme that realizes the object of the present invention is to set out from the train of thought of dynamic data processing, carry out adaptive state deletion to joint channel tracking and equalization method PSP according to the exponential fading characteristic of meteor channel, utilize real-time state to delete and shift threshold Th _k ^max in data receiving During the process, the number of states of the PSP is dynamically controlled, and the maximum likelihood reception of the data in the adaptive coding and modulation mode of the meteor trail communication is realized. As mentioned above, in the Meteor communication system, the main function of the equalization part is to estimate the vector according to the channel parameters

Perform equalization processing on the received baseband distorted signal d _k (k=1, 2...), and output decision data _s'k (k=1, 2...) to restore the original sent signal.

Assuming that the channel length is L and the decision depth of the system based on the Viterbi structure is δ, the specific implementation steps of the joint channel tracking and equalization method ASP based on adaptive state shifting can be divided into the following six steps:

其中L为流星信道长度；In the first step, the initial estimation vector of channel parameters and the received signal d _k (k=1, 2...), at each time k, use the adaptive RLS method to analyze all possible states of the received signal sequence {d _k , d _k-1 ...d _k-(L-1) } The channel parameter vector of μ _k is estimated, and the corresponding channel parameter estimation vector at time k is obtained

Where L is the meteor channel length;

通过公式 $\mathrm{\λ}({\mathrm{\μ}}_k\→{\mathrm{\μ}}_{k+1})={|d_{k+1}-\hat{h}\left({\mathrm{\μ}}_k\right)S^{\′T}\left({\mathrm{\μ}}_{k+1}\right)|}^2$ 计算出接收信号在k时刻到k+1时刻从所有可能状态μ_k转移到状态μ_k+1的路径分支度量增量λ(μ_k→μ_k+1)，式中：符号(·)^T代表转置运算，S′(μ_k+1)代表接收信号所有可能状态μ_k+1直接映射的L个信息码元的判决序列{s′_k+1-(L-1)…s′_k，s′_k+1}；In the second step, the corresponding channel parameter estimation vector at time k is obtained

by formula $\mathrm{\λ}({\mathrm{\μ}}_k\&Right\; Arrow;{\mathrm{\μ}}_{k+1})={|d_{k+1}-\hat{h}\left({\mathrm{\μ}}_k\right)S^{\′T}\left({\mathrm{\μ}}_{k+1}\right)|}^2$ Calculate the path branch metric increment λ(μ _k →μ _k+1 ) of the received signal from all possible states μ _k to the state μ _k+1 at time k to k+1, where the symbol ( ) ^T Represents the transpose operation, S′(μ _k+1 ) represents the decision sequence {s′ _k+1-(L-1) …s′ _k of L information symbols directly mapped to all possible states of the received signal μ _k+1 , s′ _k+1 };

In the third step, based on the obtained path branch metric increment λ(μ _k →μ _k+1 ), on the basis of the path branch metric Γ(μ _k ) of all possible states μ _k of the received signal at time k, use the formula $\mathrm{\Γ}\left({\mathrm{\μ}}_{k+1}\right)=\underset{{\mathrm{\μ}}_k}{\min }[\mathrm{\Γ}\left({\mathrm{\μ}}_k\right)+\mathrm{\λ}({\mathrm{\μ}}_k\&Right\; Arrow;{\mathrm{\μ}}_{k+1})]$ Calculate the path branch metric Γ(μ k+1 ) of all possible states μ _k+1 of the received signal sequence {d _k+1 , d _k …d _k-(L-2) } at time k ₊₁ ;

The fourth step is to calculate the real-time state deletion threshold Th _k+1 ^max at time k+1 by the following formula:

$\mathrm{<span\; class="notranslate">\; T</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}^{\mathrm{<span\; class="notranslate">\; max</span>}}<span\; class="notranslate">\; \&ap;</span>{{\mathrm{<span\; class="notranslate">\; <figure-callout\; id="1"\; label="C"\; filenames="C20041007300200191.png,C20041007300200201.png"\; state="\{\{state\}\}">C</figure-callout></span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}{\mathrm{<span\; class="notranslate">\; <figure-callout\; id="0"\; label="N"\; filenames="C20041007300200171.png,C20041007300200201.png"\; state="\{\{state\}\}">N</figure-callout></span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; \&Center\; Dot;</span>{\mathrm{<span\; class="notranslate">\; e</span>}}^{<span\; class="notranslate">\; -</span>\frac{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; T</span>}}_{\mathrm{<span\; class="notranslate">\; the\; s</span>}}}{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}}$

