Bandwidth slicing in software-defined 5g: A stackelberg game approach

With the explosive development of mobile Internet and the emergence of multimedia-rich applications such as augmented reality, virtual reality, video streaming, online gaming, and social networks, the traffic demands in the incoming fifth-generation (5G) era could grow significantly. To accommodate this massive amount of data traffic, Internet service providers (ISPs) must deploy new infrastructures to expand the capacity of current networks, which results in high capital and maintenance costs. Furthermore, mobile Internet applications in 5G will have diverse quality-of-service (QoS) requirements in terms of latency, burst size, throughput, and packet arrival rate, and reliable QoS provisioning depends on efficient and on-demand resource allocation [1], [2]. However, it is difficult to realize intelligent resource allocation due to the coupling between service provisioning and the taskoriented hardware.