Abstract: A selforganizing architecture for the control system of spacecraft is proposed to solve the problem that the computing efficiency is insufficient for many redundant computing units. Based on cloud computing and high speed interconnected net, the redundant backup of the single node is cancelled in the proposed architecture. The computing capability of whole system is the staunch supporter for every single node. If one node shuts down, its program code and data will be transferred to other working nodes dynamically to continue its functions. This selforganizing manner realizes the reconstruction of computing architecture for system with faults.

Key words: system architecture, selforganization, dynamic transfer, cloud computing