Periodic Reporting for period 1 - CHARITY (Cloud for Holography and Cross Reality)
Période du rapport: 2021-01-01 au 2022-06-30
While electronic means of communications have become a commodity in modern societies, their availability and those of various Internet-based tools have proven to be a vital element in the current pandemic situation, to maintain social interactions, to keep businesses going, etc. Equipped with that experience and with the pandemic threat being an ongoing concern, work on novel technologies that will make communications more immersive has received a push. Such technologies must enable immersive communication applications like Virtual Reality (VR), Augmented Reality (AR) or Holography to become largely available, reliable, and commercially sustainable. From a network standpoint, they define a new class of services where best-effort and simple traffic differentiation approaches are insufficient to meet their strict requirements. Studies showed that for an acceptable user experience with high fidelity, the latency should be less than 15ms and the bandwidth ranges from 1Gbps up to 30Gbps. Even when considering the advances of 5G, these applications pose a huge challenge to the network and the entire computation infrastructure.
In this context, the EC-funded R&I project CHARITY has set off in order to address these challenges and to develop and prototype a number of related use cases. CHARITY aspires to leverage the benefits of intelligent, autonomous orchestration of cloud, edge, and network resources, to create a symbiotic relationship between low and high latency infrastructures that will facilitate the needs of emerging applications.
Looking at the overall concept, the overarching vision of CHARITY is the development of a unified framework ensuring a complete cycle of highly interactive services management, spanning from CI/CD to life cycle management (LCM) and orchestration.
In this context, the EC-funded R&I project CHARITY has set off in order to address these challenges and to develop and prototype a number of related use cases. CHARITY aspires to leverage the benefits of intelligent, autonomous orchestration of cloud, edge, and network resources, to create a symbiotic relationship between low and high latency infrastructures that will facilitate the needs of emerging applications.
Looking at the overall concept, the overarching vision of CHARITY is the development of a unified framework ensuring a complete cycle of highly interactive services management, spanning from CI/CD to life cycle management (LCM) and orchestration.
The research work of CHARITY is organized in four research Work Packages (WP1-4), in addition to one Work Package (WP5) for dissemination and impact creation, and another (WP6) for project management.
The project work has been started with the definition of the reference scenarios and the refinement of the seven project use cases which are (a) Real-time Holographic Applications, b) Immersive Virtual Training and c) Mixed Reality Interactive Applications). Their technical as well as some general-purpose requirements have been identified and these were used to design and derive the technical specifications of the CHARITY architecture. The architecture has been designed after a thorough investigation of most relevant standards and specifications as well as different relevant architectures defined in other EU consortia. The results were documented in D1.2 and D1.3.
WP2 investigated tools and algorithms aiming at the creation of an efficient end-to-end orchestration framework that manages the creation and lifecycle management of multiple XR services over the cloud/edge continuum. For this purpose, relevant tools and algorithms have been investigated. An AI based resource-aware orchestration framework has been proposed. To ensure acceptable XR service provisioning, work on proactive prediction of QoS has been done. Several traffic routing mechanisms have been explored that can support extremely interactive, bandwidth intensive and latency sensitive end-to-end networking. Moreover, security and privacy concerns of each CHARITY use case have been analysed, and several solutions and open-source tools that can ensure the security of the Charity platform have been investigated. Work has also started to provide for the XR application developer the needed tools and instruments allowing them to exploit the capabilities developed by CHARITY.
WP3 has been investigating solutions to efficiently exploit network and computing resources (e.g. Multi/Many core CPUs, GPUs and FPGAs on edge devices, network-aware data caching/storage) and 2) how to make the application adaptive (e.g. adaptive/contextualized rendering). An initial monitoring framework based on the open-source technology Prometheus has been proposed. WP3 has been also working towards the design of the CHARITY Edge Storage Component (CHES) along with its basic functionalities, namely data storage, retrieval tasks, security and privacy protection capabilities, QoE insurance and mitigation. A first implementation is close to be released. WP3 has been also looking into solutions that make XR services resilient to fluctuations in network resources.
Discussions have been started regarding which mechanisms and components are going to be developed and their relationship with the CHARITY framework. Partners were asked to identify the components and interfaces that they will develop. In D4.1 we documented the initial mapping between components, CHARITY architecture, responsible partners and additional involved partners. Aspects related to integration, namely, the design and development of a single testbed versus multiple testbeds at partners premises have been discussed too, and how the specific components and concepts will be integrated and showcased.
