CN113922868A

CN113922868A - DVB-RCS2 RL three-turbine iterative receiver and receiving method

Info

Publication number: CN113922868A
Application number: CN202111515934.5A
Authority: CN
Inventors: 王捷; 缪开济
Original assignee: Nanjing Xianfeng Shuotong Wireless Technology Co ltd
Current assignee: Nanjing Xianfeng Shuotong Wireless Technology Co ltd
Priority date: 2021-12-13
Filing date: 2021-12-13
Publication date: 2022-01-11
Anticipated expiration: 2041-12-13
Also published as: CN113922868B

Abstract

The invention provides a DVB-RCS2 RL three-turbine iterative receiver and a receiving method, which are characterized in that on the basis of a Mesh RCST RL LM receiver, the assumed posterior probability log-likelihood ratio information of a payload symbol output by a DVB-RCS2 RL LM Turbo code decoder in an iterative manner is fully utilized, and three turbine iterative operations of iterative synchronization, iterative equalization and iterative noise interference cancellation are executed in parallel. The invention eliminates carrier frequency offset, carrier phase offset, timing deviation and amplitude variation by carrying out iterative synchronization and iterative channel estimation, realizes iterative equalization, and reduces the influence of adjacent satellite interference, ACI and IMI by using iterative noise interference estimation and cancellation technology. In addition, the present invention may also mitigate the effects of CCI using iterative noise interference estimation and cancellation techniques, thereby allowing for efficient multiple access schemes with multiple users transmitting at the same frequency or time slot.

Description

DVB-RCS2 RL three-turbine iterative receiver and receiving method

Technical Field

The invention belongs to the technical field of satellite communication, and particularly relates to a DVB-RCS2 RL three-turbine iterative receiver and a receiving method.

Background

A Satellite Digital television broadcast Return Channel (DVB-RCS) broadband Satellite communication system providing various interactive services generally consists of geostationary orbiting satellites operating in the C, X, Ku and Ka bands and a large number of Return Channel Satellite Terminals (RCSTs) distributed over a wide geographical area and employing small aperture antennas, the RCSTs communicating with HUB gateways via satellites. To efficiently utilize system network resources, all terminals use a multi-frequency time division multiple access (MF-TDMA) shared satellite return channel (DVB-RCS), where each terminal allocates a limited frequency time slot for burst (Bursts) transmissions during a limited time period. In order to effectively reduce the transmission delay of satellite communication, the DVB-RCS2 standard is designed to support RCST to have a Mesh function.

One characteristic of small aperture antennas is the antenna beam width, so Return Link (RL) receivers of the RCST supporting Mesh functionality using small aperture antennas may be subject to interference from neighboring satellites. Furthermore, in mobile satellite communications applications, small aperture antenna terminals mounted on aerial aircraft, marine vessels, trains, or ground vehicles may produce antenna pointing errors, which may also potentially exacerbate adjacent satellite interference. These combined effects result in the small aperture antenna Mesh RCST RL receiver being subject to time-varying interference from another satellite network.

The DVB-RCS RL channel of broadband satellite communication is also a fading channel, and particularly under the conditions that RCST moves and shading cannot be directly projected, the problem of multipath time delay in the channel becomes more prominent; the transmission bandwidth of the signal may be greater than the coherence bandwidth of the channel, such that broadband satellite communications over the DVB-RCS RL channel are subject to severe frequency selective fading. RCST movement also causes doppler shift of the signal carrier.

Satellite-borne high-power amplifiers (HPAs) considered for satellite use have typical traveling-wave tube non-linearity characteristics that cause distortion in the phase and amplitude of the transmitted signal constellation points. More specifically, the phase and amplitude response of a satellite non-linear power amplifier depends primarily on the envelope of its input signal. Due to random phase, frequency and power deviations between RL signals from different RCSTs, the input signal envelope of the satellite HPA varies from Block (Block) to Block, causing the satellite behavior to vary from one Block to the next.

The non-linearity of HPA also causes the Adjacent Channel Interference (ACI) and intermodulation interference (IMI) of DVB-RCS RL MF-TDMA to be exacerbated. In addition, co-channel interference (CCI) also exists when systems are designed to employ efficient multiple access schemes that allow multiple users to use the same frequency or time slot at the same time. In addition, the RL receive signal supporting the Mesh function RCST may also be subject to carrier phase noise caused by the radio frequency Local Oscillator (LO) and the receiver front-end tuner. Due to the limited bandwidth of the RL channel, the phase noise samples associated with successive symbols in a block may experience greater variation.

