Network edges drive new patterns of network traffic, and as a result of these distributed patterns, it is getting increasingly difficult to predict traffic in NRENs for two reasons. First, big data science projects like astronomical observatories or high energy physics colliders, produce massive amounts of data in spurts that need to be distributed to different institutions and data centers for processing and research. Second, the advent of 5G/IoT/MEC platforms will lead to the creation of many new services and thus many new and erratic data flows with widely differing traffic handling and latency requirements. Underlying this traffic – doing all the heavy lifting of moving bits around – is the optical network. One way to handle the situation of traffic unpredictability is to overprovision the optical network to deal with worst case requirements. But that is never the most attractive direction because of cost. Increasingly what is being talked about are organic networks, that can absorb and respond smoothly to changing traffic demands. These networks also need to adapt to the way customers are changing their enterprise architectures in and around the cloud. The solution is based on continuous and fluid reconfiguration of existing resources – while maintaining high availability – and autonomously initiating requests for additional resources. The joint presentation by Surfnet and ECI discusses both practical and industry initiatives at making this a reality.
Networks · Services · People | Code of Conduct | Privacy Policy
