Abstract: This paper addresses the automatic sensor node deployment problem of partial coverage with arbitrary percentage in wireless sensor networks. It presents the analytical relationship between sensor node deployment and coverage percentage based on the optimal network coverage topology of equilateral-hexagon structure with dense distribution of sensor nodes, through which the optimal sensor node deployment of partial coverage is obtained in the ideal case. Furthermore, due to the limited density and random distribution of sensor nodes in real systems, an optimized collaborative partial coverage (OCPC) algorithm is proposed for practical application of obtaining partial coverage with arbitrary percentage. In OCPC, the working nodes which most approximate to the optimal deployment are woken up to work and the other nodes are turned off for energy conservation via the dynamic collaboration among sensor nodes, while the requirement of both coverage and connectivity are satisfied with the fewest possible sensor nodes so that the network energy consumption can be reduced. Simulation results show that the OCPC can effectively achieve coverage of arbitrary percentage and maintain the network connectivity, and its superiority in respect of energy conservation is demonstrated by comparison with PEAS (Probing environment and adaptive sleeping) and OGDC (Optimal geographic density control) algorithms.