By combining the above factors, carrier frequency offset, carrier phase offset, timing deviation and amplitude change occur in the process of transmitting DVB-RCS RL signals, and the signals reach a receiving end through a satellite channel (including nonlinearity of satellite-borne HPA) and are seriously interfered, so that the transmission performance is reduced. For this purpose, a high-performance Mesh RCST RL receiver must be specifically and specifically designed.

The DVB-RCS2 standard specifies some pilot symbol assisted burst formats, with large pilot interval periods, and some bursts with extremely limited pilot symbols. The traditional Mesh RCST RL receiver only uses pilot frequency symbols to estimate frequency deviation, phase deviation and timing deviation, carries out channel estimation required by single carrier equalization, and has poor performance. It is well known that for synchronization with pilot symbols, the pilot period should be less than 17 symbols at a frequency offset of up to 3% of the symbol rate, otherwise 2 consecutive pilot symbols will experience a mutual phase difference exceeding that experienced by the other symbols

(ii) a And the Cramer-Rao bound of the modification of the estimated variance depends on the number of symbols used; thus, the performance of DVB-RCS RL burst synchronization with only a very limited number of pilot symbols is poor, as is the channel estimation, resulting in poor performance of the conventional Mesh RCST RL receiver.

Disclosure of Invention

Aiming at the defects in the prior art, the invention provides a DVB-RCS2 RL three-turbine iterative receiver and a receiving method, which make full use of APP LLR information soft symbols of payload symbols in bursts iteratively calculated by a Turbo code decoder of RL LM-TC specified by DVB-RCS2 standard, execute three turbine iterative operations of iterative synchronization, iterative equalization and iterative noise interference cancellation in parallel, eliminate carrier frequency offset, carrier phase offset, timing deviation and amplitude variation, realize iterative equalization, reduce the influence of adjacent satellite interference, ACI, IMI and CCI, and improve the transmission reliability and spectral efficiency.

The present invention achieves the above-described object by the following technical means.

A DVB-RCS2 RL three-turbine iterative receiving method comprises the following processes:

DVB-RCS2 RL LM received signals sampled by ADC is buffered by a burst buffer, then is subjected to spectrum inversion processing by a spectrum inversion module, then is subjected to frequency offset and phase difference correction by an NCO frequency correction module, then is subjected to re-sampling module and extraction module to obtain a sampling sample of 4 times of the actual symbol rate, then the sampling sample is transmitted to a matched filter for matched filtering, and then a timing tracking module adopts local known pilot sequences pre, post and p to correlate an input symbol sequence to realize coarse timing synchronization;

removing pilot frequency sequences pre, post and p in the symbol sequence subjected to coarse timing synchronization, converting the symbol sequence subjected to removal processing into a frequency domain through an FFT module, then carrying out frequency domain equalization through an equalizer module comprising a feedforward FDE part and a feedback FDE part, then converting the symbol sequence into a time domain through an IFFT module, carrying out interference noise cancellation, and then storing symbol demapping conversion bits soft quantity LLRs into a code block buffer by a symbol demapping module by adopting a corresponding constellation type;

decoding, wherein before decoding, a reverse puncturing module performs reverse puncturing operation on the input of an iterative decoder with different code rates, recovers the bit stream length of each redundant bit Y and W of each burst Turbo code, then transmits the bit stream lengths to the iterative decoder, and performs DVB-RCS2 standard Turbo code decoding;

LLRs of each pair of Bits output by each iterative decoding of the iterative decoder are stored in an LLR RAM by an LLR Bits operator, on one hand, the LLRs are directly used for an interference noise estimation code block to carry out iterative estimation on the superposed noise and the statistical characteristics of adjacent satellite interference, ACI, IMI and CCI to realize iterative noise cancellation and iterative interference cancellation, on the other hand, a symbol mapping module generates modulation symbols according to constellations of different modulation schemes stored in a mapping LUT, the modulation symbols are used for a timing tracking module to realize iterative timing synchronization, and are converted into a frequency domain through an FFT module to be used for a frequency offset and phase joint iterative estimation module to carry out frequency offset and phase joint iterative estimation to realize iterative frequency and phase synchronization, and meanwhile, a fine channel estimation module is used for iterative equalization;

and after the set iteration times are reached or the stopping condition is met, hard judgment is carried out on LLRs of each pair of bits, decoding bits are output to a payload CRC (cyclic redundancy check) module for check comparison, and RLE (recursive least squares) processing is output after the correctness of receiving bursts after Turbo decoding is determined.

