The K-coverage configuration guarantees that each location in an area is covered by at least K wireless sensors. To prolong the system lifetime, how to design an efficient sleep-scheduling algorithm have been attracted much attention. Many studies cannot maximize the system lifetime, while ensuring the K-coverage configuration. In this paper, we classify wireless sensors into groups, in which each group is K-covered with the minimum overlap among sensors. Different from the previous studies, our algorithm further reduces the cost of maintaining connectivity among sensors and the latency of reporting events. Experimental results demonstrate the efficiency and accuracy of our algorithm.