CHARITY has also established a communication strategy and related procedures. A project website has been established, various information material has been created and the social media presence was setup. In terms of exploitation and innovation management, CHARITY investigated the multiple market domains targeted by XR services and applications, focusing in particular on the areas interested by the use cases that will be implemented and evaluated within the project, which are the most natural and best candidates to exploit the project results. Finally, CHARITY analysed the standardization bodies in scope of the project and linked them to the CHARITY architecture components so as to adopt these standards and/or further develop them, whenever needed.
The project work has been started with the definition of the reference scenarios and the refinement of the seven project use cases which are (a) Real-time Holographic Applications, b) Immersive Virtual Training and c) Mixed Reality Interactive Applications). Their technical as well as some general-purpose requirements have been identified and these were used to design and derive the technical specifications of the CHARITY architecture. The architecture has been designed after a thorough investigation of most relevant standards and specifications as well as different relevant architectures defined in other EU consortia. The results were documented in D1.2 and D1.3.
WP2 investigated tools and algorithms aiming at the creation of an efficient end-to-end orchestration framework that manages the creation and lifecycle management of multiple XR services over the cloud/edge continuum. For this purpose, relevant tools and algorithms have been investigated. An AI based resource-aware orchestration framework has been proposed. To ensure acceptable XR service provisioning, work on proactive prediction of QoS has been done. Several traffic routing mechanisms have been explored that can support extremely interactive, bandwidth intensive and latency sensitive end-to-end networking. Moreover, security and privacy concerns of each CHARITY use case have been analysed, and several solutions and open-source tools that can ensure the security of the Charity platform have been investigated. Work has also started to provide for the XR application developer the needed tools and instruments allowing them to exploit the capabilities developed by CHARITY.
WP3 has been investigating solutions to efficiently exploit network and computing resources (e.g. Multi/Many core CPUs, GPUs and FPGAs on edge devices, network-aware data caching/storage) and 2) how to make the application adaptive (e.g. adaptive/contextualized rendering). An initial monitoring framework based on the open-source technology Prometheus has been proposed. WP3 has been also working towards the design of the CHARITY Edge Storage Component (CHES) along with its basic functionalities, namely data storage, retrieval tasks, security and privacy protection capabilities, QoE insurance and mitigation. A first implementation is close to be released. WP3 has been also looking into solutions that make XR services resilient to fluctuations in network resources.
Discussions have been started regarding which mechanisms and components are going to be developed and their relationship with the CHARITY framework. Partners were asked to identify the components and interfaces that they will develop. In D4.1 we documented the initial mapping between components, CHARITY architecture, responsible partners and additional involved partners. Aspects related to integration, namely, the design and development of a single testbed versus multiple testbeds at partners premises have been discussed too, and how the specific components and concepts will be integrated and showcased.
CHARITY has also established a communication strategy and related procedures. A project website has been established, various information material has been created and the social media presence was setup. In terms of exploitation and innovation management, CHARITY investigated the multiple market domains targeted by XR services and applications, focusing in particular on the areas interested by the use cases that will be implemented and evaluated within the project, which are the most natural and best candidates to exploit the project results. Finally, CHARITY analysed the standardization bodies in scope of the project and linked them to the CHARITY architecture components so as to adopt these standards and/or further develop them, whenever needed.
CHARITY has been researching to design a hierarchical orchestration system, along with supporting algorithms and mechanisms that manage in a cognitive manner both of the XR services and the underlying infrastructure. The XR services are dynamically divided into a set of sub-services that can be deployed unto different domains, whilst the resources and domains are dynamically agglomerated into bigger domains or vice versa. This approach introduces a potential number of possible configurations by leveraging intelligent and Reinforcement Learning-based algorithms. Moreover, CHARITY is proposing novel ANN models for resource and network state prediction mechanisms and also dynamic routing mechanisms to ensure the satisfaction of KPIs during the lifetime of XR services. CHARITY has also defined an approach for determining XR blueprints along with mechanisms that ensure their secure delivery.
In terms of supporting the CHARITY Use Cases and the CHARITY framework, several innovative technological solutions will advance the state of the art in XR application deployment, monitoring, data management, and performance optimization. Together with new graphics data compression algorithms, these will allow the development of more sophisticated and performing XR applications. SME involved in complex XR application development will greatly benefit from these technological solutions and assets.
In terms of supporting the CHARITY Use Cases and the CHARITY framework, several innovative technological solutions will advance the state of the art in XR application deployment, monitoring, data management, and performance optimization. Together with new graphics data compression algorithms, these will allow the development of more sophisticated and performing XR applications. SME involved in complex XR application development will greatly benefit from these technological solutions and assets.