Further, the process of removing the pilot sequences pre, post, and p is as follows: when a non-spread spectrum mode is adopted, a symbol sequence subjected to coarse timing synchronization is removed from known pilot sequences pre, post and p by a pre, post and p demultiplexing module and is stored in a pre, post and p buffer;

when a spread spectrum mode is adopted, a de-spreading scrambling code and de-duplication module is required to descramble an input burst symbol sequence to obtain the whole burst after de-spreading scrambling code, a pre, post and p de-duplication module is used for removing known pilot frequency sequences pre, post and p and storing the known pilot frequency sequences in a pre, post and p buffer, and the removed symbol sequence is subjected to average de-duplication on received repeated symbols to obtain a de-spread symbol sequence;

the pilot sequences pre, post and p stored in the pre, post and p buffers are transformed to the frequency domain by the FFT module, and are used for the frequency offset and phase joint estimation module to carry out initial frequency offset and phase joint estimation, so as to realize coarse frequency and phase synchronization, and are used for the coarse channel estimation module to carry out initial channel estimation required by the equalizer.

A DVB-RCS2 RL three-turbine iterative receiver for realizing the DVB-RCS2 RL three-turbine iterative receiving method comprises a burst buffer, a spectrum inversion module, an NCO frequency correction module, a frequency offset and phase joint estimation module, a re-sampling module, an extraction module, a matching filter, a timing tracking module, an FFT module, an equalizer, a demultiplexing module, a channel estimation module, a despreading, scrambling and de-repeating module, a buffer, an IFFT module, a symbol demapping module, a code block buffer, a reverse puncturing module, an iterative decoder, an LLR Bits operator, a payload CRC check module, a symbol mapping module and an interference noise estimation module.

Further, the burst buffer buffers the burst sample output by the ADC; the frequency spectrum inversion module is used for realizing frequency spectrum inversion by multiplying the Q-path signal by-1; the NCO frequency correction module corrects the frequency deviation of the demodulation signal to keep the phase continuity; the frequency offset and phase joint estimation module adopts a frequency domain algorithm, applies APP LLR information of a bearing symbol in a burst, the APP LLR information is a soft symbol iteratively calculated by a Turbo decoder, the soft symbol and pre, post and p known symbols are all used for estimating frequency deviation and phase deviation, then a symbol sampling sample in the burst is corrected according to the latest estimation, and then the next round of iteration is executed.

Further, the re-sampling module reduces the input signal sampling rate by a given scaling factor; the extraction module is responsible for extracting input symbols according to the matched values to obtain 4 times of actual symbol rate; the matched filter performs matched filtering on the input signal from the decimation module; the timing tracking module resamples the output samples of the matched filter, so that the SNR of the specified number of sample samples of each symbol is maximum, the ISI is minimum, under various channel conditions with time-varying delay in a channel, the demodulator tracks the corrected sample samples, the timing tracking algorithm uses APP LLR information of a bearing symbol in a burst, the APP LLR information is a soft symbol iteratively calculated by a Turbo decoder, the soft symbol and the known symbols of pre, post and p are used for carrying out timing deviation estimation and symbol correction, and then the next iteration is executed.

Further, the FFT module is configured to transform the time domain signal to the frequency domain; the equalizer adopts a frequency domain equalizer and comprises a feedforward FDE part and a feedback FDE part; the demultiplexing module is a Pre, post and p demultiplexing module and is used for removing the known pilot sequences Pre, post and p from the input burst symbol sequence; the channel estimation module adopts a frequency domain algorithm to carry out channel estimation and is divided into two sub-modules of coarse channel estimation and fine channel estimation, the coarse channel estimation algorithm only adopts known pilot frequency sequences pre, post and p, the fine channel estimation algorithm uses APP LLR information of a bearing symbol in a burst, the APP LLR information is a soft symbol iteratively calculated by a Turbo decoder, the soft symbol and the known symbols of pre, post and p are used together for carrying out channel estimation, and then the next iteration is executed.

