In this paper, several formation controllers are developed with constant and variable connected communication topology for six-degree-of-freedom (6-DOF) networked spacecraft flying in deep space with constant and time-varying communication delays considered. In the control algorithm, any explicit leader does not exist in the formation team and it is required that each spacecraft communicate only with its neighbors and therefore the required communication loads of the whole formation is greatly reduced. The proposed control strategies allow that each spacecraft can track its desired position and attitude and simultaneously the whole group moves to the desired formation and obtain its desired relative attitudes with its neighbors. Simulation results demonstrate the effectiveness of the proposed controllers.imaging and gravitational detection, such as Terrestrial Planet Finder (TPF) [3], LISA [4], Deep Space-3 [5], and DARWIN [6]. In this paper, a deep space formation flying is considered.The manners of cooperation between spacecrafts determine the architecture of a formation, which has been grouped in [7] into five main categories: leader-follower, behavioral, virtual structure, cyclic and MIMO. These categories are actually according to the communication model of the spacecraft in formation. In applications, a MIMO formulation gives better performance in the ideal case but communication amongst the spacecrafts is required [8]. Ref.[9] considered the synchronization problems for networked spacecraft in deep space based on a constant all-to-all communication topology but left communication delays unaddressed. In general, it is impractical to have every spacecraft communicate with every other spacecraft. On the other hand, the lack of sufficient number of communication channels may cause several problems such as deterioration of the overall control performance, inability to detect obstacles and formation reconfiguration, and therefore to develop a control algorithm that meets all the requirements is of great importance for a successful formation flying.The objective of this paper is to study the formation control problems for networked spacecrafts in deep space with fixed, switching, or dynamically changing communication topology with constant and timevarying communication delays. Since, for example, in the TPF case and other similar space missions, it is often the case that due to the dynamic nature of each spacecraft's states, the sensing and interspacecraft communication topology also changes over time. When the information networks are fixed, static graph-based [9][10][11][12][13][14] ideas are often used for specifying which communication links are present in the formation control. In [10] the author obtained a controller based on a connected information graph with one agent designated as active leader while the other designated as the follower. Ref.[14] used basic graph theory to define the communication links amongst spacecrafts in formation to study the distributed attitude alignment of spacecrafts under dire...