Role of AI and ML in Network OperationsAI 與 ML 在網路維運中的角色

Here you will receive an overview of the role artificial intelligence and machine learning play in network operations, with emphasis on the Cisco products that incorporate these cutting-edge technologies, demonstrating their pivotal role in modern network operations.

本單元將概略介紹人工智慧（AI）與機器學習（ML）在網路維運中扮演的角色，並著重介紹採用這些尖端技術的 Cisco 產品，展示它們在現代網路維運中的關鍵地位。

AI and ML are transforming network operations by enhancing efficiency, reliability, and security, shifting from reactive to proactive management. This transformation has significantly benefited several key areas that you can see in the following figure.

AI 與 ML 正在改變網路維運，透過提升效率、可靠性與安全性，使管理方式從被動反應轉為主動預防。這項轉變已為以下圖中所示的幾個關鍵領域帶來顯著效益。

AIOps, short for Artificial Intelligence for IT Operations, refers to the application of artificial intelligence and machine learning techniques to enhance and automate IT operations. Cisco has developed a suite of products that stand at the forefront of the AIOps revolution in network engineering.

AIOps（Artificial Intelligence for IT Operations，IT 維運人工智慧）是指運用人工智慧與機器學習技術來強化並自動化 IT 維運。Cisco 已開發出一系列站在網路工程 AIOps 革新前沿的產品。

These solutions are crafted to make networks more autonomous and efficient, with the ability to self-optimize and quickly resolve issues.

這些解決方案的設計目標是讓網路更自主、更有效率，具備自我最佳化並快速解決問題的能力。

Cisco Catalyst Center (previously known as Cisco DNA Center) is central to Cisco intent-based networking. It offers centralized management, automation, and orchestration across the entire network. Within Cisco Catalyst Center, the AI Network Analytics feature provides a robust array of capabilities, including intelligent issue detection through AI-driven baselining and anomaly detection, proactive insights for trend and pattern identification, and comparative benchmarking to evaluate network performance against peers or other sites. By continuously collecting and analyzing network data, Cisco AI Network Analytics adapts to evolving network conditions, enabling IT teams to proactively address potential issues and improve overall network performance.

Cisco Catalyst Center（前身為 Cisco DNA Center）是 Cisco 意圖式網路的核心，提供整個網路的集中式管理、自動化與協調。在 Cisco Catalyst Center 中，AI Network Analytics 功能提供一系列強大的能力，包括透過 AI 驅動的基準建立與異常偵測進行智慧問題偵測、用於趨勢與模式識別的主動洞察，以及用於評估網路效能相對於同儕或其他站點表現的比較基準測試。透過持續收集並分析網路資料，Cisco AI Network Analytics 能因應不斷變化的網路狀況做出調整，使 IT 團隊得以主動處理潛在問題並提升整體網路效能。

In the figure, you can see the average Client RSSI (Received Signal Strength Indicator) over a week for two buildings in San Francisco and San Jose, grouped into three distinct categories (Low, Medium High). The plot was produced by Cisco AI Network Analytics engine and showcases how AI helps engineers to assess average signal strength received by clients in different locations.

圖中顯示舊金山與聖荷西兩棟建築物一週內的平均用戶端 RSSI（接收訊號強度指標），分為三個不同的類別（低、中、高）。此圖表由 Cisco AI Network Analytics 引擎產生，展示 AI 如何協助工程師評估不同位置用戶端接收到的平均訊號強度。

The Cisco Meraki advanced WAN analytics use ML algorithms to enhance management, troubleshooting, and optimizing connectivity and uptime. AI within Cisco Meraki enhances security through real-time threat detection and microsegmentation and powers smart cameras and environmental sensors for monitoring safety and compliance. AI and ML also provide operational insights, facilitating rapid anomaly response and process optimization.

Cisco Meraki 進階 WAN 分析運用 ML 演算法強化管理、疑難排解，並最佳化連線與正常運行時間。Cisco Meraki 中的 AI 透過即時威脅偵測與微分段強化安全性，並驅動用於監控安全與合規性的智慧攝影機與環境感測器。AI 與 ML 也提供維運洞察，促成快速的異常回應與流程最佳化。

The following figure shows one of the Meraki AI powered features—the Auto RF solution that automatically optimizes the wireless radio parameters, such as selecting the best available channel and adjusting the power level, allowing you to maximize performance and minimize interference.

