NOMA is a key technology for future wireless communication, which can improve the spectral efficiency (SE) of mobile network. In this paper, a subchannel assignment algorithm is applied to maximize the energy efficiency of downlink heterogeneous NOMA network. Different from previous works, the subchannel assignment problems and power allocation problem are formulated as non-convex problem, then we transform the original problems to convex optimization problems and solve it using difference of convex functions (DC) programming. The simulation results confirm that the applied scheme not only enhance the sum rate of heterogeneous NOMA network but also energy efficiency (EE) of small cell base station (SBS).