Furthermore, the de-spreading scrambling code and de-repeating module adopts a spread spectrum mode to descramble the input burst symbol sequence to obtain the whole burst after de-spreading scrambling code, and after known pilot frequency sequences pre, post and p are removed, the de-repeating is realized by averaging the received repeated symbols; the buffer is a pre, post and p buffer and is used for buffering known pilot sequences pre, post and p obtained by demultiplexing; the IFFT module transforms the frequency domain signal to a time domain; the demapping module performs symbol demapping operation by adopting a correct constellation type, and the conversion symbols are represented by LLRs; the code block buffer is used for buffering code block soft bits output to the iterative decoder; and the reverse puncturing module recovers the bit stream length of each redundant bit Y and W of each burst Turbo code by performing reverse puncturing operation on the input of the iterative decoder with different code rates.

Further, the iterative decoder performs Turbo code decoding in the DVB-RCS2 standard, inputs LLRs of received systematic bits and check bits into corresponding component decoders, calculates LLRs from receiver sample samples by considering channel gain and noise variance; decoding is performed in an iterative manner, with each iteration both component decoders updating the LLRs and interleaving or deinterleaving accordingly; exchanging the difference between the LLRs of the input and the output between the component decoders; and (4) performing further processing on the LLRs of each pair of bits output by each iterative decoding, and performing hard decision on the LLRs of each pair of bits after the set iteration times are reached or a stopping condition is met, so as to output a decoded bit.

Further, the LLR Bits operator stores the soft decision LLRs output by the decoder for each iteration into a RAM unit of an LLR table for use by a required module; the effective load CRC check module calculates burst-borne CRC for the bit sequence output by the iterative decoder, checks and compares the burst-borne CRC and the burst-borne CRC, and determines the correctness of the received frame decoded by the Turbo; the symbol mapping module generates modulation symbols according to constellations of different modulation schemes stored in the LUT; the interference noise estimation module estimates the statistical characteristics of the superimposed adjacent satellite interference, ACI, IMI, and CCI interference, and improves estimation performance by using the output of the previous iterative decoder.

The invention has the following beneficial effects:

the DVB-RCS2 standard stipulates RL Linear Modulation (LM) to adopt Turbo coding, so the invention fully utilizes the assumed posterior probability log-likelihood ratio (APP LLR) information of the payload symbols iteratively output by the DVB-RCS2 RL LM Turbo code decoder, namely soft symbols iteratively calculated by the Turbo decoder, and executes three Turbo iterative operations of iterative synchronization, iterative equalization and iterative noise interference cancellation in parallel on the basis of a Mesh RCST RL LM receiver. Specifically, the method comprises the steps of performing iterative synchronization and iterative channel estimation, eliminating carrier frequency offset, carrier phase offset, timing deviation and amplitude change, realizing iterative equalization, and simultaneously using an iterative noise interference estimation and cancellation technology to reduce the influence of adjacent satellite interference, ACI and IMI, thereby improving the transmission reliability of DVB-RSC2 RL, and allowing the use of higher-order modulation schemes with fewer pilot symbols; and secondly, the influence of CCI can be relieved by using the iterative noise interference estimation and cancellation technology, so that an efficient multiple access scheme that a plurality of users transmit in the same frequency or time slot is allowed to be adopted.

The DVB-RCS2 RL three-turbine iterative receiving method is adopted on the Mesh RCST, so that the purpose of improving the utilization efficiency of the DVB-RCS RL frequency spectrum can be achieved. Iterative equalization, iterative synchronization, iterative noise interference estimation and cancellation three-turbine iterative technology obviously improve the performance of the Mesh RCST RL LM receiver. And the iterative equalizer, the iterative synchronizer, the iterative noise interference estimation and counteractor and the Turbo decoder run in parallel, so that the throughput cannot be reduced.

Drawings

FIG. 1 is a first partial schematic diagram of a three-turbine iterative receiver according to the present invention;

fig. 2 is a second partial schematic diagram of a three-turbine iterative receiver according to the present invention.

Detailed Description

The invention will be further described with reference to the following figures and specific examples, but the scope of the invention is not limited thereto.

