Optical Label Processing Techniques for Intelligent Forwarding of Packets in All-Optical Packet Switched Networks

Background

To date, several optical labeling techniques have been proposed and demonstrated. Those techniques can be divided in two areas: all-optical label processing techniques, in which all the operation required to recognize the labels are performed in the optical domain, and opto-electronics label processing techniques, in which optical and electronics processing are combined to optimally perform the label recognition.

Several works have been presented in literature that demonstrate all-optical label processing (Vegas Olmos, 2004; Klonidis 2005; Wang 2006, 2010; Takenaka 2006; Wai 2005; Herrera 2008; Calabretta 2004, 2005, 2006, 2007, 2008; Scaffardi 2010; Hamilton 2002; Wada 2007; Le Minh 2006). Despite the ultrafast operation of processing the labels in the optical domain, the main issue in all-optical label processors is the low amount of labels that can be processed and the lack of mature digital photonic circuits (logics) for complex operation required for implementing algorithms and schedulers for routing the optical packets. Moreover, the complexity of those circuits and the lack of mature and generic photonic integration technology prevented the realization of small footprint, and low power consumption circuits for processing large number of labels. Future breakthrough in the field of photonic integration could lead to efficient and scalable all-optical label processing circuits.