Service restoration plays a vital role in improving the resiliency and reliability of distribution network. It is necessary to develop robust service restoration scheme for restoring the out-of-service loads in a short time. In this paper, a mixed-integer second-order cone programming (MISOCP) model is proposed for the service restoration of distribution network. This model relaxes the branch power flow equations into convex second-order cone format. The employment of the convex formulations is of great significance because it ensures that the obtained solution is global optimal. The service restoration is formulated as a multi-objective optimization problem considering minimization of de-energized loads and total number of switching operations. The simulation conducted on three test systems shows that the MISOCP method can provide robust and reliable restoration schemes.