Intelligent O&M
Overview of Intelligent O&M
In the big data era, traditional O&M based on specified rules cannot meet network O&M requirements, and the insufficiency of automatic O&M capabilities becomes increasingly prominent. Traditional O&M faces the following challenges:
- Traditional O&M is based on SNMP and data is collected in minutes. Data cannot be obtained in real time when a fault occurs.
- In traditional O&M, only device metrics are monitored. However, user experience may be poor when the metrics are normal. Traditional O&M lacks means of correlatively analyzing clients and the network.
- In traditional O&M, network faults can be detected only after receiving users' complaints. As a result, faults cannot be effectively and proactively identified and analyzed.
Huawei's intelligent network analysis platform iMaster NCE-CampusInsight disrupts the traditional monitoring mode that focuses on resource status. This platform applies AI to the O&M field and adopts Telemetry technology to collect performance indicators and log data of network devices. By using big data, AI algorithms, and more advanced analytics technologies as well as leveraging scenario-specific continuous learning and expert experience accumulation, iMaster NCE-CampusInsight frees O&M personnel from nerve-wracking alarms and noises, making user network experience visualized and O&M automated and intelligent.
Intelligent O&M Solution Architecture
iMaster NCE-CampusInsight uses Huawei's big data analytics platform, receives device data through Telemetry technology, and analyzes and displays network data through intelligent algorithms.
Figure 1 Logical architecture of iMaster NCE-CampusInsight
The overall solution architecture contains three layers. The bottom layer is the campus network device layer. It provides data collection capabilities in multiple dimensions, such as client, radio, AP, switch, and user log, and sends the data to iMaster NCE-CampusInsight through Telemetry technology. The middle layer is the iMaster NCE-CampusInsight data analysis layer. It delivers big data storage (for real-time traffic preprocessing, offline traffic distributed processing, and data storage services) and data analysis services (for pattern recognition and intelligent engine). The top layer is the service provisioning layer. It offers customers final data analysis services, including network visibility, campus service analysis, intelligent wireless network, and user application experience.
Category |
Function |
Description |
|---|---|---|
Network visibility |
Network health |
Client access experience: Analyze the access success rate and time consumption fulfillment rate of clients from the association, authentication, and DHCP phases to measure the network access performance. Client roaming experience: Measure the roaming quality of wireless clients based on the roaming success rate and time consumption fulfillment rate, and identify wireless roaming issues. Client throughput experience: Analyze the coverage, capacity health, and fulfillment rate of throughput to check whether the signal coverage meets requirements, whether the network is overloaded, and whether the throughput decreases. Device in-service rate: Collect statistics on the device in-service rate, evaluate the overall network availability, and identify device out-of-service issues. |
Integrated topology |
The integrated topology displays quality analysis and horizontal comparison for each site based on the wireless service access success rate, intelligently analyzes issue patterns, identifies issue boundaries, and provides preliminary analysis of root causes. |
|
Campus service analysis |
Issue analysis |
iMaster NCE-CampusInsight analyzes, identifies, and collects statistics on connection, air interface performance, roaming, and device issues based on data such as performance indicators and logs, and displays information about affected APs and clients based on each issue indicator. |
Access analysis |
The Connectivity module evaluates the overall network connection quality from aspects such as the client access fault event trend. Client access fault event statistics: includes the number of association failures, number of authentication failures, number of DHCP failures, and total number of access times. Client access fault event trend: supports time range selection, displays the distribution of devices and clients that fail to be connected in the trend chart, and displays the distribution of devices and clients with issues in an area chart. |
|
Performance analysis |
The performance experience module evaluates user experience based on RSSI, negotiated rate, and packet loss rate, displays the number and trend of clients with good and poor experience at each time point within a period of time, and analyzes the client distribution trend by single indicator. Poor experience analysis based on APs and clients helps administrators identify APs and clients with the poorest experience. |
|
