In Ad Hoc networks, topology control is designed to save energy and increase network capacity by enabling nodes to use proper transmission power, which is usually much smaller than the maximal transmission power. However, the randomicity of network deployment and node movement results in a non-uniform distribution of nodes. In a region where nodes are sparse, the distance between two nodes may be much longer than that in a dense region, so the transmission power is not able decrease and does not depend on which topology control algorithm is adopted. For the first time, this paper proposes movement control algorithms to improve performance of topology control by moving a subset of nodes to desiring positions. Based on the minimum spanning tree of the network topology graph, addition links are determined. Moreover, addition links are shortened and large communication range may be reduced by moving some nodes. Three movement algorithms, PMST-P, PMST-UV and LMST-LUV, are realized, and their performance is compared with each other with respect to maximum communication range, total moving distance and etc. using simulations.