Networks are evolving constantly. The number of subscribers is increasing and new services require more network capacity. Higher capacity demands are tackled with network re-segmentation and node splitting, which in practice leads to laborious and costly on-site re-adjustment of network devices. Not only will it result in increased risk for human errors and network unreliability, but also higher operating costs. An elegant way to reduce this manual work is to bring intelligence to the network. By intelligence, we mean intelligent network devices that can be automatically aligned and management software to control and monitor the whole network. Adding intelligence to networks can lead to astonishing results: we typically see 50 % savings in operating costs and payback times of 1-3 years for the additional investments required to enable the intelligence. In this paper, we outline how such an intelligent network can be constructed and how intelligence affects network management.
Network devices are getting more complex. Although the devices are robust and designed to stand rough conditions, technical problems can occur – configuration faults, hardware defects or changes in the system structure that require re-adjustment of the devices. With constantly increasing capacity demands, broadband network operators need to invest heavily both in new network hardware and the maintenance and renewal of their network. More devices mean more frequent maintenance operations, which translates to the need to have more on-site personnel to handle these operations. Additionally, with the size of the network expanding, different types of challenges begin to surface.