Among them, T _s is the symbol width; N ₀ is the power spectral density of Gaussian white noise, τ is the attenuation coefficient of the meteor trail; C ₁ is the link constant of the meteor communication, which is offset from the correct state set by the communication system The probability P _c ^r has the following relationship,

${\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; c</span>}}^{\mathrm{<span\; class="notranslate">\; r</span>}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 2</span>}}\mathrm{<span\; class="notranslate">\; erfc</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; )</span>$

In the formula, erfc( ) is the complementary error function, so C ₁ can be obtained by looking up the complementary error function table from _Pc ^r ;

In the fifth step, from the path branch metric Γ(μ _k+1 ) obtained in the third step, find out the state μ _k+1 ^min with the minimum path branch metric Γ(μ _k+1 ) at time k+1, and then use the The real-time state pruning threshold Th _k+1 ^max obtained in the four steps deletes the state whose branch metric is greater than the sum of Γ(μ _k+1 ^min ) and the threshold Th _k+1 ^max to complete the state pruning at this moment;

The sixth step is to perform path backtracking on the state μ _k+1 ^min of the minimum branch metric, and output the corresponding decision signal s′ _(k+1)-δ at this moment, where δ is the system decision depth.

In the first step in the above method, the adaptive RLS method is used to carry out the channel parameter vectors of all possible states μ _k of the received signal sequence {d _k , d _k-1 ... d _k-(L-1) } at time k Estimated by the following four formulas:

$\mathrm{<span\; class="notranslate">\; e</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; -</span>{\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}}^{<span\; class="notranslate">\; *</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; S</span>}}^{<span\; class="notranslate">\; \&prime;</span>\mathrm{<span\; class="notranslate">\; T</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span>$

$\mathrm{<span\; class="notranslate">\; g</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; S</span>}}^{<span\; class="notranslate">\; \&prime;</span>\mathrm{<span\; class="notranslate">\; T</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; w</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; S</span>}}^{<span\; class="notranslate">\; \&prime;</span><span\; class="notranslate">\; *</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; S</span>}}^{<span\; class="notranslate">\; \&prime;</span>\mathrm{<span\; class="notranslate">\; T</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span>}$

$\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; w</span>}}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; g</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; S</span>}}^{<span\; class="notranslate">\; \&prime;</span><span\; class="notranslate">\; *</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span>$

$\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; g</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; e</span>}}^{<span\; class="notranslate">\; *</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span>$

式中符号符号From the above four formulas, the corresponding channel parameter estimation vector at time k is obtained

Symbols in the formula

的初值为信道参数的初估矢量，即 $\hat{h}\left({\mathrm{\&mu;}}_0\right)={\hat{h}}_0.$ ( ) ^* represents the conjugate operation, e(μ _k ) is the error signal, that is, the received signal d _k and the expected output signal ${\hat{h}}^{*}\left({\mathrm{\&mu;}}_{k-1}\right)S^{\&prime;T}\left({\mathrm{\&mu;}}_k\right)$ , w is the forgetting factor, g(μ _k ) is the Kalman gain vector, p(μ _k ) is the correlation matrix, s′(μ _k ) represents the state μ _k at time k for the received data sequence {d _k , d _k-1 ... d _k-(L-1) } decision sequence {s′ _k-(L-1) ... s′ _k-1 , s′ _k }. In the adaptive RLS method, when k=1 starts the first operation, the initial value p(μ ₀ ) of p(μ _k ) takes the identity matrix,

The initial value of is the initial estimation vector of channel parameters, namely $\hat{h}\left({\mathrm{\&mu;}}_0\right)={\hat{h}}_0.$

In the sixth step of the above method, the path backtracking is performed on the state μ _k+1 ^min of the minimum branch metric at time k+1. The specific backtracking process is: if the system decision depth is δ, first find out the arrival state μ _k The previous moment of ₊₁ ^min , that is, the state μ _k at time k, and then find out the state at the previous moment of reaching state μ _k , that is, the state μ k-1 at time _k-1 ... and so on, until (k The state μ _{(k+1)-δ at time +1)-δ} is used for symbol judgment, and the corresponding decision signal s _(k+1)-δ ′ is output at this time.

