An Enhanced K-means Clustering Algorithm with Non-Orthogonal Multiple Access (NOMA) for MMC Networks

In this paper, we address the need for motivating applications, such as sensor nodes, to operate under the Massive Machine to Machine Communication (MMC) mode in the future 5G cellular wireless network, while under constraints such as energy-efficiency, etc. It has been shown that the Random Access CHannel (RACH) in LTE-A cannot handle the possible simultaneous access of network resources by the billions of devices expected within an MMC Network. Thus, an enhanced K-means clustering algorithm accompanied by Non-Orthogonal Multiple Access (NOMA) is proposed, where each strong channel gain device is allocated to the appropriate cluster as a cluster head (CH) to enhance the network sum throughput. A performance analysis is conducted, where our proposed scheme was shown to have a higher network sum throughput than the traditional K-means, with a minimum rate requirement of 100-2000 kbps.