深空通信中基于Spinal码的传输机制研究

发布时间：2018-03-29 07:29

本文选题：深空通信　切入点：Spinal码　出处：《哈尔滨工业大学》2014年硕士论文

【摘要】：随着航天技术的发展,世界各国越来越重视深空探测。而深空通信在深空探测任务中起着关键作用。但深空环境的大尺度距离跨度、高动态、传播环境特性复杂等特征,对深空数据信息的高质、高效、高可靠实时传输提出了极大挑战。除了传统的应对方式(增加接收/发射天线尺寸、提高载波、增加发射功率等)之外,高效的信道编译码技术和合理的传输机制也起着关键的作用。新近提出的Spinal码是一种全新的、在BSC(二进制对称信道)、AWGN(加性高斯白噪声信道)上均能实现近容量限传输的无速率编码方式,相比传统的高增益固定速率编码(如LDPC码),Spinal码在极宽的信噪比范围内尤其是低信噪比情况下均获得了更好的性能,同时编译码复杂度远低于传统的高增益固定速率编码。鉴于此,本文将Spinal码引入深空通信中,并重点结合应用层数据压缩、传输层数据纠删和数据链路层/物理层Spinal编码以及LTP文件传输协议,设计了一种面向DTN(Delay/Disruption Tolerant Network)协议栈框架的跨层联合传输机制,以实现深空数据信息的可靠、高效传输。本文首先利用Markov预测和反馈信息实现了发送策略的动态调整,从而不等待反馈,持续发送数据。然后,本文建立了一种跨层联合优化模型。以探测图像为例,对应用层的图像压缩、传输层的数据纠删以及数据链路层/物理层的Spinal编码进行联合优化,使得传输每一幅图像所需的符号数最小。在应用层和传输层,利用了信息与应用数学领域近年来新提出的压缩感知(Compressed Sensing,CS)技术进行图像压缩。与传统压缩相比,CS压缩编码复杂度低,可以实现高效率压缩,同时,CS压缩还具有潜在的纠删功能,因此将CS应用到应用层的图像压缩和传输层的数据纠删中。最后,在基于Spinal码的Markov预测和跨层联合优化模型的基础上设计了一种跨层联合优化传输机制,使每次传输所发送的编码符号数最少,在传输过程中不等待反馈,收到反馈时处理反馈,根据反馈信息和Markov预测决定下一时刻的发送策略,从而持续发送图像数据。通过对吞吐量的仿真分析,将理想状态的收发无延时交互传输机制、本文所用的跨层联合优化传输机制和基于反馈重传的传输机制等机制作为对比,结果表明在延时巨大、误码率非常高的深空通信中,跨层联合传输机制接近于收发无延时交互传输机制,高于预测重传机制6.5%,比无预测追加机制高13.9%,比无预测重传机制高20%。
[Abstract]:With the development of space technology, countries in the world pay more and more attention to deep space exploration. Deep space communication plays a key role in deep space exploration missions. It poses a great challenge to the high quality, high efficiency and high reliability of real-time transmission of deep space data information. In addition to the traditional response methods (increasing the size of receiving / transmitting antennas, increasing carrier, increasing transmission power, etc.), Efficient channel coding and decoding technology and reasonable transmission mechanism also play a key role. Recently proposed Spinal code is a new one. In BSC (binary symmetric channel) AWGN (additive Gao Si white noise channel) can realize near-capacity limited transmission rate free coding. Compared with the traditional high gain fixed rate code (such as LDPC code / Spinal code), it has better performance in a wide range of signal-to-noise ratio (SNR), especially in the case of low signal-to-noise ratio (SNR). At the same time, the complexity of encoding and decoding is much lower than that of traditional high gain fixed-rate coding. In view of this, Spinal code is introduced into deep space communication, and combined with application layer data compression. Data erasure in transport layer, Spinal coding in data link layer / physical layer and LTP file transfer protocol, a cross-layer joint transmission mechanism for DTN(Delay/Disruption Tolerant Network) protocol stack framework is designed to realize the reliability of deep space data information. In this paper, the Markov prediction and feedback information are used to realize the dynamic adjustment of the transmission strategy, so that the data can be continuously transmitted without waiting for feedback. Then, a cross-layer joint optimization model is established, and the detection image is taken as an example. Image compression in application layer, data erasure in transmission layer and Spinal coding in data link layer / physical layer are jointly optimized to minimize the number of symbols needed to transmit each image. In this paper, the compressing sensing CS (compressed sensing) technique proposed in recent years in the field of information and applied mathematics is used for image compression. Compared with traditional compression, CS compression has lower complexity and can achieve high efficiency compression. At the same time, CS compression also has potential erasure function. Therefore, CS is applied to the application layer image compression and data erasure in the transmission layer. Finally, based on the Markov prediction and cross-layer joint optimization model based on Spinal code, a cross-layer joint optimization transmission mechanism is designed. The number of coded symbols transmitted in each transmission is minimized, and the feedback is processed when the feedback is received, and the transmission strategy at the next moment is determined according to the feedback information and the Markov prediction. Through the simulation and analysis of the throughput, we compare the ideal state transmission mechanism, the cross-layer joint optimization transmission mechanism and the feedback retransmission mechanism. The results show that in the deep space communication with huge delay and high bit error rate, the cross-layer joint transmission mechanism is close to the transceiver and non-delay interactive transmission mechanism, which is higher than the predictive retransmission mechanism 6.5, 13.9 higher than the non-predictive supplementary mechanism, and 20 times higher than the non-predictive retransmission mechanism.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN927.3

【参考文献】