Through the above six steps, the decision signal s _(k+1)-δ ′ is finally output, thereby completing a joint data detection of the ASP method. Then return to step (1) to start joint data detection at the next moment. By analogy, the equalization for the received signal sequence d _k (k=1, 2...) can be completed, and finally output the decision data sequence s' _k (k=1, 2...) in the meteor trail communication system shown in Figure 3 , so as to realize the maximum likelihood reliable reception of meteor communication data.

The present invention has following effect:

(1) Low computational complexity

Because the present invention is Path(ASP)=M M ^L-1 =M ^L+1-1 to all state path branch measures total number of calculations at each moment, wherein l is the correct state deletion probability P according to system requirements _c ^r deleted state numbers, and the total path branch metric calculation times at each moment of the existing PSP method is Path(PSP)=M M ^L =M ^L+1 , so the path branch metric calculation of the method of the present invention is relative to Existing PSP methods reduce M by a factor ^{of 1} . The simulation data shows that the computational complexity of the present invention is significantly lower than that of the existing PSP method, as shown in Fig. 5 and Fig. 6 . In actual communication, for the data frames of different modulation modes of meteor trail communication adaptive variable rate, the ASP method of the present invention can effectively control the computational complexity and storage space of the joint maximum likelihood equalization, so that each frame data processing has approximately the same The computing overhead, so as to make full use of the limited DSP instruction resources of the system, it is easy to achieve the best receiving quality.

(2) Good system performance

The invention not only has low computational complexity, simple method and easy realization, but also has system performance close to that of the existing joint channel tracking and maximum likelihood equalization method. Simulation experiments show that, for different correct state pruning probabilities P _c ^r of the present invention, under the QPSK modulation mode, when the signal-to-noise ratio E _s /N ₀ is between 10.5dB and 11.5dB, the bit error rate reaches 1× 10 ^-3 , as shown in Figure 7. In the 16QAM modulation mode, when the signal-to-noise ratio E _s /N ₀ is greater than 18dB, the bit error rate also reaches 1×10 ^-3 , as shown in Figure 8 .

Description of drawings

Figure 1 is an adaptive change diagram of the meteor communication transmission rate

Figure 2 is a structure diagram of meteor communication data frame

Figure 3 is a structural block diagram of the meteor trail communication system

Fig. 4 is the realization flowchart of the inventive method

Figure 5 is a comparison diagram of computational complexity under QPSK modulation mode

Figure 6 is a comparison diagram of computational complexity under 16QAM modulation mode

Figure 7 is a graph of the bit error rate of the computer simulation system under the QPSK modulation mode

Figure 8 is the BER curve of the computer simulation system under the 16QAM modulation mode

Detailed ways

对接收到的基带畸变信号d_k(k＝1，2…)进行均衡处理，输出判决数据s′_k(k＝1，2…)以恢复出原始的发送信号，实现数据的可靠接收。本发明从动态数据处理的思路出发，根据流星信道特点对联合信道跟踪与均衡方法PSP进行简化，提供一种基于自适应状态删移的联合信道跟踪与最大似然均衡ASP方法。Referring to FIG. 3 , the principle of the existing Meteor communication system is as described in the background art. Among them, the main function of the equalization part is based on the initial estimation vector of the channel parameters

Perform equalization processing on the received baseband distorted signal d _k (k=1, 2...), and output decision data s' _k (k=1, 2...) to restore the original transmission signal, and realize reliable data reception. The present invention starts from the idea of dynamic data processing, simplifies the joint channel tracking and equalization method PSP according to the characteristics of the meteor channel, and provides a joint channel tracking and maximum likelihood equalization ASP method based on adaptive state deletion.

The basic train of thought of the inventive method is that all possible state numbers of the received data of joint channel tracking and equalization method PSP at every moment are M ^L , wherein M is the modulation base number of the received signal, and L is the meteor channel length, so every For a moment, the huge computational complexity and storage space of the PSP method make it difficult to realize in the Meteor communication system. The combined channel tracking and maximum likelihood equalization ASP method based on adaptive state deletion provided by the present invention calculates the real-time state deletion threshold Th at each time k according to the correct state deletion probability P _c ^r required by the meteor system _k ^max , using this threshold to delete the part of the state with the least possibility of correctness in real time during the equalization process of the received data, so as to adaptively control the number of states of the received signal, effectively reduce the system complexity and storage space, and obtain a relative The optimal receiving performance realizes the joint maximum likelihood reception of meteor trail communication adaptive coding and modulation data.

