An Efficient Solution for Traffic Cognition in Future IoE Networks

Network traffic cognition is essential for network operations, such as resource management and network monitoring. With the advancements in 6G and Internet of Everything (IoE) technologies, the diversity of applications and the volume of traffic will expand at an unprecedented rate, posing critical challenges for traffic cognition. Existing traffic cognition methods struggle to improve granularity and speed simultaneously. In this paper, based on the theory of traffic fractals, the variation traits of traffic flows are studied on spatial and temporal observation scales respectively. Consequently, evolutionary fractals (EFs) are established through the Legendre transformation of these variation traits, resulting in the innovative EF cognition model (EFCM). Compared with traditional fractals, EFs analyze traffic flows using both spatial and temporal information, capturing more detailed characteristics and thus obtaining higher precision for fine-grained cognition. Additionally, discrete EFs, unlike traditional continuous fractals, significantly enhance the speed of cognition computation. The proposed EFCM achieves precise and rapid cognition at a fine-grained level, and the experimental results demonstrate its superiority over other methods in terms of cognition granularity and speed.