下圖展示 Meraki AI 驅動功能之一——Auto RF 解決方案，該方案能自動最佳化無線電參數，例如選擇最佳可用頻道並調整功率等級，讓您能將效能最大化並將干擾降到最低。

Cisco Nexus Dashboard provides a unified management and monitoring pane for data center networks, hosting applications like Cisco Nexus Dashboard Insights (NDI). AI and ML in Cisco NDI are used for Event Analytics to refine control-plane event analysis. These technologies detect correlations between configuration changes, control-plane faults, and events, identifying anomalies that could disrupt network operations.

Cisco Nexus Dashboard 為資料中心網路提供統一的管理與監控面板，並可承載 Cisco Nexus Dashboard Insights（NDI）等應用程式。Cisco NDI 中的 AI 與 ML 用於事件分析，以精進控制平面事件分析。這些技術能偵測組態變更、控制平面故障與事件之間的關聯性，找出可能中斷網路維運的異常情形。

Cisco AppDynamics is a comprehensive application performance management and IT operations analytics platform. It uses advanced AI and ML technologies to enhance the observability and operational efficiency of both applications and infrastructure. Using machine learning, AppDynamics automatically detects performance issues by establishing dynamic baselines that account for historical data, including time-of-day and seasonal variations, facilitating immediate anomaly detection. It also uses ML-driven root cause analysis to pinpoint the underlying causes of anomalies and integrates log analysis tools to identify outliers and detect log patterns. These AI and ML capabilities ensure optimal application performance by quickly identifying and addressing issues.

Cisco AppDynamics 是一套全方位的應用程式效能管理與 IT 維運分析平台。它運用先進的 AI 與 ML 技術，強化應用程式與基礎架構的可觀測性與維運效率。AppDynamics 運用機器學習，透過建立考量歷史資料（包括一天中的時段與季節性變化）的動態基準，自動偵測效能問題，促成即時異常偵測。它也運用 ML 驅動的根本原因分析找出異常的根本原因，並整合記錄分析工具以找出離群值並偵測記錄模式。這些 AI 與 ML 能力藉由快速找出並解決問題，確保應用程式效能達到最佳狀態。

The following figure shows a drill-down of a Transaction snapshot in Cisco AppDynamics where Machine Learning algorithms in the background work on root cause analysis and help you pinpoint and resolve issues.

下圖顯示 Cisco AppDynamics 中某筆交易快照的深入分析，其中背景執行的機器學習演算法進行根本原因分析，協助您找出並解決問題。

Cisco ThousandEyes is a network intelligence platform that offers visibility into the digital delivery of applications and services over the internet. It helps organizations monitor, troubleshoot, and optimize their network infrastructures, including traditional networks, cloud networks, and SaaS applications. Using AI and ML technologies, Cisco ThousandEyes automatically analyzes historical and current data to detect anomalies and disruptions within the network infrastructure, enhancing the ability to maintain and improve network performance.

Cisco ThousandEyes 是一套網路智慧平台，提供對應用程式與服務透過網際網路傳遞情形的可視性。它協助組織監控、疑難排解並最佳化其網路基礎架構，包括傳統網路、雲端網路與 SaaS 應用程式。Cisco ThousandEyes 運用 AI 與 ML 技術，自動分析歷史與目前的資料，以偵測網路基礎架構中的異常與中斷情形，強化維持與提升網路效能的能力。

Cisco Secure Network Analytics (formerly Stealthwatch) is a security product that uses machine learning to monitor network traffic in real-time, quickly identifying and responding to anomalies that could indicate security threats. By analyzing historical network traffic data that is categorized as normal or malicious, the system learns to recognize patterns and signatures that are associated with various types of cyber threats, such as malware, ransomware, DDoS attacks, and unauthorized access attempts. When it detects traffic that matches these threat characteristics, it alerts network administrators to take appropriate action. This ability to distinguish between benign and potentially harmful network activity enables organizations to proactively defend their networks against both known and emerging threats.

Cisco Secure Network Analytics（前身為 Stealthwatch）是一款安全產品，運用機器學習即時監控網路流量，迅速找出並回應可能顯示安全威脅的異常情形。透過分析被分類為正常或惡意的歷史網路流量資料，此系統學會辨識與各類網路威脅（例如惡意軟體、勒索軟體、DDoS 攻擊與未經授權的存取嘗試）相關的模式與特徵。當系統偵測到符合這些威脅特徵的流量時，會警示網路管理員採取適當行動。這種區分良性與潛在有害網路活動的能力，讓組織能主動防禦已知與新興威脅。