With reference to Fig. 4, the present invention has provided the joint channel tracking and equalization method based on self-adaptive state deletion, i.e. the concrete implementation flow of data equalization method ASP, and its process is:

First, system initialization is performed on the Meteor communication receiver to realize signal acquisition and synchronization, and obtain the initial channel parameter estimation vector

Then, on this basis, the joint channel tracking and equalization processing based on adaptive state pruning is performed on the payload data d _k (k=1, 2...) after the frame header, and the corresponding decision data s′ _k (k = 1, 2...). Assuming that the channel length is L, and the decision depth of the system is δ, the specific implementation is as follows:

(1) Calculate the channel parameter estimation vector

之后，首先根据信道参数的初估矢量

和接收信号d_k(k＝1，2…)，在每一个k时刻，采用自适应的递推最小二乘RLS方法，通过以下四式对接收信号序列{d_k，d_k-1…d_k-(L-1)}所有可能状态μ_k的信道参数矢量进行估计，得到k时刻相应的信道参数估计矢量

即Realize signal acquisition and synchronization in Meteor communication system, and obtain initial channel parameter estimation vector

After that, firstly, according to the initial estimation vector of channel parameters

and the received signal d _k (k=1, 2...), at each time k, using the adaptive recursive least squares RLS method, the received signal sequence {d _k , d _k-1 ...d _k-(L-1) } Estimate the channel parameter vectors of all possible states μ _k , and obtain the corresponding channel parameter estimation vector at time k

Right now

$\mathrm{<span\; class="notranslate">\; e</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; -</span>{\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}}^{<span\; class="notranslate">\; *</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; S</span>}}^{<span\; class="notranslate">\; \&prime;</span>\mathrm{<span\; class="notranslate">\; T</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span>$

$\mathrm{<span\; class="notranslate">\; g</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; S</span>}}^{<span\; class="notranslate">\; \&prime;</span>\mathrm{<span\; class="notranslate">\; T</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; w</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; S</span>}}^{<span\; class="notranslate">\; \&prime;</span><span\; class="notranslate">\; *</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; S</span>}}^{<span\; class="notranslate">\; \&prime;</span>\mathrm{<span\; class="notranslate">\; T</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span>}$

$\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; w</span>}}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; g</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; S</span>}}^{<span\; class="notranslate">\; \&prime;</span><span\; class="notranslate">\; *</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span>$

$\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; g</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; e</span>}}^{<span\; class="notranslate">\; *</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span>$

In the formula: the symbol ( ) ^* represents the conjugate operation, e(μ _k ) is the error signal, that is, the received signal d _k and the expected output signal ${\hat{h}}^{*}\left({\mathrm{\&mu;}}_{k-1}\right)S^{\&prime;T}\left({\mathrm{\&mu;}}_k\right)$ , w is the forgetting factor, g(μ _k ) is the Kalman gain vector, p(μ _k ) is the correlation matrix, S′(μ _k ) represents the state μ _k at time k for the received data sequence {d _k , d _k-1 ... d _k-(L-1) } decision sequence {s′ _k-(L-1) ... s′ _k-1 , s′ _k }. In the adaptive RLS method, when k=1 starts the first operation, the initial value p(μ ₀ ) of p(μ _k ) takes the identity matrix, The initial value of is the initial estimation vector of channel parameters, namely $\hat{h}\left({\mathrm{\&mu;}}_0\right)={\hat{h}}_0.$

进一步计算出接收信号在k时刻到k+1时刻从所有可能状态μ_k转移到状态μ_k+1的路径分支度量增量λ(μ_k→μ_k+1)。具体的实施方式是：(2) Calculate the path branch metric increment λ(μ _k →μ _k+1 ) according to the channel parameter estimation vector corresponding to all states μ _k at time k obtained in step (1)

Further calculate the path branch metric increment λ(μ _k →μ k+1 ) for the received signal to transfer from all possible states μ _k to the state μ _k+ 1 at time k to k ₊ 1. The specific implementation method is:

的基础上，通过下式计算出k时刻到k+1时刻从状态μ_k转移到状态μ_k+1的路径分支度量增量λ(μ_k→μ_k+1)：in estimated

On the basis of , the path branch metric increment λ(μ _k →μ _k+1 ) for transitioning from state μ _k to state μ _{k+1 from time k to k+1} is calculated by the following formula:

