User Clustering in mm Wave Quality of Service-Based Non-orthogonal Multiple Access (QNOMA) for Vehicular Network

Due to dynamic nature topology in vehicular network and exponentially huge amount of traffic, it is very difficult to provide services with high data rate to these vehicles. In a heterogeneous network, a V2I system with NOMA capabilities can achieve low latency and high reliability. In this paper, we focus V2I communication system in 5G. The performance of V2I is measured in terms of sum rate according to QoS requirement of different vehicles inside a cell. The intelligent transport system is required to cover the challenges of existing wireless networks. The most common problem in vehicular network is the requirement of data rate differently. In mm wave NOMA, the users which are closer to the base station received better rate comparatively to the other users. This become the bottleneck to meet the QoS demand for far user in mm wave NOMA. We solved this problem by using clustering technique to divide the number of users into different clusters. We proposed clustering scheme and compare with other two unsupervised machine learning techniques included K means and hierarchical. After clustering, SIC ordering is arranged according to their target rates of vehicles in QNOMA. In the end, we compare these clustering technique according to the data rate requirement of different vehicles. The proposed user clustering algorithm significantly improves the sum rate of the network according the QoS requirements.