LAVIE - Uma arquitetura híbrida LEO-GEO para a internet interplanetária

Abstract

Space has been a source of unknowns for humanity since ancient times. The difficulty in studying it and its potential for discoveries fuel the desire for more knowledge. With time and technological advances, it became possible to send different equipment to space, such as rockets, satellites and probes. This brought opportunities for several applications, such as satellite communication, climate and disaster monitoring, geolocation systems and even space exploration and data collection missions. With this, the need arises for a structured network for sending and receiving this data, which was called Interplanetary Internet. However, space is not a conducive environment for the operation of computer networks, involving problems such as long distances between links, large and variable delays, intermittent connections and extreme weather conditions. In addition, the communication model currently used is inadequate for broader use, as the data is sent directly from the antennas on terrestrial ground to the space equipments, which requires a great synchrony in the line of sight of the sending and receiving parties. Several authors have developed solutions that aim to improve interplanetary communication, although the vast majority are planned for specific proprietary applications and with tests that evaluate only a specific metric. This dissertation work proposes the LAVIE architecture, an architecture based on hybrid satellite constellations, for interplanetary communication that improves the transfer of images with spatial links and is a model of easy replicability to other planets. The proposal is validated through a performance analysis comparing LAVIE with the interplanetary communication model currently used. Where promising results were obtained regarding heavy communication at LAVIE, reaching a maximum of 47% of bandwidth usage and 1477 Kb/s of throughput, very acceptable values when compared to traditional architecture. In addition, the data were measured with a repetition factor equal to 50, to obtain a standard deviation and confidence interval with a confidence level of 95%, which showed no more than 1% variation, indicating reliable data.