Scalable two-stage Clos-network switch and module-first matching
Roberto Rojas‐Cessa, Chuan-Bi Lin · 2006
Clos-network switches were proposed as a scalable architecture for the implementation of large-capacity circuit switches. In packet switching, the three-stage Clos-network architecture uses small switches as modules to assemble a switch with large number of ports or aggregated ports with high data rates. However, the configuration complexity of packet Clos-network switches is high as port matching and path routing must be performed. In the majority of the existing schemes, the configuration process performs routing after port-matching is achieved, and thus making port matching expensive in hardware and time complexity for a large number of ports. Here, we reduce the configuration complexity by performing routing first and port matching afterwards in a three-stage Clos-network switch. This approach applies the reduction concept of Clos networks to the matching process. This approach results in a feasible size of schedulers for up to Exabit-capacity switches, an independent configuration of the middle stage modules from port matches, a reduction of the matching communication overhead between different stages, and a release of the switching function to the last-stage modules in a three-stage switch. By using this novel matching scheme, we show that the number of stages of a Clos-network switch can be reduced to two, and we call this the two-stage Clos-network packet switch.