ISG ENI Use Cases

From wiki
Revision as of 18:09, 15 October 2021 by Forbesr (talk | contribs) (ISG ENI Use Cases)
Jump to: navigation, search

ISG ENI Use Cases

ISG ENI - Industry Specification Group on Experiential Networked Intelligence

ENI has published versions 3 of the Use cases

  • This page highlights prominent use cases

ENI Use Cases

1 Introduction

The ENI ISG has identified a series of use cases, exemplar of the benefits of applying an ENI system to understand the operating status of networks and networked applications in near-real-time, to reconfigure these networks according to business goals, and to find bottlenecks of service and/or failure of the network. The possibility of taking all these advantages on-demand, in response to changing of contextual information, expands the applicability of the ENI system concept in a few sets of types of scenarios, or categories, identified by the ENI community.

See also

2 Infrastructure Management

This category of use cases covers the processes related to the management of the network infrastructure. The use cases in this category use policies for managing the network infrastructure, enabled by placing analytics in the control loop and using the results of the analytics-based AI mechanisms as part of the input to policy-based management of the infrastructure. Closed control loops continually monitor and adjust performance as necessary. These uses cases include:

  • The intelligent link load-balancing and bandwidth allocation between Internet Data Centres (IDCs). An ENI System uses policies to balance network traffic, ensuring bandwidth for the tenant and improving bandwidth utilization among IDC tenants,
  • Smart resource allocation, which uses an ENI System to avoid service degradation and/or disruption during planned events that consume resources and threaten SLAs. AI mechanisms plan different context-based scenarios, and policies implement network reconfiguration to ensure SLAs, and
  • The reduction of energy costs, by using the ENI System to move services to selected servers and idling or powering down others. The ENI system then predicts peak hours, waking up the necessary number of servers to share the load.

3 Network Operations

Use cases described in this category are concerned with running the network, where the runtime contexts of the network are extracted and analyzed, and the management operations are performed and optimized dynamically at runtime. These uses cases include:

1. Dynamic Policy-based IP address allocation, where AI and policies dynamically assign and optimize IP address pools,

2. An ENI System enables radio coverage and capacity optimization to adjust to context changes in order to provide the required capacity in coverage areas, to minimize interference and maintain an acceptable quality of service by instructing base stations to adjust the appropriate RF parameters,

3. Smart network rollouts, so that operators can define different policies for different types of rollouts and for different types of resources, applied by an ENI System,

4. An ENI System provides an efficient, context-sensitive optimization framework for the next generation fronthaul interface, which is required to support the functional split between remote and centralized units in a Radio Cloud Centre,

5. An ENI System uses AI and policies to dynamically manage network slicing to adapt to context-sensitive load changes automatically,

6. An ENI System is able to automate cloud resource composition processes, which have become an essential requirement given the growing use of these type of resources for the provision of network services, by applying policies defined for them.

7. An ENI System can detect subtle patterns that can have a large impact on user experience, by discovering how user perception of service QoE relates to expectations for application performance and network performance.

8. An ENI system identifies traffic with opaque payloads or port allocations by learning other relevant patterns. This helps on network traffic classification that has a key role in operation and management, challenged by the growing use of encryption down the protocol stack.

9. High-precision time synchronization is one of the key requirements of 5G networks and many of the services it supports. An ENI System can be used to accurately predict time offset and time skew rate, by reducing clock deviations from the standard time, which is particularly relevant as high-precision time synchronization is one of the key requirements of 5G networks and many of the services it supports, and

10. An ENI System supports tailor-made QoS requirements for each mobile gaming wireless link, by addressing network latency, which is crucial in user experience for real time online games, directly affecting the performance of players in these games.

4 Service Orchestration and Management

This category of use cases shows how an ENI System can assist service orchestration and management of processes such as service activation, by using the operator's business channels or customer portals as well as providing differentiated SLAs for different applications. This includes:

  • Context-aware VoLTE service experience optimization, by applying an ENI System to collect and analyze RAN data, and optimize the VoLTE service experience adaptively and responsively in contrast to time-consuming and inefficient manual field tests,
  • Intelligent network slicing management, with advanced automation and AI algorithms applied holistically to achieve runtime deployment and adaptation of the network slice instances,
  • Intelligent carrier-managed SD-WAN, where an ENI System uses AI, policies, and context-awareness to monitor the networks and help to optimize the services and resources of SD-WAN user enterprises ,
  • Intelligent caching based on prediction of content popularity, reducing the backhaul bandwidth cost of cellular networks, and reducing the content access delay of mobile users, can be achieved by using artificial intelligence prediction modules and other supporting modules, provided by the ENI System, and
  • Service experience optimization of E2E slicing involving both OSS and BSS, can be achieved by using E2E QoE prediction and interaction with the client/tenant. Based on the predicted QoE, the ENI System provides (re)configuration recommendations and/or commands to the OSS and BSS.

5 Assurance

Use cases in this category are concerned with the functionality of network monitoring, trending, and prediction, as well as taking policy-based actions by using knowledge learned from the network to facilitate network maintenance. This includes:

· Network fault identification and prediction, to proactively identify and forecast status of a service that is not performing as expected, and repair the service before customer SLAs are violated,

· Assurance of service requirements, to support context-aware resource allocation in a virtualized environment. An ENI System continuously monitors services, predicting faults and triggering the most appropriate and optimal actions for mitigation, such as slice prioritization enforcement, and

· Network fault root-cause analysis and intelligent recovery, where AI algorithms are used by the ENI System to calculate the self-recovery policies based on alarm, network topology, and network service data collection.

6 Network Security This category uses AI to address network security, and includes: · Policy-based network slicing for IoT security, using AI to address specific situations that involve DDOS attacks and providing automatic and dynamic actions in different contexts. · Limiting profit in cyber-attacks that leverages NFV technologies by the ENI System to address dynamic detection and mitigation of threats such as ransomware and crypto-mining.

More information is available on the ETSI portal at:

Use Cases Specification (In Revision)

  • ETSI GS ENI 001
Work Item Title: Experiential Networked Intelligence (ENI); ENI use cases
Target procedure or publication version: V3.2.1
Version 3.1.1 available click [here]