A site or a node in the static WDM point-to-point optical network uses a combination of wavelength division multiplexing in a waveguide and space-division multiplexing across multiple waveguides to establish a unique link to every other site in the macrochip. A WDM multiplexer merges groups of ‘N’ transmit waveguides into a single N-wavelength WDM waveguide. This runs east–west over to the column containing the destinations of those ‘N’ transmit waveguides. Through an interlayer coupler, these N-wavelength bundle drops to the second SOI routing layer and then runs north–south. At each of the four destination chips, a WDM drop filter pulls off the appropriate wavelength, routes it back through the first SOI wafer, and then onto the target bridge chip. The network has no in-line switching and is non-blocking, yielding …show more content…
A path-setup packet is sent on electronic control network. It includes information on the destination address of node B and additional control information such as priority, ‘1’ hot source address and flow id. The path set-up packet is routed in the electronic control network. It reserves the photonic switches as it travels along the path. The next hop is determined at every router in the path according to some routing algorithm used. By the time the path-setup packet reaches the destination, the photonic path is reserved. The, a fast light pulse is sent on the photonic path from node B to node A to indicate that the path has been reserved. The photonic message is now sent from node A, following the path from switch to switch, until it reached node B. Once the packet reaches node B, the message transmission is completed. Next, the path-teardown packet is sent from node B to node A on the electronic control network to relieve the path. Also, the photonic message is checked for errors and a small acknowledgement packet is sent from node B to node A on the electronic control