The operating state estimation of power system is a critical process for providing a best-fit state estimation based on the meter measurements in the field and the configuration of power-grid network. To deal with the bad meter measurements caused by various faults in state estimation, power system researchers have developed numerous detection schemes in the past. Liu et al.  recently proposed a new stealthy false data injection attack, which can bypass the existing bad data detection schemes and arbitrarily manipulate the states of power system, posing dangerous threats to the control of a power system. Nevertheless, their results did not show the detailed information of meters to be compromised. In this paper, we tend to tackle this issue and develop mechanisms to efficiently compute the optimal set of meter measurements given a number of state variables to be manipulated in a power-grid network. We formalize the problem of finding the optimal set of meter measurements as a known NP-hard problem and propose a heuristic approach to derive the near-optimal set of meter measurements efficiently. We implement our proposed scheme on the IEEE 9-bus, 14-bus, 30-bus, 118-bus and 300-bus systems and our data shows its efficiency and effectiveness.