purpose: Automated deployment of a new services taking into account vertical requirements, QoS and QoE for every Network service running at the Greek site (under the 5G-EVE architecture)
AI-enhanced MANO can automate the deployment of a new critical service (e.g. Airport Evacuation use case) in a secure and robust way to achieve the use case specific requirements and optimize resource usage and services downtime due to network or general infrastructure problems. For the deployment of a new critical service the steps of AI-enhanced MANO are:
- Critical service requests resources from MEC
- Various metrics, KPIs and SLAs are gathered
- AI-enhanced MANO algorithm decides and performs all deployments and migration actions
AI-Mano REST APIs using the OpenAPI V3 standard is located at: https://ai-mano.5gtours.wings-ict-solutions.dev/
The specification includes:
- Available endpoints (e.g. /kpis) and operations on each endpoint (e.g. GET /kpis, POST /kpis, DELETE /kpis, etc.)
- Operation parameters Input (e.g. kpiID, Identifier of KPI)
- Responses and output for each operation (with samples)
- Schemas, description and example values
- Authentication methods
Final version of API is available at the link: https://ai-mano.5gtours.wings-ict-solutions.dev/
Open Air Interface
The OAI code is available freely on Github, currently with many active branches. Within one of them, a stable version of FeMBMS is available (Branch ‘Develop’). The Further evolved multimedia broadcast multicast service (FeMBMS) is the LTE broadcast-only mode standardized in 3GPP Rel.14 where several modifications were made to enhance the system architecture and media formats flexibility for content providers, broadcasters, and mobile network operators.
- To obtain an SDR receiver compliant with the 3GPP Release 16 specifications with:
- At least one OFDM (Release 16) numerology.
- CAS subframe with Release 16 improvements (New AL and PBCH Repetition).
- To improve the synchronization mechanism and the stability.
- To contribute with OAI (Branch Develop) with the final version of the code.
- New OFDM numerologies → Test phase (ongoing).
- CAS subframe Release 16 → Test phase (completed).
- Initial synchronization → Improved (completed).
- Sync tracking → Debug phase (ongoing).
- New OFDM numerologies → Debug phase.
- CAS subframe Release 16 → Done (to be further tested).
- Synchronization → Done (to be further tested).
Service Layer SDK descriptions
- AI-Agents in OSM is fully Open-source
- Atos contribution, see https://osm.etsi.org/wikipub/index.php/OSM_PoC_11_Deployment_of_AI-Agents_in_OSM
- ETSI ENI PoC for Autonomous Network Slice Management and Vertical closed-loop integration is Open-source
- several partners contributing, see https://eniwiki.etsi.org/index.php?title=Poc_09:_Autonomous_Network_Slice_Management_for_5G_Vertical_Services
- Multicast/broadcast support will be proprietary implementation, but open Northbound API for the multicast service for 3GPP Release 17 will be hosted by 3GPP (EXP/Enensys contributing)
- More information here 3GPP 5G OPEN API
- AI-enhanced MANO implementation on top of 5G-EVE infrastructure in the Greek node intend to produce a github repository of REST APIs and manuals, as well as sample code to use those APIs (Wings contributing).