A new region based energy conscious sink movement (RESM) is presented to enhance lifetime of heterogeneous Deep learning network using stable election protocol. The sink movement is based on the total energy of the individual regions of the entire sensing field. The sink always moves to the centre of the lowest non-zero energy region which is the main contribution in this research. Based on the residual energy of the regions, this sink movement method enhances the network lifetime. The simulation study results proved that this approach prolongs the network lifetime compared with the other sink movement methods. The lifetime improvement over LEACH, SEP and MSE is 23%, 6%, and 3.3% respectively.