$\mathrm{<span\; class="notranslate">\; \&lambda;</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; \&Right\; Arrow;</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>{<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; -</span>\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; S</span>}}^{<span\; class="notranslate">\; \&prime;</span>\mathrm{<span\; class="notranslate">\; T</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; |</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}$

In the formula, the symbol (·) ^T represents the transpose operation, S′(μ _k+1 ) represents the decision sequence of L information symbols directly mapped to all possible states of the received signal μ _k+1 , and L is the equivalent meteor channel length;

(3) Calculate the path branch metric Γ(μ _k+1 )

According to the step (2), the path branch metric increment λ(μ _k →μ _k+1 ) of the state μ _k transferred to the state μ k ₊₁ is obtained, and the path branch metrics Γ of all possible states μ _k of the received signal at the existing k time ₍ μ _k ), calculate _the path branch metric _{Γ (} μ _{k +1} ). The specific implementation method is:

On the basis of calculating the path branch metric increment λ(μ _k →μ _k+1 ), the surviving path branch metric function Γ(μ _k+1 ) of the state μ _k+1 is further calculated, which is obtained by reaching the state μ _{k +1} for the minimum surviving path state decision, i.e.

$\mathrm{<span\; class="notranslate">\; \&Gamma;</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}}{\mathrm{<span\; class="notranslate">\; min</span>}}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; \&Gamma;</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; \&lambda;</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; \&Right\; Arrow;</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span>$

$<span\; class="notranslate">\; =</span>\underset{{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}}{\mathrm{<span\; class="notranslate">\; min</span>}}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; \&Gamma;</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span>{<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; -</span>\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; S</span>}}^{<span\; class="notranslate">\; \&prime;</span>\mathrm{<span\; class="notranslate">\; T</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; |</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; ]</span>$

(4) Calculate the real-time state deletion threshold Th _k+1 ^max .

The real-time state pruning threshold Th k ₊₁ ^max at time k+1 is calculated by the following formula:

$\mathrm{<span\; class="notranslate">\; T</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}^{\mathrm{<span\; class="notranslate">\; max</span>}}<span\; class="notranslate">\; \&ap;</span>{{\mathrm{<span\; class="notranslate">\; <figure-callout\; id="1"\; label="C"\; filenames="C20041007300200191.png,C20041007300200201.png"\; state="\{\{state\}\}">C</figure-callout></span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}{\mathrm{<span\; class="notranslate">\; <figure-callout\; id="0"\; label="N"\; filenames="C20041007300200171.png,C20041007300200201.png"\; state="\{\{state\}\}">N</figure-callout></span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; \&CenterDot;</span>{\mathrm{<span\; class="notranslate">\; e</span>}}^{<span\; class="notranslate">\; -</span>\frac{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; T</span>}}_{\mathrm{<span\; class="notranslate">\; the\; s</span>}}}{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}}$

Among them, T _s is the symbol width; N ₀ is the power spectral density of Gaussian white noise, τ is the attenuation coefficient of the meteor trail; C ₁ is the link constant of the meteor communication, which is offset from the correct state set by the communication system The probability P _r ^c has the following relationship,

${\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; c</span>}}^{\mathrm{<span\; class="notranslate">\; r</span>}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 2</span>}}\mathrm{<span\; class="notranslate">\; erfc</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; )</span>$

In the formula, erfc( ) is the complementary error function, so C ₁ can be obtained by looking up the complementary error function table from _Pc ^r ;

(5) Perform state deletion

According to the path branch metrics Γ(μ _k +1 ) of all possible states μ _k+1 of the received signal at time k+1 obtained in step (3) and the real-time state deletion threshold Th _k+1 ^max obtained in step (4), find Out of the state μ k+ ₁ ^min with the minimum path branch metric Γ(μ _k+1 ) at time k+1, delete the state whose branch metric is greater than the sum of Γ(μ _k+1 ^min ) and the threshold Th _k+1 ^max , to complete the state deletion at this moment.

(6) Output decision signal s′ _(k+1)-δ

Perform path backtracking on the state μ _k+1 ^min of the minimum branch metric, and output the corresponding decision signal s′ _(k+1)-δ at this moment, where δ is the system decision depth. The specific _backtracking process is: if the system decision depth is δ, first find out the state μ k at the moment k before reaching the state μ _k+1 ^min , and then find out the state at the moment before reaching the state μ _k , That is, the state μ k-1 at time _k-1 ... and so on, until the state μ (k+1)-δ at time _(k+1)-δ is found, the symbol judgment is performed on it, and the corresponding judgment at this time is output Signal s′ _(k+1)-δ , so as to complete a joint data detection of the ASP method at time k+1.

