Traffic Matching in 5G Ultra-Dense Networks

Driven by enormous traffic demand, traditional cellular networks are evolving into an ultra-dense architecture. To take full advantage of such an ultra-dense architecture and efficiently serve the traffic with spatiotemporal fluctuation, the transmission mechanisms should be redesigned under the constraints of backhaul and energy consumption. In this article, we summarize and classify the spatiotemporal arrival properties of different traffic in ultra-dense networks, and optimize several promising technologies, such as dynamic time-division duplexing and full duplex radio, to adapt the network service to match the traffic. A new approach based on combining stochastic geometry and queueing theory is proposed to provide a useful guidance for the design of ultra-dense networks when the traffic is spatiotemporally fluctuating. Successive efforts, such as the analysis of different QoS requirements and more complicated but practical scenarios, still require attention along this line of research.