As shown in fig. 1 and 2, the DVB-RCS2 RL three-turbine iterative receiver according to the present invention includes a burst buffer, a spectrum inversion module, a Numerically Controlled Oscillator (NCO) frequency correction module, a frequency offset and phase joint estimation module, a re-sampling module, an extraction module, a matched filter, a timing tracking module, a Fast Fourier Transform (FFT) module, an equalizer, a demultiplexing module, a channel estimation module, a de-spreading scrambling and de-repeating module, a buffer, an Inverse Fast Fourier Transform (IFFT) module, a symbol de-mapping module, a code block buffer, a de-puncturing module, an iterative decoder, an LLR Bits operator, a payload CRC check module, a symbol mapping module, and an interference noise estimation module.

The burst buffer is used for buffering burst sampling samples sampled and output by the ADC; the frequency spectrum inversion module is used for realizing frequency spectrum inversion by multiplying the Q-path signal by-1, and solving the problem that a demodulator possibly generates an inverted frequency spectrum or a non-inverted frequency spectrum; the NCO frequency correction module is used to correct the frequency offset that the demodulated signal may have (up to half the ADC sampling frequency) to keep the phase continuous. The frequency offset and phase joint estimation module adopts a frequency domain algorithm, applies the posterior probability log-likelihood ratio (APP LLR) information of the bearing symbols in the burst, the information is soft symbols which are iteratively calculated by an iterative decoder (namely a Turbo decoder), the soft symbols and pre, post and p known symbols are all used for estimating frequency deviation and phase deviation, then symbol sampling samples in the burst are corrected according to the latest estimation, and the next round of iteration is executed.

The re-sampling module is used for reducing the sampling rate of the input signal according to a given scale factor; the extraction module is responsible for extracting input symbols according to the matched values to obtain 4 times of actual symbol rate; the matched filter is used for performing matched filtering on the input signal from the decimation module. The timing tracking module is used for resampling output samples of the matched filter, so that the SIGNAL-to-NOISE RATIO (SNR) of a specified number of sampling samples of each Symbol is maximum, intersymbol Interference (ISI) is minimum, the demodulator tracks the corrected (maximum SNR) sampling samples under various channel conditions with time-varying delay in a channel, the timing tracking algorithm uses APP LLR information of a carrying Symbol in a burst, the information is soft symbols calculated by Turbo decoder iteration, the soft symbols and pre, post and p known symbols together perform timing deviation estimation and Symbol correction, and then next iteration is performed.

The FFT module is used for transforming the time domain signal to a frequency domain; the equalizer adopts a frequency-domain equalizer (FDE) and is used for completing equipment for performing equalization compensation on time delay and frequency characteristic distortion of a data transmission channel, wherein the equipment comprises a feedforward FDE part and a feedback FDE part; the demultiplexing module is a Pre, post, p demultiplexing module, and is configured to perform processing of removing a preamble (Pre), a postamble (post), and a pilot (p) symbol from the input burst symbol sequence. The channel estimation module adopts a frequency domain algorithm to carry out channel estimation and is divided into a coarse channel estimation submodule and a fine channel estimation submodule; the coarse channel estimation algorithm only adopts pre, post and p known symbols; the fine channel estimation algorithm uses APP LLR information of the bearing symbols in the burst, the information is soft symbols iteratively calculated by the Turbo decoder, the soft symbols and pre, post and p known symbols are used for channel estimation, and then the next iteration is executed.

A despreading scrambling code and de-repeating module adopts a spread spectrum mode to descramble an input burst symbol sequence to obtain the whole burst (comprising a lead code, a pilot frequency and a postamble) after despreading scrambling code; after removing the known pilot pre, post and p symbols, the de-duplication is realized by averaging the received repeated symbols, and then the de-duplicated symbols are processed. The pre, post and p buffers are used for buffering the known pilot symbols of the pre, post and p obtained by the demultiplexing; the IFFT module is used for transforming the frequency domain signal to a time domain; the symbol demapping module is used for performing symbol demapping operation by adopting a correct constellation type, and converting symbols to be represented by Log Likelihood Ratio Statistics (LLRs); the code block buffer is used for buffering code block soft bits (data) output to the iterative decoder; and a De-puncturing (De-puncturing) module recovers the bit stream length of each redundant bit Y and W of each burst Turbo code by performing anti-puncturing operation on the input of the Turbo decoder with different code rates, wherein Y and W both represent redundant check bits output by the Turbo code encoder.

