Internet of Things (IoT), as a novel technology, has many challenges that massive connectivity and low latency are two most important of them. Machine to machine (M2M) communication is regarded as a new technology to realize IoT for the 5G wireless networks. In this paper, we are designed a new clustering scheme in a millimeter-wave (mmWave) non-orthogonal multiple access (NOMA) transmission to reduce the system overhead for massive connectivity, which depends on the distance between machine type communication (MTC) device and the base station (BS). The clustering scheme includes a near, a far and a mid-cell MTC device per cluster with a non-overlapping frequency band for both far and mid-cell MTC devices to reduce inter-cluster interference. The proposed three clustering schemes in different conditions are as 1) the random near, the random mid-cell and the random far MTC devices (RNRMRF); 2) the nearest near, the nearest mid-cell and the nearest far MTC devices (NNNMNF); and 3) the nearest near, the farthest mid-cell and the farthest far MTC devices (NNFMFF). The outage probabilities of the clustering schemes are compared analytically for the line of sight paths. Also, the non-line of sight paths for channel gain and the impact of these paths over outage probability are analyzed. The computer numerical and simulation results show the accuracy of the developed analytical and approximation results in high signal-to-noise ratio cases. As a result, our clustering schemes are offered to settle a trade-off between system overhead and outage performance. Besides as an asymptotic computational algorithm, we provided the order of complexity to assess the difficulty of the system.

物联网（IoT）作为一项新技术，面临着许多挑战，其中巨大的连接性和低延迟是其中两个最重要的挑战。机器对机器（M2M）通信被认为是为5G无线网络实现IoT的新技术。在本文中，我们为毫米波（mmWave）非正交多路访问（NOMA）传输设计了一种新的群集方案，以减少大规模连接的系统开销，这取决于机器类型通信（MTC）设备之间的距离和基站（BS）。集群方案包括每个集群的近，远和中小区MTC设备，具有用于远和中小区MTC设备的非重叠频带，以减少集群间干扰。在不同条件下提出的三种聚类方案是：1）随机近，随机中小区和随机远MTC设备（RNRMRF）；2）最近的，最近的中间小区和最近的MTC设备（NNNMNF）；3）最近的，最远的中型小区和最远的MTC设备（NNFMFF）。对于视线路径，对聚类方案的中断概率进行了分析比较。同样，分析了通道增益的视线路径以及这些路径对中断概率的影响。计算机数值和仿真结果表明，在高信噪比情况下，所开发的分析和近似结果的准确性。结果，我们提供了集群方案，以解决系统开销与中断性能之间的折衷问题。除了作为渐近计算算法外，我们还提供了复杂度的顺序来评估系统的难度。