We present the architecture and function of a Network Control Server (NCS) prototype that has been developed for deployment in an NGN to keep that in an acceptable operating condition. For this task, the NCS gathers statistical data from the network and uses this information to generate a global view of the current operating condition. Using this information, the NCS follows its internal strategy and uses optimization algorithms to adapt the network to changing operating situations.
Among the benefits of the presented approach are an increased network resilience with a quick reaction towards changes in the offered traffic, a reduced QoS request blocking rate and a more efficient network operation. These have been evaluated in simulations and shown in a functional lab prototype.