Research On The Spinal Code Based Transmission Mechanism For Deep Space Communications

As the space technology develops rapidly, many countries pay more attention to the deep space exploration. And the deep space communications play an important role in the deep space exploration. In the deep space, high bit error rate, long time delay and so on are its remarkable features, consequently, There are very huge challenges in the deep space communications. Faced with the challenges, in additional to the traditional methods( increase the size of receive/send antenna, raise the carrier frequency, enhance the sending power), efficient channel encoding/decoding technique and reasonable transmission mechanisms are very important.Spinal code is proposed in 2012. It is a new class of rateless code which can be transmitted closed to the capacity limitation in both the BSC and AWGN channel. Compared with the coding algorithm of the conventional high gain fixed rate code such as LDPC code, Spinal code obtains a better performance in a very wide range of SNR especially in the low SNR in both the BSC and AWGN channel, moreover,its complexity of encoding and decoding is linear to the length of the message. Consequently, this paper chooses the Spinal code as the channel encoding technique, combining with the data compression in the application layer, data correction in the transmission layer,Spinal encoding in the data link layer and physical layer and LTP(Licklider Transmission Protocol),and designs a DTN(Delay/Disruption Tolerant Network)-oriented cross-layer transmission mechanism to ensure the deep space data communicates efficiently.This paper uses Markov prediction and the feedback to adjust the sending strategy dynamically for sending data continuously without waiting the feedback. Then, This paper builds a cross-layer joint optimization model. Taking the deep space images as the example, combining the image compression in the application layer with image correction in the transmission layer,Spinal encoding in the data link layer and physical layer is to make the symbols each image needed least. In the application layer and the transmission layer,CS(Compressed sensing) is used to compress images. Compared with the conventional compression, The CS’s encoding complexity is lower and can realize image compression efficiently, at the same time, it also has the error correcting capability. Consequently, the CS can be used in the application layer and transmission layer. Finally, a cross-layer joint optimization transmission mechanism is designed in this paper. In this mechanism, the encoding symbols sent in each transmission is least. the sender does not wait for the feedback. When the sender receives the feedback, it will deal with the feedback, after that, according to the feedback and Markov prediction, the sender decides the sending strategy in the next moment. Consequently, this cross-layer transmission realizes continuous transmission of images data. There are five mechanisms in this dissertation. They are ideal mechanism, cross-layer optimized mechanism, prediction and retransmission mechanism, retransmission without prediction and additional symbols mechanism.According to the throughput result, The mechanism of cross-layer transmission is close to the ideal transmission mechanism. And compared with other transmission mechanisms, it has a 6.5% lead on prediction and retransmission mechanism, 13.9% lead on additional symbols mechanism without prediction and 20% lead on retransmission without prediction mechanism.