Lab 1 - Getting Started

Setup access details

The setup used for this lab session is a dCloud demo instance, where all the data represents real-life data captured from production systems.

Note

The lab setup format (1) was chosen in order to focus on the outcome of the Cisco AI Analytics use cases, rather than the configuration required to support it.
In this way you can analyze the results you can expect to see on a real network (2) without the need to wait for the cloud data pipelines to produce results following the service activation (3).
If you have any questions, please feel free to reach out to the proctors.

1. Ready to use demo setup
2. The data behind the AI/ML-Driven insights you'll analyze in this lab is taken from production network datasets
3. Depending on the specific use case, it takes between a few hours and 4 weeks until you can see results; the longer times apply to long-term analytics features, such as the AP Perfomance Advisories

Please click on the button on the lab session page to begin the lab session.

Then use the following credentials on the Cisco Catalyst Center login page to get started:

• Username: demo
• Password: demo1234!

Network Diagram

The Cisco AI Analytics solution is accessible via the Cisco Catalyst Center appliance, leveraging the powerful Cisco AI Cloud platform to enable AI-driven use cases ranging from network monitoring, troubleshooting and security.

Cisco AI Analytics leverages the data platform on Cisco Catalyst Center in order to collect the relevant telemetry data from the managed network devices.

Data processing for the AI-powered use cases is offloaded to the Cisco AI Cloud platform, where both raw data export and processed data retrieval is handled by a lightweight agent installed on the on-prem appliance.

Use cases

Cisco AI Analytics was released in 2019, covering AI-driven use cases extending the Cisco DNA Assurance solution, mostly focusing on Wi-Fi troubleshooting.

In the following years it has expanded to cover additional areas, such as security with the AI-driven features of Cisco AI Endpoint Analytics or even Wi-Fi full closed-loop, with Cisco AI-enhanced RRM. In 2023 the coverage has expanded into cross-domain network monitoring, with the launch of Event Analytics.

Note

This lab guide is designed to give you a comprehensive, hands-on experience focused on a specific subset of the Cisco AI Analytics use cases. For the remaining use cases not featured in the lab exercises, we provide a high-level overview, for your reference. As you progress through the lab, you are encouraged to explore these additional use cases independently to deepen your understanding of the product's full potential.

Each section has direct links to the detailed use case information and lab instructions.

You can explore the use cases below in any order, based on your preference and interest; at the end of each use case specific lab section, you'll find a link to come back to this section.

Assurance

The following use cases are part of the Assurance feature set and are covered under the DNA Advantage license.

Wireless LAN

• AI-Driven Anomaly Detection
  Generate AI-driven baselines to determine the normal behavior of the network and use them as a reference to raise relevant issues, reducing alerting noise. Focus on troubleshooting, offering guided root cause analysis for faster problem resolution. Click here to deep dive into AI-driven issues

• Baseline Dashboard
  Explore the network performance across sites, use the AI-driven baselines to observe the predicted vs. the actual network behavior with respect to multiple Wi-Fi Onboarding and Roaming related KPIs. Focus on network monitoring, observing the performance evolution over time as well as comparing the performance of different SSID within the same building. Click here to deep dive into the Baseline Dashboard

• AP Performance Advisories
  Long-term AI-driven analytics identifying APs delivering sub-optimal client experience. Analyze the problematic radios by comparing them with reference radios in the network. Focus on performance optimization, knowing what radios require attention in order to improve the client experience. Click here to deep dive into AP Performance Advisories

Cross-domain

• Event Analytics (New in version 2.3.7!)
  The latest use case added to the Cisco AI Analytics suite. Focus on network monitoring, enabling visibility and correlation across different network event types, covering both wired and wireless. Click here to deep dive into Event Analytics

Security

The following use cases are part of the Cisco AI Endpoint Analytics feature set and are covered under the DNA Advantage license.

• AI Endpoint Analytics Smart Grouping
  Identify groups of similar endpoints on the network, where the endpoint classification data is either missing or incomplete; automatically generate classification rules, leveraging crowdsourcing and direct user feedback to enhance visibility on endpoints accessing the network. Focus on endpoint visibility. Click here to deep dive into Smart Grouping

• AI Endpoint Analytics Spoofing Detection
  Analyze the behavior of endpoints belonging to specific classes (e.g., printers, IP phones, IP cameras, IoT building automation devices..) to determine whether the network traffic patterns match those of legitimate endpoints and raise spoofing alerts in case of behavior mismatch. Focus on security alerting, by identifying (and optionally blocking) endpoints exhibiting suspicious behavior. Click here to deep dive into Spoofing Detection

Additional use cases

The Cisco AI Analytics solution covers additional use cases that are not covered in this lab; the list below is provided for reference.

• Network Heatmaps
  Powerful visualization tool allowing to identify the top network devices, as well as observing trends and correlating performance across different network entities. The network heatmaps cover both APs with an extended set of Wi-Fi and system KPIs, as well as switches, exposing information about temperature sensors. Focus on network monitoring.

• Trends
  Identify outliers in the network, either an AP within a building, or a building compared to all the managed buldings, whose behavior has constantly deviated more than the rest of the network. Focus on network monitoring.

• Network Comparisons
  Automatically group entities (APs and Wi-Fi clients) across the entire network in three main buckets based on the performance with respect to multiple KPIs; then use this information to compare performance across buildings, AP models or client types. Focus on performance optimization.

• Peer Comparisons
  Compare the distribution of Wi-Fi related KPIs on the local network against the aggregated distribution of other, external, similar (by type and size) networks. Focus on performance optimization, by extending the visibility beyond the local network.

Use case Navigation

In this section we will see where to find different use cases described above in Cisco Catalyst Center. Upon logging in to dCloud instance of Cisco Catalyst Center you will see below home page

Click on menu on the top left corner of screen.

Security related use cases i.e AI Endpoint Analytics Smart Grouping and AI Endpoint Analytics Spoofing Detection are located under Policy > AI Endpoint Analytics section and all other use cases are located under Assurance section.

Next let's first explore Assurance section.

Event Analytics and AI-Driven issues are located under Issues and Events section. The rest of the use cases are located under AI Network Analytics section as highlighted in image below. The AP Performance Advisories and Trends use cases are located under Trends and Insights.