An Architecture for M2M Communications Over Cellular Networks Using Clustering and Hybrid TDMA-NOMA

Machine-to-Machine (M2M) communications have now become a highly promising affiliate of the future fifth generation (5G) cellular networks. This paper proposes a novel architecture for M2M communications over cellular networks. $k$-mean clustering for machines as well as cluster head (CH) reselection method is applied in order to balance the power consumption within the machines to increase their battery life. For communication between CH and member machines, time division multiple access (TDMA) is proposed. On the other hand, for communication between CH and BS, non-orthogonal multiple access (NOMA) technology is considered. Performance of the proposed architecture is evaluated through extensive MATLAB simulations. Results clearly demonstrates the capability of the proposed architecture in improving the lifetime of machines and reducing communication delays. Comparisons with other counterparts also attests superior energy efficiency and delay performance of the proposed architecture.