In this paper, we present three approaches dealing with the decentralized control of interconnected systems. The first one treats the decentralized quadratic optimal control. The second approach concerns the decentralized pole-placement control. The third one, which is our main contribution, focuses on a decentralized stabilization approach that guarantees the stability of the global interconnected system. The decentralized local gains are calculated and formulated via the resolution of linear matrix inequalities (LMIs) problem. A numerical simulation-based comparison of the three methods is performed on an interconnected double-parallel inverted pendulum.