Intelligent Data Flow Management Based on Optical Label Switching Technology for Photonic Transport Network

Optical transport networks are developed rapidly to ever increasing traffic requirements. With the advent of new technologies such as IoT, artificial intelligence, cloud computing and tactile internet, existing optical fiber infrastructure should be improved substantially. Telecommunication operators should pay attention to global data volume which increases exponentially. In this paper, we present the new architecture of optical transport network based on all-optical label swapping technology. This approach allows operating optical label in the optical domain without opto-electronic conversion. The structure of the switch, label splitter and wave router are presented. We propose the machine learning algorithm which allows saving optical resources of all optical label swapping network based on the length of the payload. Proposed algorithm is able to improve over time and by learning from previous experience.