I. Introduction
In RECENT years, multi-agent networks have been widely developed and utilized in various areas, such as consensus of multiple ground vehicles, formation of unmanned aerial vehicles and spacecraft formation control [1]–[7]. In addition, graph theory is an essential tool to deal with the multi-agent problems, and some considerable results on graph theory have been proposed by Ren et al. [8], Porfiri et al. [9], and Belykh et al. [10], [11]. Meanwhile, multi-agent systems (MASs) with linear dynamics have been intensively studied in the past decade [12]–[14]. The LMI technique is effective in tackling the cooperative control problem for linear MASs. However, the high order of LMIs will lead to high numerical complexity [15]. Therefore, reducing the order of LMIs deserves further study.