Software Defined Cognitive Networking: Intelligent Resource Provisioning For Future Networks

Lead Research Organisation: University of Northampton
Department Name: Faculty of Arts, Science and Technology


The non-cooperative competition of network resources between a growing number of adaptive media applications has a significant detrimental impact on user experience and network efficiency. This can lead to knock-on effects to the digital economy and digital public services, which are increasingly dependent on high quality and reliable media streaming. Existing network infrastructures often prioritise improved network coverage and fast packet forwarding functions, which do not always effectively contribute to the improved user experience. Ultimately, the quality of user experience and network efficiency are the two of the most important benchmarks for online media distribution. Future network management must leverage application and user-level cognitive factors in order to allocate scarce network resources effectively and intelligently. this First Grant project aims at developing software defined cognitive networking (SDCN) to ensure the user experience, user-level fairness and network efficiency of online adaptive media using SDN-assisted and QoE-aware resource management. SDCN will lay the groundwork for a great leap from the conventional resource provisioning and traffic engineering schemes to context-aware network management.

In order to achieve its objective, the project will develop a cognitive model based on the analysis of human factors of adaptive media experience, iterations of subjective experiments, and data modelling. The model will enable a non-intrusive QoE assessment service that monitors adaptive media flows and estimates their perceptual user experience using a number of application, service, and network-level metrics. The model will use a purpose-built multi-objective resource allocation function to derive optimal solutions to provision available network resource for the improved user experience, fairness and network efficiency in a network segment.

Planned Impact

The project aims at delivering scientific, societal, industrial and economic impact.
For scientific impact, the project will deliver impact through publications, workshops, academic visits, and student training. The team will aim at publishing SDCN research outcomes such as design, implementation, and experimental results in conference and journals in the areas of multimedia, HCI and networking. This will stimulate further collaborations and innovations on this emerging topic. The team will also participate relevant events hosted by the EPSRC or EPSRC-supported projects. We will visit project partner and other UK research institutions who work on relevant topics to exchange research ideas and to form future collaborations. We will also engage university students to inspire future research on networking and computing technologies. This activity will contribute to the PI's teaching on networking and media technologies. Students will have the opportunities to use SDCN as a use case to study data modelling, online media, software-defined networks, and problem-solving skills.
Public engagement and social enterprise are the main tools for societal impact. We will use a dedicated project website, social media and audio-visual presentations to engage research communities, university students, and the public. As part of the host institution's STEM-themed activities with local schools, the PI will also create opportunities for pupils to engage with novel research from SDCN. We will look into exploiting and transforming technologies developed in SDCN project for the improvements of citizen and environmental wellbeing. Challenges related to data communications in public transportation, health and safety are the expected areas of the impact on social enterprise.
The PI and his team will maximise the industrial impact through industrial engagement, industrial workshop and technology and knowledge transfer. Our partnership with HPE Aruba will provide us with business insights, collaborations with their research and development departments, and access to research facilities. Through industrial engagement, the SDCN team will explore opportunities to pilot technologies derived from the project in industrial prototypes or trial user services. The PI will organise a workshop on cognitive networking in conjunction with a conference with significant industrial attendance. The workshop will provide an ideal platform to stimulate discussions and collaborations with elite researchers and engineers on future intelligent infrastructure design. It will also help disseminating research outcomes from SDCN and other EPSRC-supported projects. We'll release technical demonstrations in the form of live demo, offline sandpit playground, or audio-visual presentation. The intellectual property of our development has the potential for commercial exploitation through licensing.
With respect to the economic impact, the development of SDCN will contribute to the growth of media and internet sector in the UK, and cherish new markets in the area of IoT and creative media while maximising the value of any investment in the networking infrastructure. The project will also provide a strong use case that further drives the growth in fog computing and NFV in the UK.
Description The background research has allowed us to focus our target use case around smart homes where we see increasing complexities in resource management and heterogeneous user devices. Our research also takes into account the impact of connected smart systems at home as well as the implications of 5G to home entertainment. The research has also seen some synergy with virtual reality (VR)-related research within the team on topics such as how the communication networks support distance VR creative process. A new research topic has also stemmed from this grant locally at the University of Northampton where we see a large scale software-defined networking implementation across its new Waterside campus. There have been discussions around algorithmic accountability, fairness, and ethics while network traffic on campus is managed. Our research outcomes are published in IEEE CCNC and ACM Multimedia FAT/MM. The multi-disciplinary research has also led to partnerships with a media company and helped us engaged with the art society, where our findings on human factors reach to wider areas.
Exploitation Route We strive to publish our research findings in reputable outlets and share any open-source libraries or toolkit developed within the project.
Sectors Digital/Communication/Information Technologies (including Software)

