A Cross-layer Optimization Of High Efficient Transmission For Deep Space Detected Image

In the information obtained by deep space exploration,the deep space detected image information that transmitted through the downlink is the most intuitive and core information of human beings to understand and explore space.The deep space link with characteristics of long delay,easy interruption and complex time-varying causes great challenge to the downlink image transmission.Therefore,it is necessary to study the efficient and reliable image transmission scheme for deep space detection image,which plays an important role in the development of deep space exploration.By introducing cross-layer transmission design,it can bring a large transmission performance gain to the terrestrial communication system.However,in the field of deep space communication,the study and application of the cross-layer transmission is less.To improve the transmission performance of the downlink image transmission in deep space communications,this paper proposes a cross-layer image transmission scheme.The proposed scheme is based on the delay tolerant network framework.In detail,the compressed sensing(CS)is adopted in the application layer for image compression,spinal codes is adopted in the physical layer for channel decoding,and licklider transmission protocol(LTP)is used in the transport layer for transmission control.Meanwhile,due to the limited error correction ability of the underlying channel,and the CS has certain error-tolerant decoding ability.Based on the above cross-layer transmission scheme,this paper introduces the error-tolerant transmission strategy,using the error-tolerant ability of the CS reconstruction,share the pressure with the process of Spinal decoding.Furthermore,the error-tolerant scheme of cross-layer transmission is proposed.By jointly optimizing across the application,the transport,and the physical layers,the transmission model that maximizes transmission throughput to meet minimum reconfiguration quality is established.At the same time,the Gilbert-Elliot model is introduced.For different channel states,the problem is solved by optimizing the proposed models.Based on the optimization of transmission model,the channel state prediction is used,and the transmission parameters are dynamically adjusted.The proposed transmission scheme realizes the high efficiency of the downlink image transmission.In order to verify and evaluate the proposed scheme,this paper builds a semi-physical simulation platform for cross-layer transmission.The simulation platform adopted the MVC design pattern,including the construction of the client,the engine and the link.The software implementation of cross-layer transmission scheme is carried out.Then the hardware platform implementation of the deep space detected image compression and transmission system is implemented,and the construction of the semi-physical simulation platform is completed by combining the software with the hardware platform.The simulation experiment used the Earth-Mars links as the communication scenarios to simulate the downlink of image transmission.Under the different channel state,the simulation experiments in the different transmission schemes have been done.And the performance of the proposed scheme compared with a variety of transmission scheme has been evaluated.Simulation results show that the proposed scheme which adopts error-tolerant transmission strategy can significantly improve the performance of transmission efficiency based on comparisons with the other schemes.In particular,compared with the typical transmission scheme which is commonly used in existing deep space communications,the throughput is improved significantly under the same conditions.