To sustain communication reliability and use the harvested energy effectively, it is crucial to consider energy sharing between energy harvesting tags (EHT) in a multiple access network, which are basic building blocks for the internet of things (IoT) applications. This technique also achieves higher throughput compared with the non-cooperative strategies despite energy losses occurred during energy transfer. We propose an energy cooperative communication strategy for a multiple access network of tags that depends on the harvested battery energy. We develop an optimal transmission policy for EHTs that maximizes the long-term joint average throughput using a Markov decision process (MDP) model. Simulation results show that the proposed energy cooperative policy produces improved performance than traditional policies.