5G-MEDIA: Remote Production

5G Context

Today’s broadcasts of events require large teams on location, at least one OB (outside broadcasting) van, as well as  time-consuming installation of audio and video equipment and control room set-up for audio-video engineers and directing teams. On top of this, the broadcasting and media industry needs to find ways of reducing steadily rising costs and complexity. 

Combining 5G technologies in smart ways can help lower costs and enable time-saving production methods like remote production. This is the context for the 5G-MEDIA pilot, which implements remote and smart production over 5G networks for low-latency and high-bandwidth media streaming. In practical terms, this means using the control room at the broadcaster’s facilities remotely so less equipment and crew are needed on site during the production process. 

Goals of the Pilot

• Overcoming limitations in traditional broadcast productions by enabling remote productions from anywhere without needing dedicated infrastructure at the event site. 

• Ensuring the efficient development and deployment of software-based virtualised media functions and dynamic network and transmission optimisation based on the demands of the transfer through the use of MAPE (Monitoring Analyse Planning Execute) and CNO (Cognitive Network Optimizer).

• Significantly lowering costs, personnel, time and complexity for remote production with high impacts on the media industry. 

• Target KPIs: low latency and high throughput. 

• Performance KPIs: keep the maximum tolerable packet loss rate at the application layer within the maximum tolerable end-to-end latency for that application.

Architecture and Pilot Overview

The architecture for the 5G-MEDIA pilot revolves around the set-up of a Remote Production, where cameras and audio equipment at the venue are connected via a 5G network to media production applications. 

These media production applications are deployed and orchestrated by the 5G-MEDIA Service Virtualisation Platform (SVP). The SVP ensures that the media processing functions are embedded within the network and cloud infrastructure enabling low latency and high throughput as required by live streaming and media processing. 

Deployment

The 5G-MEDIA pilot deployment fat the RTVE broadcaster’s site in Madrid focuses on key performance and business KPIs.

• Continuous monitoring of the performance KPI parameters for maximum tolerable packet loss rate and feeding them into the optimisation algorithm of the CNO Engine. MAPE applies a reinforcement learning algorithm to adjust video compression levels to maximise Quality of Experience (QoE) in the presence of dynamically varying background traffic and congestion levels. 

• Monitoring of performance KPIs like service deployment time, that is, the time for setting up end-to-end workflow for supporting media services and virtualisation infrastructure scalability, that is, the ability to support, seamlessly instantiate, migrate and up or downscale media-related virtualised services.  

• Assessing business KPIs, such as the reduction of network and service management OPEX for remote broadcast productions stemming from the use of virtualisation and MANO (management and network orchestration) platforms for deploying virtualised functions and tuning technical performance parameters to meet business-level policies. 

Outcomes and Impacts

Achievements from the 5G-MEDIA pilot, live demonstration and extensive evaluation in real-world environments (TID and RTVE infrastructure) include:

• Development and flexible deployment of virtualised and flexible media services: Compression Engines, Media Process Engine, Speech to Text.

• Support of the SMPTE ST2110 video over IP standard.

• Definition and implementation of a QoE Probe and Publisher supporting optimisation.

• Implementation of Machine Learning algorithms to adapt video quality profiles and compression levels to available network and computational resources based on Technical Guidelines from the European Broadcasting Union (EBU), RTVE and IRT.

Vertical partners involved: IRT GmbH, RTVE 

Also involved: BitTubes GmbH, Madrid University of Technology, Singular Logic, University College London

Other partners: Engineering Ingegneria Informatica SpA (Coordinator), CERTH - Centre for Research and Technology Hellas, Hellenic Telecommunications Organisation S.A. (OTE), IBM, Interinnov SAS, NEXTWORKS SrL, NETAS Telecommunications Inc., Telefonica R&I

5G-MEDIA Website | @5GPPP5GMEDIA