The iterative decoder is used for performing Turbo code decoding in DVB-RCS2 standard, inputting LLRs of received systematic bits and check bits into corresponding component decoders, and calculating the LLRs from receiver sampling samples by considering channel gain and noise variance; decoding is performed in an iterative manner, and at each iteration, both component decoders update the LLRs and must interleave or deinterleave accordingly; exchanging the difference between the LLRs of the input and the output, namely the extrinsic information, between the component decoders; each pair of bit LLRs output by iterative decoding can be used for further processing, and after the set iteration times are reached or the stop condition is met, each pair of bit LLRs is subjected to hard decision to output the decoded bit.

The LLR Bits operator is used for storing the soft decision LLRs output by the Turbo decoder iterated each time into an RAM unit of an LLR table for a needed module to use; the payload CRC check module is used for calculating burst-borne CRC for the bit sequence output by the Turbo decoder, checking and comparing the burst-borne CRC and the burst-borne CRC, and determining the correctness of the received frame after Turbo decoding. The symbol mapping module maps the data according to a Look-UP-Table (LUT) stored in a Look-UP-Table (LUT)

The constellation of the modulation schemes QPSK, 8-PSK, and 16-QAM generates modulation symbols.

Since it is usually difficult to obtain accurate channel information at the receiver side, it is difficult to completely suppress the influence of interference such as adjacent satellite interference, ACI, IMI, CCI, etc. only by using channel estimation; the conventional receiving method usually directly treats such interference signals as noise, and when the interference is strong, the interference cannot be reduced by increasing the transmission power, so as to obtain ideal performance. The interference noise estimation module estimates the statistical characteristics of the superimposed adjacent satellite interference, ACI, IMI, CCI and other interferences, and the iterative receiver can provide the output of the previous iterative decoder, so the interference noise estimation module can utilize the output of the previous iterative decoder to greatly improve the estimation performance, thereby further improving the detection and decoding performance.

The receiving method of the DVB-RCS2 RL three-turbine iterative receiver is shown in FIGS. 1 and 2, and because the overall schematic diagram of the DVB-RCS2 RL three-turbine iterative receiver is more complex, in order to show the invention more clearly, the overall schematic diagram of the DVB-RCS2 RL three-turbine iterative receiver is divided into two partial schematic diagrams of FIG. 1 and FIG. 2 for detailed description; wherein, the number "1" in fig. 1 and 2 both represents the signal transmitted from the demultiplexing module to the FFT module in the non-spread spectrum mode, the number "2" in fig. 1 and 2 both represents the signal after the known pilot sequence stored in the buffer is transformed to the frequency domain by the FFT module, and the number "3" in fig. 1 and 2 both represents the signal after the modulation symbol generated by the symbol mapping module is transformed to the frequency domain by the FFT module; the number "4" in fig. 1 and 2 indicates a modulation symbol generated by the symbol mapping module. The specific working process of the DVB-RCS2 RL three-turbine iterative receiver provided by the invention is as follows:

the DVB-RCS2 RL LM receiving signal sampled by ADC is buffered by a burst buffer, then the spectrum inversion module carries out spectrum inversion processing, the NCO frequency correction module carries out frequency offset and phase difference correction, then a re-sampling module and an extraction module obtain a sampling sample with 4 times of the actual symbol rate, then the sampling sample is transmitted to a matched filter for matched filtering, and then a timing tracking module adopts local known pilot sequences pre, post and p to carry out correlation on an input symbol sequence to realize coarse timing synchronization.

When a non-spread spectrum mode is adopted, a symbol sequence subjected to coarse timing synchronization is removed from known pilot sequences pre, post and p by a pre, post and p demultiplexing module, stored in a pre, post and p buffer, converted to a frequency domain by an FFT module, subjected to initial frequency offset and phase joint estimation by a frequency offset and phase joint estimation module to realize coarse frequency and phase synchronization, and subjected to initial channel estimation required by an equalizer by a coarse channel estimation module; when the spread spectrum mode is adopted, a de-spreading scrambling code and de-duplication module is required to descramble an input burst symbol sequence to obtain the whole burst (including a lead code, a pilot frequency and a post-pilot code) after de-spreading scrambling code, a pre, post and p de-duplication module is used for removing known pilot frequency sequences of pre, post and p and storing the known pilot frequency sequences in a pre, post and p buffer, and the removed symbol sequence is subjected to de-duplication by averaging received repeated symbols to obtain the de-spread symbol sequence.

