Study of a protocol and a priority paradigm for deep space data communication

Communicating data in deep-space entails the following constraints: large signal propagation delays---on the order of minutes or hours, coupled with the disruptions due to planetary orbital dynamics high channel error characteristics scarce and asymmetric data bandwidth availability, etc. Communication protocols developed for the terrestrial Internet perform poorly, or not at all, in this environment. We have co-designed a protocol, namely the Licklider Transmission Protocol (LTP), for reliable data communications in deep-space. We compare its performance to that of the well-known terrestrial Internet protocol---TCP---for various communication channel characteristics.Deep-space missions often generate much more data for communication to Earth than what is allowed by the channel data rate. Therefore, it becomes important to decide which data to send, and when this is done largely as a manual process at present. To help address this problem, we propose a two-dimensional priority paradigm for applications, aimed at optimizing the overall data communication performance. The two dimensions are: Intrinsic Value, a measure of how innately valuable the application data is, and Immediacy, a measure of how urgently a unit of application data needs to be communicated. We integrate this priority paradigm with LTP, and study candidate Forward Error Correction (FEC) mechanisms for implementing the paradigm such as Convolutional codes, Reed-Solomon codes, and Digital Fountain codes, for various channel characteristics. Finally, we recommend appropriate FEC mechanisms for the priority requirements of applications under different channel characteristics to optimize the volume and value of the data received.