Artificial Intelligence and machine learning are expected to be some of the big topics at next week\u2019s Cisco Live event and the company is already talking about how those technologies will help drive the next generation of .\n\u201cArtificial intelligence will change how we manage networks, and it\u2019s a change we need,\u201d wrote John Apostolopoulos Cisco CTO and vice president of Enterprise Networking in a blog about how Cisco says these technologies impact the network.\u00a0\n\nAI is the next major step for networking capabilities, and while researchers have talked in the past about how great AI would be, now the compute power and algorithms exist to make it possible, Apostolopoulos told Network World.\u00a0\nTo understand how AI and ML can boost IBN, Cisco says it's necessary to\u00a0 understand four key factors an IBN environment needs: infrastructure, translation, activation and assurance.\nInfrastructure can be virtual or physical and include wireless access points, switches, routers, compute and storage. \u201cTo make the infrastructure do what we want, we use the translation function to convert the intent, or what we are trying to make the network accomplish, from a person or computer into the correct network and security policies. These policies then must be activated on the network,\u201d Apostolopoulos said.\nThe activation step takes the network and security polices and couples them with a deep understanding of the network infrastructure that includes both real-time and historic data about its behavior. It then activates or automates the policies across all of the network infrastructure elements, ideally optimizing for performance, reliability and security, \u00a0Apostolopoulos wrote.\nFinally assurance maintains a continuous validation-and-verification loop.\u00a0 IBN improves on translation and assurance to form a valuable feedback loop about what\u2019s going on in the network that wasn\u2019t available before. \u00a0\u00a0\n\n\n\n\n\nApostolopoulos used the example of an international company that wanted to set up a world-wide video all-hands meeting.\u00a0 Everyone on the call had to have high-quality, low-latency video, and also needed the capability to send high-quality video into the call when it was time for Q&A.\n\u201cBy applying machine learning \u00a0and related machine reasoning, assurance can also sift through the massive amount of data related to such a global event to correctly identify if there are any problems arising. We can then get solutions to these issues \u2013 and even automatically apply solutions \u2013 more quickly and more reliably than before,\u201d Apostolopoulos said.\nIn this case, assurance could identify that the use of WAN bandwidth to certain sites is increasing at a rate that will saturate the network paths and could proactively reroute some of the WAN flows through alternative paths to prevent congestion from occurring, Apostolopoulos wrote.\u00a0 \u00a0\n\u201cIn prior systems, this problem would typically only be recognized after the bandwidth bottleneck occurred and users experienced a drop in call quality or even lost their connection to the meeting. It would be challenging or impossible to identify the issue in real time, much less to fix it before it distracted from the experience of the meeting.\u00a0 Accurate and fast identification through ML and MR coupled with intelligent automation through the feedback loop is key to successful outcome.\u201d\nApostolopoulos said AI can accelerate the path from intent into translation and activation and then examine network and behavior data in the assurance step to make sure everything is working correctly. Activation uses the insights to drive more intelligent actions for improved performance, reliability and security, creating a cycle of network optimization.\nSo what might an implementation of this look like? Applications that run on Cisco\u2019s DNA Center may be the central component in an IBN environment. Introduced on 2017 as the heart of its IBN initiative, Cisco DNA Center features automation capabilities, assurance setting, fabric provisioning and policy-based segmentation for enterprise networks.\n\u201cDNA Center can bring together AI and ML in a unified manner,\u201d Apostolopoulos said.\u00a0 \u201cIt can store data from across the network and then customers can do AI and ML on that data.\u201d\nCentral to Cisco's push is being able to gather metadata about traffic as it passes without slowing the traffic, which is accomplished through the use of ASICs in its campus and data-center switches.\n\u201cWe have designed our networking gear from the ASIC, OS and software levels to gather key data via our IBN architecture, which provides unified data collection and performs algorithmic analysis across the entire network (wired, wireless, LAN, WAN, datacenter),\u00a0 Apostolopoulos said. \u201cWe have a massive collection of network data, including a database of problems and associated root causes, from being the world\u2019s top enterprise network vendor over the past 20-plus years.\u00a0 And we have been investing for many years to create innovative network-data analysis and ML, MR, and other AI techniques to identify and solve key problems.\u201d\nMachine learning and AI can then be applied to all that data to help network operators handle everything from policy setting and network control to security.\n\u201cI also want to stress that the feedback the IT user gets from the IBN system with AI is not overwhelming telemetry data,\u201d Apostolopoulos said. Instead it is valuable and actionable insights at scale, derived from immense data and behavioral analytics using AI.\u00a0\nManaging and developing new AI\/ML-based applications from enormous data sets beyond what Cisco already has is a key driver behind it\u2019s the company\u2019s Unified Compute System (UCS) server that wasa rolled out last September. While the new server, the UCS C480 ML, is powerful \u2013 it includes eight Nvidia Tesla V100-32G GPUs with 128GB of DDR4 RAM, 24 SATA hard drives and more \u2013 it is the ecosystem of vendors \u2013 Cloudera, HortonWorks and others that will end up being more important.\u00a0\nEarlier this year Cisco forecast that AI and ML will significantly boost network management this year.\n\u201cIn 2019, companies will start to adopt Artificial Intelligence, in particular Machine Learning, to analyze the telemetry coming off networks to see these patterns, in an attempt to get ahead of issues from performance optimization, to financial efficiency, to security,\u201d said Anand Oswal, senior vice president of engineering in Cisco\u2019s Enterprise Networking Business. The pattern-matching capabilities of ML will be used to spot anomalies in network behavior that might otherwise be missed, while also de-prioritizing alerts that otherwise nag network operators but that aren\u2019t critical, Oswal said.\n\u201cWe will also start to use these tools to categorize and cluster device and user types, which can help us create profiles for use cases as well as spot outlier activities that could indicate security incursions,\u201d he said.\nThe first application of AI in network management will be smarter alerts that simply report on activities that break normal patterns, but as the technology advances it will react to more situations autonomously. The idea is to give customers more information so they and the systems can make better network decisions. Workable tools should appear later in 2019, Oswal said.