Protocol trace |
Client access phases including association, authentication, and DHCP are displayed in terms of different protocols. Refined analysis for individual faults that occur during client access is provided based on the protocol interaction result and duration at each phase. The analysis includes the most possible root causes and rectification suggestions for client access failures. Currently, protocol trace supports the following user authentication methods: 802.1X, Portal (Portal 2.0/HTTPS), HACA, and MAC authentication. |
|
Intelligent wireless network |
Intelligent radio calibration |
iMaster NCE-CampusInsight collects KPIs and radio parameters reported by devices and uses intelligent algorithms to calculate the load prediction information in the next calibration period. In addition, it accurately identifies the network topology and edge AP list by utilizing the big data analytics algorithm, and pushes information to devices in response to their requests. Wireless devices perform intelligent radio calibration based on the information delivered by iMaster NCE-CampusInsight and the network information collected in real time. After the radio calibration is complete, the devices periodically report the KPI information and calibration logs of the current network to iMaster NCE-CampusInsight. iMaster NCE-CampusInsight then compares and displays the wireless network parameters before and after the radio calibration. |
WLAN topology |
Network plan import: After the network plan file made by the WLAN Planner is imported, iMaster NCE-CampusInsight displays data such as sites, pre-deployed APs, obstacles, background images, and scale planned in the file. Network comparison: After pre-deployed APs are associated with real APs, the planned data and actual data are compared in terms of the power, channel, frequency bandwidth, number of clients, negotiated rate, and signal strength, and the comparison result is displayed. Wi-Fi heatmap display: The radio heatmap can be displayed based on the AP location. |
|
User application experience |
Application analysis |
Based on the monitoring and analysis of audio and video service sessions, the SIP session statistics, service traffic trend, and session details list can be displayed, helping users quickly learn about the quality status of audio and video services. |
Deployment Design of the Intelligent O&M Solution
The intelligent O&M solution consists of iMaster NCE-CampusInsight, iMaster NCE-Campus, and devices. Currently, iMaster NCE-CampusInsight can manage and perform intelligent analysis on Huawei cloud switches and APs.
- Network bandwidth design
Devices need to periodically report data to iMaster NCE-CampusInsight. Therefore, the campus network needs to reserve bandwidth for data reporting. The average bandwidth consumed by each device is 3 kbit/s.
- Installation location design
iMaster NCE-CampusInsight and iMaster NCE-Campus can be deployed at different locations. They can collaborate with each other as long as network connectivity is achieved. To avoid the instability of the intermediate network, you are advised to deploy them in the same location, for example, a DC.
- Server deployment and selection design
The iMaster NCE-CampusInsight server can be deployed in cluster or standalone mode. Select the deployment mode and node type based on the network scale and configure the corresponding software and hardware resources. For details about the software and hardware requirements, see the iMaster NCE-CampusInsight product documentation of the corresponding version.
Precautions for Intelligent O&M Design
- Precautions for network deployment design:
- During network deployment, ensure that APs' IP addresses are reachable to iMaster NCE-CampusInsight so that the APs can send KPI performance data and logs to it.
- During network deployment, make sure that the device clock is synchronous with the clock of iMaster NCE-CampusInsight. You are advised to deploy an NTP server on the network to synchronize the system clock of the network.
- Precautions for designing the integrated topology function:
- When group fault analysis is performed based on the integrated topology function, only the common tree network topology is supported.
- Precautions for designing the protocol trace function:
- To identify DHCP-related connectivity issues and protocol trace, you need to use the WAC as a DHCP server or enable DHCP snooping on the WAC.
- Precautions for designing the audio and video quality analysis function:
- The audio and video quality analysis function requires devices to send related logs to iMaster NCE-CampusInsight. You are advised to configure the same log sending interval for switches and WLAN devices. The maximum difference between the log sending intervals cannot exceed 20s.
- Precautions for designing the intelligent radio calibration function:
- Intelligent radio calibration and traditional radio calibration cannot be deployed on APs in the same region at the same time.