Whether a non-spread spectrum mode or a spread spectrum mode is adopted, symbol sequences with known pilot sequences pre, post and p removed are transformed to a frequency domain through an FFT module, then frequency domain equalization is carried out through an equalizer module comprising a feedforward FDE part and a feedback FDE part, then the frequency domain equalization is transformed to a time domain through an IFFT module, interference noise cancellation is carried out, and then a symbol demapping module adopts a correct constellation type to store symbol demapping conversion bits soft quantity LLRs into a code block buffer;

decoding, wherein before decoding, a reverse puncturing module performs reverse puncturing operation on the input of a Turbo decoder with different code rates to recover the bit stream length of each redundant bit Y and W of each burst Turbo code, and then the Turbo code is transmitted to an iterative decoder, namely the Turbo code decoder, and the Turbo code decoding of DVB-RCS2 standard is performed;

LLRs of each pair of Bits output by each iterative decoding of the Turbo decoder are stored in an LLR RAM by an LLR Bits operator, and can be directly used for iterative estimation of statistical properties of superimposed noise, adjacent satellite interference, ACI, IMI, CCI and the like by an interference noise estimation code block to realize iterative noise interference cancellation; or by symbol mapping module according to values stored in mapping LUT

QPSK, 8-PSK, 16-QAM and the likeA constellation of the modulation scheme generates a modulation symbol, on one hand, the modulation symbol is used for a timing tracking module to realize accurate iterative timing synchronization, on the other hand, the modulation symbol is used for a frequency offset and phase joint estimation module to perform frequency offset and phase joint iterative estimation after being converted to a frequency domain through an FFT module, so that accurate iterative frequency and phase synchronization is realized, and meanwhile, the modulation symbol is used for a fine channel estimation module to perform accurate iterative channel estimation and is used for iterative equalization. Therefore, the invention can realize the iterative synchronization, the iterative equalization and the iterative noise interference cancellation, and the three Turbo iterative operations and the parallel operation of the Turbo decoder.

And after the set iteration times are reached or the stopping condition is met, carrying out hard judgment on LLRs of each pair of bits, outputting decoding bits to a payload CRC (cyclic redundancy check) module, carrying out check comparison, determining the correctness of the received burst after Turbo decoding, and outputting to carry out RLE (recursive least squares) processing.

The present invention is not limited to the above-described embodiments, and any obvious improvements, substitutions or modifications can be made by those skilled in the art without departing from the spirit of the present invention.

Claims

1. A DVB-RCS2 RL three-turbine iterative receiving method is characterized by comprising the following processes:

2. The DVB-RCS2 RL three-turbine iterative reception method according to claim 1, wherein the pilot sequences pre, post, p are removed as follows: when a non-spread spectrum mode is adopted, a symbol sequence subjected to coarse timing synchronization is removed from known pilot sequences pre, post and p by a pre, post and p demultiplexing module and is stored in a pre, post and p buffer;

3. A DVB-RCS2 RL three-turbine iterative receiver implementing the DVB-RCS2 RL three-turbine iterative reception method according to any one of claims 1 to 2, comprising a burst buffer, a spectrum inversion module, an NCO frequency correction module, a frequency offset and phase joint estimation module, a re-sampling module, an extraction module, a matching filter, a timing tracking module, an FFT module, an equalizer, a demultiplexing module, a channel estimation module, a despreading frequency scrambling and de-repetition module, a buffer, an IFFT module, a symbol demapping module, a code block buffer, an anti-puncturing module, an iterative decoder, an LLR Bits operator, a payload CRC check module, a symbol mapping module, an interference noise estimation module.

4. The DVB-RCS2 RL three-turbine iterative receiver of claim 3, wherein the burst buffer buffers burst sample samples output by the ADC sampling; the frequency spectrum inversion module is used for realizing frequency spectrum inversion by multiplying the Q-path signal by-1; the NCO frequency correction module corrects the frequency deviation of the demodulation signal to keep the phase continuity; the frequency offset and phase joint estimation module adopts a frequency domain algorithm, applies APP LLR information of a bearing symbol in a burst, the APP LLR information is a soft symbol iteratively calculated by a Turbo decoder, the soft symbol and pre, post and p known symbols are all used for estimating frequency deviation and phase deviation, then a symbol sampling sample in the burst is corrected according to the latest estimation, and then the next round of iteration is executed.

