An improved discrete fruit-fly optimization algorithm（ FOA） is proposed

aiming at problems of low area coverage rate and poor utilization rate of node caused by random deployment of wireless sensor networks（ WSNs）. Adaptive classify smell-based random search step size and different random search methods for different steps adopted

migration-based operation and elite library are introduced in multiple population co-evolution mechanism

which improves precision and efficiency of optimization. Simulation result shows that this new algorithm effectively resolve problem of WSNs coverage

it has the highest nodes usage while maximizing the network coverage rate and extend the network lifetime.