Parallel Activation Architecture

Figure 2 illustrates the Parallel Activation architecture discussed in [4]. In this approach, as soon as a failure occurs, nodes downstream to the failed headend sense a loss of light. The egress node among these nodes, on sensing LOL, multicasts an alternate path setup message to each node on the alternate path. The message also contains the address of the node upstream to the node to which the message is sent. On receiving such a message a node on the alternate path extracts the address of its upstream neighbor from it and exchanges wavelength and port information with it. After this, each node sets up its cross connects and a complete alternate light-path is established. The time required to complete each phase of the restoration process is as follows:

$$t_{detect} = t_{LOL}$$ (8)

$$t_{FIS} = d_{OTN}(l_w - f)$$ (9)

$$t_{setup} = 2 \cdot d_{DCN}(h)$$ (10)

$$t_{restore} = t_{LOL} + d_{OTN}(l_w - f) + 2 \cdot d_{DCN}(h) + t_{switch}$$ (11)

Figure 2: The parallel activation architecture.