Through the above six steps, the joint channel tracking and equalization ASP processing based on adaptive state shifting at time k+1 can be completed, and the corresponding decision signal s′ _(k+1)-δ can be output. Then return to step (1) to start joint data detection at the next moment. By analogy, the equalization of the received signal sequence d _k (k=1, 2...) can be completed, and finally the corresponding decision data sequence s' _k (k=1, 2...) shown in Figure 3 is output, thereby effectively reducing the Under the conditions of computational complexity and storage space, the maximum likelihood reliable reception of Meteor communication data is realized.

With reference to Fig. 5 and Fig. 6, the present invention provides the normalized computational complexity under the quadrature phase-shift keying (QPSK) modulation mode and under the hexadecimal phase-shift keying (16QAM) modulation mode respectively by computer simulation curve. As can be seen from Fig. 5 and Fig. 6, the existing PSP method has the largest complexity, and the computational complexity of PSP and MSP does not change with the signal-to-noise ratio, while the computational complexity of the ASP method of the present invention increases with the increase of the signal-to-noise ratio decreases gradually and tends to a minimum value. The reason is that as the signal-to-noise ratio increases, the branch metric difference between the correct state and the wrong state increases, and the number of remaining states after pruning decreases, so the complexity also decreases until only the only until the correct state. Since the number of signal states of the 16QAM modulation method is much larger than that of QPSK, the branch metric comparison and pruning of each state also bring a certain amount of computing overhead, so the minimum complexity of the 16QAM modulation method is greater than the minimum value of QPSK. It can be seen that the method of the present invention has obvious reduction in computational complexity.

After the meteor trail communication system adopts the method of the present invention, its performance can be given by the following computer simulation experiment.

Experiment 1:

The simulated channel adopts the Watterson model recommended by CCIR. There is the following relationship between the meteor channel fading rate f _r and the random sequence generation rate f _c of the equivalent Rayleigh multipath fading channel.

${\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; r</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1.47</span>}<span\; class="notranslate">\; *</span>\frac{\mathrm{<span\; class="notranslate">\; 4.4</span>}{\mathrm{<span\; class="notranslate">\; <figure-callout\; id="2"\; label="f\; c"\; filenames="C20041007300200201.png"\; state="\{\{state\}\}">f</figure-callout></span><figure-callout\; id="2"\; label="f\; c"\; filenames="C20041007300200201.png"\; state="\{\{state\}\}">\; </figure-callout>}}_{\mathrm{<span\; class="notranslate">\; </span>}}}{\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; \&pi;</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>\sqrt{\mathrm{<span\; class="notranslate">\; 2</span>}}}$

Under the conditions of the typical 1Hz fading rate of meteor trail communication, the meteor trail attenuation coefficient τ=0.3s, and the multipath channel length L=4, the system-level simulation of joint channel tracking and maximum likelihood ASP data equalization is carried out to obtain QPSK Figure 7 is the relationship curve between ^the system bit error rate and the correct state deletion probability _Pcr under the modulation mode. The data transmission rate used in the simulation is 32kbps, and each symbol is sampled at 5 points, the following formula is

$\mathrm{<span\; class="notranslate">\; g</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; sin</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; \&pi;t</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; \&pi;t</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; T</span>}}<span\; class="notranslate">\; \&Center\; Dot;</span>\frac{\mathrm{<span\; class="notranslate">\; cos</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; \&alpha;\&pi;t</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 4</span>}{\mathrm{<span\; class="notranslate">\; \&alpha;</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}{\mathrm{<span\; class="notranslate">\; <figure-callout\; id="2"\; label="t"\; filenames="C20041007300200201.png"\; state="\{\{state\}\}">t</figure-callout></span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; T</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; )</span>}$

