Aiming at solve the problems of uneven clustering

unreasonable cluster head selection and excessive energy loss caused by random selection of initial cluster centers in the clustering process of K-Means（ KM）algorithm in GEC algorithm

an improved algorithm KM-LEACH is proposed in this paper.Firstly

KM clustering algorithm is used for clustering

and aiming at the problem that random selection of initial clustering centers in KM algorithm can easily lead to local minimum solution

genetic algorithm is used to improve the algorithm

and the optimal initial clustering centers are selected to achieve global optimization; then residual energy and location influence factor are introduced to select cluster heads reasonably.After that

the data is transmitted to cluster head in the form of time division multiple access（ TDMA）

which reduces the number of network congestion and energy consumption of data transmission.The improved KM clustering algorithm can be clustered and evenly clustered at one time to reduce the energy loss in the clustering process

and the cluster head can be reasonably selected by introducing the influence factor to balance the energy loss in the network.The simulation results show that KMLEACH algorithm can save.