In this study we proposed a scheme to achieve adequate coverage while maximizing network lifetime in a sensor network. The proposed approach is based on controlling the number of active sensors using information theory. It assumes that the area of coverage is divided into grids and uses a threshold probability to decide whether the sensors should be ON or OFF. Based on the number of ON sensors in each grid we define an entropy value which we use to adjust the threshold probability for the next epoch. The control mechanism is devised to maximize entropy while keeping the number of ON sensors at a minimum. This approach improved the coverage and network lifetime.