The raised cosine impulse response g(t) of the transmission filter, the carrier frequency is 40MHz, the noise is additive Gaussian white noise, and the system data transmission is not processed by interleaving and error correction coding. For the convenience of comparison, the Viterbi equalization method, the PSP method and the MSP method under the known channel are also simulated in FIG. 7 . Therefore, the bit error rate curve under known channels can be regarded as the performance upper bound of various methods. It can be seen from Figure 7 that in the correct state, the probability of deletion $P_c^r=0.001$ The performance of the ASP method is close to that of the PSP method; $P_c^r=0.01$ When the bit error rate performance of the ASP method is only 0.8dB lower than that of the PSP method; the system bit error rate performance of the MSP method is between $P_c^r=0.01$ and $P_c^r=0.001$ However, its computational complexity increases rapidly with the increase of the modulation number M. It can be seen that when the signal-to-noise ratio E _s /N ₀ is between 10.5dB and 11.5dB in the QPSK modulation mode, the bit error rate of the ASP method can reach 1×10 ^-3 . After product code or Turbo code encoding, the system bit error rate can reach 10 ^-6 , so that the reliable reception of Meteor communication data can be realized.

Experiment 2:

Similar to Experiment 1, the simulation channel also uses the Watterson model recommended by CCIR, and the system simulation conditions such as channel fading rate, channel length, symbol sampling point, and carrier frequency are the same. Under the modulation mode of 16QAM, the data transmission rate used in the simulation is 64kbps, and the system transmission data is not processed with interleaving and error correction coding, and the obtained bit error rate curve is shown in Figure 8. For the convenience of comparison, the Viterbi equalization method, the PSP method and the MSP method under the known channel are also simulated in FIG. 7 . Therefore, the bit error rate curve under known channels can be regarded as the performance upper bound of various methods. Similar to the results in Figure 7, it can be seen from Figure 8 that the correct state deletion probability $P_c^r=0.001$ The performance of the ASP method is close to that of the PSP method; $P_c^r=0.01$ When the bit error rate performance of the ASP method is lower than that of PSP by 1.5dB; the system bit error rate performance of the MSP method is between $P_c^r=0.01$ and $P_c^r=0.001$ However, its computational complexity increases rapidly with the increase of the modulation number M. It can be seen that in the 16QAM modulation mode, the bit error rate also reaches 1×10 ^-3 when E _s /N ₀ is greater than 18dB, so the bit error rate of the system after coding with a code efficiency of 1/2 convolutional code or Turbo code can reach up to 10 ^-6 .

Both of the above two experimental results prove that the method of the present invention can ensure the high-quality data reception of meteor trail communication.

Claims (3)

1. data balancing method that is used for meteoric trail communication, carry out as follows:

1. estimate vector by channel parameter first With received signal d _k(k=1,2 ...), at each k constantly, adopt self adaptation RLS method sequence { d to received signal _k, d _K-1D _K-(L-1)Institute might state μ _kThe channel parameter vector estimate, obtain the constantly corresponding channel parameter estimation vector of k

, wherein L is the meteor channel length;

2. by the constantly corresponding channel parameter estimation vector of the k that obtains

, pass through formula $\mathrm{\&lambda;}({\mathrm{\&mu;}}_k\&RightArrow;{\mathrm{\&mu;}}_{k+1})={|d_{k+1}-\hat{h}\left({\mathrm{\&mu;}}_k\right)S^{\&prime;T}\left({\mathrm{\&mu;}}_{k+1}\right)|}^2$ Calculate received signal and when k, be carved into k+1 constantly from the possible state μ of institute _kTransfer to state μ _K+1Path branches tolerance increment λ (μ _k→ μ _K+1), in the formula: symbol () ^TRepresent the transposition computing, S ' (μ _K+1) represent the received signal institute might state μ _K+1The direct judgement sequence { s of Ying She a L information code element _K+1-(L-1)' ... s _k', s _K+1';

3. by the path branches tolerance increment λ (μ that obtains _k→ μ _K+1), received signal institute might state μ constantly at existing k _kPath branches tolerance Γ (μ _k) the basis on, utilize formula $\mathrm{\&Gamma;}\left({\mathrm{\&mu;}}_{k+1}\right)=\underset{{\mathrm{\&mu;}}_k}{\min }[\mathrm{\&Gamma;}\left({\mathrm{\&mu;}}_k\right)+\mathrm{\&lambda;}({\mathrm{\&mu;}}_k\&RightArrow;{\mathrm{\&mu;}}_{k+1})]$ Obtain k+1 received signal sequence { d constantly _K+1, d _kD _K-(L-2)Institute might state μ _K+1Path branches tolerance Γ (μ _K+1);

