Research On Unequal Error Protection Rateless Codes In Deep Space Communication

Deep space exploration is the focus all over the world as the third milestone of thedevelopments of aerospace science and technology. With the space exploration plans byvarious countries made, the data volume and business types are increasing with eachpassing day. Especially, the image data of exploration transmitted back on down link ofdeep space channel, is the most visible and important information for cognitions of deepspace. Meanwhile, in the manned space mission, the transmission of voice and videobetween interagency task systems are the necessary guarantee for mission accomplishedand personnel monitors.For efficient and reliable transmissions of multimedia business, the forwarderasure correction on the application layer attracts extensive attentions. Therateless codes, LT codes, have been employed in deep space communicationswith their characteristics of low complexities and needless of priori knowledgeof channel state. With the research of rateless codes developing, this work focusesto how to further promote the quality of service (QoS) of the deep space data.Firstly, based on preliminary investigation, an unequal error protection (UEP)scheme based on independent window LT (IW-LT) code is proposed. Accordingto the different levels of importance, the original data are divided intoindependent windows and encoding by LT codes. The asymptotic decodingperformance of IW-LT code is derived by AND-OR tree analysis technique.Moreover, to accurately match the expected UEP gain and decoding overhead,multi-objective genetic algorithms are introduced to optimize the degreedistribution and probabilities of selecting windows, which is enhanced theflexibility of IW-LT UEP scheme. The simulation results show that, compared toexisting rateless coding UEP scheme, IW-LT code can obtain lower decodingerror rate under the fixed parameters of UEP property.Secondly, due to the high-speed mobile of spacecraft, astral periodic motionand random interferences in the interstellar space, the bit error ratio of deepspace link is fluctuated widely, even caused interrupt easily. The insufficientnumber of output symbols can be obtained at the receivingterminal, which onlyshould be used for recovery as more as best original packets. Therefore, it is ofinterest to understand the intermediate performance-how many inputs can berecovered by the decoder from an insufficient number of output symbols. In thispaper, we employ NSGA-II multi-objective genetic algorithms optimizationmethod to find several degree distributions with optimum packet recovery rates,and change the transmission order of the encoded symbols according to their structure, their transmission history, and an estimate of the channel’s erasure rate.As the simulation results, we showed that the proposed UEP LT algorithm canincrease the intermediate recovery rate of rateless codes significantly, to improvethe data real-time recovery rate on the random interruption channel of deep spaceenvironments.