Research On Power Control Techniques For Formation Of Small Satelites

Space technology is an important manifestation of the country's comprehensivenational strength. With the recent successful launch of a large number of spacecraft andsatellites, marking the pace of human space exploration is moving. Formation of smallsatellite network also been rapid development in recent years, which means multiplesmall satellites form a "virtual satellite". The formation of small satellite networks inflight formation will change according to changes in the task and the environment,which calls orbital reconstruction, brings a high relative velocity. And conventionalpower control techniques cannot meet this situation. This paper mainly studies theformation of small satellite network power control technology.This paper studied the SINR balancing algorithm which is widely used in existingnetwork power control technology. In order to cope with the two cases during fight information of small satellite networks.1) When irradiated of small satellites from the sunis limited, its solar panels electricity supply is seldom. At this point needs to lose somecommunication performance at the expense of saving its transmit power.2) When thesmall satellite networks acquires for high-speed transfer rate or a low error ratetransmission, which need to ensure a high quality of communication by increasingSINR through increasing its transmit power, Game theory was introduced to deal withthese two issues. Two selection of utility function is made to meets the needs these cases:Koskie-Gajic Algorithm and Zhao Chenglin Algorithm, which can satisfy these twocases very well. In order to improve its convergence rate, dynamic game is approved,which can improve the convergence rate without changing the convergence propertiesResearch on algorithm performance comparison is also done.AR model is approved to deal with the power control deviation caused byhigh-speed moving happened during formation reconstruction, through which thefluctuation range of SINR is reduced. Then AR predictor is combined with dynamicgame to better deal with the power control deviation. The later player performs better ina dynamic game, which is also proved in this paper.