Research On Rateless Codes For Space Based Internet Of Things With Ultra-reliable Transmission

With the expansion of the scope of crucial areas,and the continuous improvement of the demand for response capabilities,the transmission of short packet applications such as sensing data and signaling has expanded from terrestrial communication networks to space-based internet of things(S-IoT).The strong channel codes in short block-length regime is the fundamental guarantee for realizing ultra-reliability transmission in S-IoT.However,due to the environmental constraints such as the difficulty of accurate acquisition of the channel state information(CSI),the power constraint at the S-IoT receiver/transmitter,and the severe fading of the large propagation distance,the existing short block-length channel coding scheme cannot meet the ultra-reliability requirements of S-IoT.To solve these challenges,the thesis aiming to improve the support capability of coding technology in S-IoT communication systems,which is limited by the difficulties of channel coding and decoding in short block-length regime.Based on the coding optimization theory and the coding performance evaluation method in finite block-length,a novel rateless coding scheme for S-IoT is developed.A rateless coding structure based on the cascaded codes is proposed to solve the error performance requirement in short block-length regime.Coding gain exploit and complexity optimization for rateless coding and decoding in the low signal-to-noise ratio(SNR)regions are studied to improve coding and decoding performance under S-IoT resource constraints.For satellite-ground communication scenarios,a practical retransmission control mechanism based on rateless coding is proposed to improve the efficiency in S-IoT large-scale propagation scenarios.Specifically,the following researches are carried out:Through the concatenation of non-binary channel coding scheme and analog fountain codes(AFC),a novel rateless coded modulation structure which can approach the finite block-length bound is proposed.According to the Hamming distance spectrum,an optimal degree distribution search strategy based on the length of information bits is designed,and a short block-length channel coding scheme that can approach the error performance limit in bit-level rate compatible mode is obtained.To achieve the performance approximation in the mid-to-high signal-to-noise ratio(SNR)with rateless characteristic,the AFC coding algorithm is also optimized.The proposed cascaded coding structure supports the improvement of the error performance of the rateless code in the physical channel.For the high-gain coding requirements under low SNR region,a weight-adaptive AFC coding scheme based on limited feedback is proposed.Through the theoretical derivation of AFC mutual information with different weight sets,the theoretical optimal weight selection strategy under different SNR regions is designed.For the high reliability requirements of short block-length AFC,the extrinsic information transfer(EXIT)chart given the block-length is analyzed.By modifying the weight selection strategy in short block-length regime,the modifyed strategy can improve the AFC coding gain without precise CSI,and achieve the approximation to the finite block-length bound with low coding complexity.The proposed scheme supports the requirement of rateless coding gain improvement under S-IoT resource constraints.Focusing on the proposed cascaded rateless coding structure,a novel ordered statistics decoding(OSD)based on the soft information of AFC decoding is proposed.The output distribution of AFC decoder is modeled by density evolution theory,OSD decoding optimization algorithm is designed based on the statistical characteristics of log-likelihood ratio(LLR),and cyclic redundancy check(CRC)is introduced to ensure the decoding performance in short block-length.Compared with the traditional rateless code decoding algorithm,the proposed algorithm significantly improves the decoding performance in short block-length regime.Compared with the traditional OSD algorithm,the proposed algorithm has the advantage of low complexity under the same decoding performance.The proposed support the high-reliability decoding performance requirements of S-IoT nodes with limited power.Combined with the proposed concatenated rateless coding and decoding optimization algorithm,a rateless transmission scheme based on delayed CSI feedback is proposed for the requirement of no retransmission in large-scale satellite-to-ground propagation scenarios.Through the Euclidean distance analysis,the low peak-to-average power ratio(PAPR)and high reliability constellation mapping of AFC is proposed.The achievable rate with block-length constraint is deduced from the information density,and the modulation and coding schemes(MCSs)under the fading channel are designed.By adapting to the transmission mechanism based on delayed CSI regions,the proposed rateless transmission scheme can improve the capability of reliable transmission in typical S-IoT scenarios,while achieving near-capacity performance with few retransmission,which support the improvement of the applicability of the rateless coding in the S-IoT.