The energy of the sensor network nodes is limited, and in order to save the energy consumption of the sensor nodes, an improved compressed sensing algorithm combining the LEACH model with the CS theory based on the spatial-temporal correlation of nodes was proposed. The LEACH algorithm was used to cluster the net- work nodes and select the cluster head. Then, the cluster head node was sampled by the compressed sensing theory. The sampled data was passed to the remote sink node through multi-hop routing. Finally, a small amount of data from the cluster head used a reconstruction algorithm to accurately recover the original signal. The simu- lation results show that this method can effectively reduce the amount of data transmission, save the energy con- sumption of nodes and prolong the lifetime of the wireless sensor network.