For static or dynamic networks entirely located within a specific region (Network-0 and Network-3), the concept of path ownership is straightforward. In a network entirely positioned within the static region, the path is owned by the static region. Similarly, a network located within the dynamic region is owned by the reconfigurable module in that region. During child implementations of reconfigurable modules, the QoS for the NoC path within the static region remains consistent and is adhered to as defined in the parent implementation. However, for the path owned by the reconfigurable module within the dynamic region, the reconfigurable module can define and configure the QoS parameters for that NoC path.
The following figure shows that for the NoC path entirely located within the static region (Network-0), the DFX flow maintains the entire path with the QoS requirements as initially defined in the parent implementation. In contrast, for the NoC paths that are entirely located within the dynamic region (Network-3), you have the flexibility to choose to reconfigure the entire path. This includes the option to add more NMUs or NSUs, resulting in configurations like Network-3a and Network-3b. In child implementations, you can introduce entirely new NoC paths within the dynamic region, as represented by Network-4.