On the impact of client to line port blocking in the line interface count and footprint of transport networks

In order to cope with client traffic heterogeneity while reducing the number of line interfaces/wavelengths used in the network, low-data-rate traffic must be efficiently groomed into higher-bit-rate line signals. Traffic grooming can be performed via different node architectures, which provide different trade-offs between their complexity/cost and the number of client to line port interconnections supported. Particularly, non-blocking grooming at a node can be attained by having a common backplane for all client and line interface modules deployed in the same network element. Another architecture consists of using various backplanes and preventing the interconnection between client and line interface modules connected to different backplanes, thereby restricting non-blocking grooming to interface modules sharing the same backplane. Finally, grooming can be confined to a single muxponder module or cascade of muxponder modules, further restricting the possible client to line port interconnections. In this work we evaluate the impact of the grooming architecture in the minimum number of line interfaces and modules required.