5. The DVB-RCS2 RL three-turbine iterative receiver of claim 3, wherein the re-sampling module reduces an input signal sampling rate by a given scaling factor; the extraction module is responsible for extracting input symbols according to the matched values to obtain 4 times of actual symbol rate; the matched filter performs matched filtering on the input signal from the decimation module; the timing tracking module resamples the output samples of the matched filter, so that the SNR of the specified number of sample samples of each symbol is maximum, the ISI is minimum, under various channel conditions with time-varying delay in a channel, the demodulator tracks the corrected sample samples, the timing tracking algorithm uses APP LLR information of a bearing symbol in a burst, the APP LLR information is a soft symbol iteratively calculated by a Turbo decoder, the soft symbol and the known symbols of pre, post and p are used for carrying out timing deviation estimation and symbol correction, and then the next iteration is executed.

6. The DVB-RCS2 RL three-turbo iterative receiver of claim 3, wherein the FFT module is configured to transform the time domain signal to the frequency domain; the equalizer adopts a frequency domain equalizer and comprises a feedforward FDE part and a feedback FDE part; the demultiplexing module is a Pre, post and p demultiplexing module and is used for removing the known pilot sequences Pre, post and p from the input burst symbol sequence; the channel estimation module adopts a frequency domain algorithm to carry out channel estimation and is divided into two sub-modules of coarse channel estimation and fine channel estimation, the coarse channel estimation algorithm only adopts known pilot frequency sequences pre, post and p, the fine channel estimation algorithm uses APP LLR information of a bearing symbol in a burst, the APP LLR information is a soft symbol iteratively calculated by a Turbo decoder, the soft symbol and the known symbols of pre, post and p are used together for carrying out channel estimation, and then the next iteration is executed.

7. The DVB-RCS2 RL three-turbine iterative receiver of claim 3, wherein the despreading and scrambling module and the de-repeating module use a spreading mode to descramble an input burst symbol sequence to obtain a whole burst after despreading and scrambling, and after removing known pilot sequences pre, post, and p, the de-repeating is realized by averaging received repeated symbols; the buffer is a pre, post and p buffer and is used for buffering known pilot sequences pre, post and p obtained by demultiplexing; the IFFT module transforms the frequency domain signal to a time domain; the demapping module performs symbol demapping operation by adopting a correct constellation type, and the conversion symbols are represented by LLRs; the code block buffer is used for buffering code block soft bits output to the iterative decoder; and the reverse puncturing module recovers the bit stream length of each redundant bit Y and W of each burst Turbo code by performing reverse puncturing operation on the input of the iterative decoder with different code rates.

8. The DVB-RCS2 RL three-Turbo iterative receiver according to claim 3, wherein the iterative decoder performs Turbo code decoding in the DVB-RCS2 standard, inputs LLRs of received systematic bits and check bits into the corresponding component decoders, calculates LLRs from receiver sample samples by considering channel gain and noise variance; decoding is performed in an iterative manner, with each iteration both component decoders updating the LLRs and interleaving or deinterleaving accordingly; exchanging the difference between the LLRs of the input and the output between the component decoders; and (4) performing further processing on the LLRs of each pair of bits output by each iterative decoding, and performing hard decision on the LLRs of each pair of bits after the set iteration times are reached or a stopping condition is met, so as to output a decoded bit.

9. The DVB-RCS2 RL three-turbo iterative receiver of claim 3, wherein the LLR Bits operators store soft-decision LLRs output by the decoder for each iteration into RAM cells of an LLR table for use by required modules; the effective load CRC check module calculates burst-borne CRC for the bit sequence output by the iterative decoder, checks and compares the burst-borne CRC and the burst-borne CRC, and determines the correctness of the received frame decoded by the Turbo; the symbol mapping module generates modulation symbols according to constellations of different modulation schemes stored in the LUT; the interference noise estimation module estimates the statistical characteristics of the superimposed adjacent satellite interference, ACI, IMI, and CCI interference, and improves estimation performance by using the output of the previous iterative decoder.