4. calculating k+1 real-time status constantly by following formula deletes and moves thresholding Th _K+1 ^Max,

${\mathrm{Th}}_{k+1}^{\max }\&ap;{C_1}^2{N}_0\&CenterDot;e^{-\frac{(k+1)T_s}{\mathrm{\&tau;}}}$

Wherein, T _sBe symbol width; N ₀Be the power spectral density of white Gaussian noise, τ is the attenuation coefficient of meteoric trail; C ₁Be the link constant of meteor communication, the correct status that it and communication system are set is deleted and is moved probability P _c ^rFollowing relational expression is arranged,

$P_c^r=\frac{1}{2}\mathrm{erfc}\left(C_1\right)$

Erfc () is the compensating error function in the formula, so C ₁Can be by P _c ^rObtain by looking into the compensating error function table;

5. the path branches that is 3. obtained by step is measured Γ (μ _K+1), find out k+1 and have minimal path branch metric Γ (μ constantly _K+1) state μ _K+1 ^Min, the real-time status that is 4. obtained by step is deleted and is moved thresholding Th again _K+1 ^Max, with branch metric greater than Γ (μ _K+1 ^Min) and thresholding Th _K+1 ^MaxThe state of sum is left out, and the state of finishing this moment is deleted and moved;

6. to the state μ of minimum branch metric _K+1 ^MinCarry out the path and recall, export the constantly corresponding decision signal s of this k+1 _(k+1)-δ', wherein δ is the system determination degree of depth.

2. method according to claim 1, wherein the 1. described employing self adaptation of step RLS method is to k moment received signal sequence { d _k, d _K-1D _K-(L-1)Institute might state μ _kThe channel parameter vector estimate, be to draw the constantly corresponding channel parameter estimation vector of k by following four formulas

$e\left({\mathrm{\&mu;}}_k\right)=d_k-{\hat{h}}^{*}\left({\mathrm{\&mu;}}_{k-1}\right)S^{\&prime;T}\left({\mathrm{\&mu;}}_k\right)$

$g\left({\mathrm{\&mu;}}_k\right)=\frac{p\left({\mathrm{\&mu;}}_{k-1}\right)S^{\&prime;T}\left({\mathrm{\&mu;}}_k\right)}{w+S^{\&prime;*}\left({\mathrm{\&mu;}}_k\right)p\left({\mathrm{\&mu;}}_{k-1}\right)S^{\&prime;T}\left({\mathrm{\&mu;}}_k\right)}$

$p\left({\mathrm{\&mu;}}_k\right)=\frac{1}{w}[p\left({\mathrm{\&mu;}}_{k-1}\right)-g\left({\mathrm{\&mu;}}_k\right)S^{\&prime;*}\left({\mathrm{\&mu;}}_k\right)p\left({\mathrm{\&mu;}}_{k-1}\right)]$

$\hat{h}\left({\mathrm{\&mu;}}_k\right)=\hat{h}\left({\mathrm{\&mu;}}_{k-1}\right)+g\left({\mathrm{\&mu;}}_k\right)e^{*}\left({\mathrm{\&mu;}}_k\right)$

In the formula: symbol () ^*Represent conjugate operation, e (μ _k) be error signal, i.e. received signal d _kWith the desired output signal Difference, w is a forgetting factor, g (μ _k) be the kalman gain vector, p (μ _k) be correlation matrix, S ' (μ _k) represent k state μ constantly _kFor receiving data sequence { d _k, d _K-1D _K-(L-1)Judgement sequence { s _K-(L-1)' ... s _K-1', s _k';

When k=1 begins to start computing for the first time, p (μ _k) initial value p (μ ₀) get unit matrix,

Initial value be that channel parameter first estimated vector, promptly $\hat{h}\left({\mathrm{\&mu;}}_0\right)={\hat{h}}_0.$

3. method according to claim 1, the wherein 6. described state μ of step to k+1 minimum branch metric of the moment _K+1 ^MinThe detailed process of carrying out recalling in the path is: if the system determination degree of depth is δ, at first find out arrival state μ _K+1 ^MinPrevious moment, i.e. k state μ constantly _k, and then find out arrival state μ _kThe previous moment state, i.e. k-1 state μ constantly _K-1And the like, find (k+1)-δ state μ constantly always _(k+1)-δ, it is carried out bit decision, export corresponding decision signal s of this moment _(k+1)-δ'.

Images