Daemon Methodology
Outline of Objectives within Methodology Steps

(1) Designing a “NI-native architecture” for B5G systems

While current efforts at integrating NI in mobile networks aim at tweaking machine learning solutions so that they fit networking environments, DAEMON upends the approach, and seeks to update the network architecture so that it natively supports NI operations.

(1.1) Unlocking the full potential of NI via coordination of intelligence

DAEMON will propose a novel architectural design that enables comprehensive coordination across the many NI instances that operate across the network infrastructure. To ensure its practical viability and maximize impact, the DAEMON architecture will leverage on and be fully compatible with cutting-edge standardization activities by O-RAN, 3GPP and ETSI.

(1.2) Enabling NI deep into the edge and user plane

DAEMON will develop and demonstrate novel coordinated NI solutions that operate at the VNF level and are implemented in the user plane, as well as in an original Beyond Edge micro-domain that extends the B5G architectural view unprecedently close to the end-terminal.

(2) Developing specialized NI-assisted network functionalities for B5G systems

DAEMON designates a concrete list of key network functionalities for which it will devise and implement NI algorithms capable of taking full advantage of the proposed NI-native architecture, and spanning varied planes, domains and operation timescales.

(2.1) Developing NI-assisted functionalities for B5G performance and efficiency

DAEMON will demonstrate how a properly designed NI can substantially increase the capacity, reduce the latency and improve the efficiency of mobile network systems, helping them to reach the outstanding targets set for B5G technologies.

(2.2) Developing NI-assisted functionalities that enable the sustainable operation of B5G systems

The NI developed by DAEMON will hinge upon new understanding of the energy footprint of network functions, and support the automatic adjustment of the infrastructure energy consumption in balance with desired performance.

(2.3) Developing NI-assisted functionalities that ensure an extreme reliability of zero-touch B5G

The NI developed by DAEMON will be designed for resilience, improving the dependability of network operations

(3) Establishing fundamental guidelines for a pragmatic design of NI

DAEMON will challenge current practices in machine learning for network automation, and innovate the way NI is conceived, by leveraging emerging trends in AI and re-thinking how AI is applied to the network infrastructure environment.

(3.1) Understanding the limits of AI

When designing the NI that will assist the different network functionalities, DAEMON will carefully assess in which cases powerful but non-interpretable AI models such as Deep Learning (DL) ones are appropriate, and when statistical, analytical or hybrid models should be preferred.

(3.2) Designing loss functions that are tailored to the networking context

To ensure that AI models abide by the specific requirement set out by different network functionalities, DAEMON will address the largely unexplored “loss-metric mismatch” problem in the design of NI, devising custom and/or self-learned loss functions.

(3.3) Developing AI models for adaptable NI

NI sets unconventional requirements for AI models, in terms of computational complexity and execution times. DAEMON will carry out seminal work on the design of AI models that can adaptively trade off accuracy for the alternative metrics above, therefore establishing a new paradigm for machine learning techniques conceived to support NI by-design.

(4) Demonstrating the viability and performance of NI-native B5G networks

DAEMON will provide objective evidences of the advantage of a structured, deep, and sensible integration of NI into network infrastructures. This will be achieved by meeting a comprehensive range of Key Performance Indicators (KPIs) on the performance, reliability and sustainability of the proposed solutions, in realistic experimental or measurement data-driven settings

(4.1) Network KPIs

DAEMON will meet the highly challenging requirements imposed by B5G
networks while providing very large gains in terms of efficient utilization of resources and energy consumption.

(4.2) Experimental and data-driven evaluation

DAEMON will validate all NI solutions devised in the project by means of experimental prototypes or of large-scale evaluations based on measurement data.

(5) Industrial and scientific impact

DAEMON will produce the necessary standards, patents and contributions to industry fora in order to:

  1. Attract industry to use the project technology.
  2. Foster the widespread adoption of the technology beyond the manufacturers and operators of the consortium.
  3. Protect the project findings to secure the commercial advantage of the project partners.

It will also produce top-quality publications to ensure the long-term impact of the project results on research and development community focused on B5G systems.

Scroll to Top