Description Our research on human factors and AI have seen impacts on local businesses. In a new partnership with a media company, the team is using its research on video QoE and AI to develop an object detection model to automate brand recognition for product-placement evaluation. The project is expected to lead to a KTP application in 2020. Our work on human attention research also led to a collaboration with the fine art community in developing a new form of artwork that is generated using eye gaze analysis and machine learning. A public exhibition that showcases the joint work has been organised in Milton Keynes Gallery Project Space in Feb 2020. The exhibition was co-sponsored by the project alongside HTC, HP, and Google Tilt Brush.
First Year Of Impact 2020
Sector Creative Economy,Digital/Communication/Information Technologies (including Software),Education,Leisure Activities, including Sports, Recreation and Tourism,Culture, Heritage, Museums and Collections
Impact Types Societal,Economic

Description AI-Assisted Product Detection In Audio-Visual Content
Amount £5,000 (GBP)
Organisation Higher Education Innovation Funding (HEIF) 
Sector Public
Country United Kingdom
Start 12/2019 
End 06/2020
Description QoE in networked media 
Organisation Lancaster University
Country United Kingdom 
Sector Academic/University 
PI Contribution Led research on cross-device multimedia experience and user-centred software defined network designs.
Collaborator Contribution A research team and state-of-the-art testbed on SDN/P4 which allows research experimentations and user testing.
Impact Fawcett, L., Mu, M., Hareng, B., and Race, N., "REF: Enabling Rapid Experimentation of Contextual Network Management using Software Defined Networking", in IEEE Communications Magazine, 2017 Mu, M., et al. "Closing the Gap: Human Factors in Cross-device Media Synchronization", in IEEE Journal of Selected Topics in Signal Processing, 2017. DOI: 10.1109/JSTSP.2016.2638358 Mu, M., Broadbent, M., Hart, N., Farshad, A., Race, N., Hutchison, D. and Ni, Q., "A Scalable User Fairness Model for Adaptive Video Streaming over SDN-Assisted Future Networks", in IEEE Journal on Selected Areas in Communications. 34, 2168-2184, 2016. DOI: 10.1109/JSAC.2016.2577318 Fawcett, L., Mu, M., Broadbent, M., Hart, N., and Race, N., SDQ: Enabling Rapid QoE Experimentation using Software Defined Networking, in proceedings of IFIP/IEEE International Symposium on Integrated Network Management (IEEE IM), Lisbon, Portugal, 05/2017 Sani, Y., Mu, M., Mauthe, A., and Edwards., C., A Bio-inspired HTTP-based Adaptive Streaming Player(GRAND CHALLENGE AWARD), in Proceedings of 2016 IEEE International Conference on Multimedia and Expo (IEEE ICME 2016), Seattle, USA, 07/2016 Mu, M., et al., QoE-aware Inter-stream Synchronization in Open N-Screens Cloud, in Proceedings of the 13th Annual IEEE Consumer Communications & Networking Conference (IEEE CCNC), Las Vegas, USA, 01/2016
Start Year 2015
Description Quality of Experience in 3D Point Cloud 
Organisation University of Koblenz and Landau
Country Germany 
Sector Academic/University 
PI Contribution Establish novel research topics and cross-country funding opportunities in the area of 3D point cloud and user-centred designs. Our contributions include expertise in games design, objective and subjective QoE evaluation, eye-tracking in 3D and virtual reality, and machine learning.
Collaborator Contribution Newly established research lab and post-graduate research team on 3D point cloud capturing and distribution systems.
Impact no
Start Year 2019
Description Software defined networks 
Organisation Hewlett Packard Enterprise (HPE)
Country United Kingdom 
Sector Private 
PI Contribution Post-graduate researchers working on a SDN/P4 testbed environment and studying SDN-assisted fairness in multimedia networks.
Collaborator Contribution Valuable advice from an industrial standpoint. Significant inputs to joint publications.
Impact Fawcett, L., Mu, M., Hareng, B., and Race, N., "REF: Enabling Rapid Experimentation of Contextual Network Management using Software Defined Networking", in IEEE Communications Magazine, 2017 EPSRC SDCN project
Start Year 2016
Description Using AI to assist the evaluation of product placement in TV and films 
Organisation Big Film Group
Country United Kingdom 
Sector Private 
PI Contribution We use our knowledge in multimedia and AI to assist a product placement company to automate the evaluation of brands and products in TV and films using deep learning.
Collaborator Contribution Contribute area knowledge, sample content, and data to support research activities.
Impact Object Detection ML model for brand logos
Start Year 2019
Description Digital Northampton Merged Futures 2019 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach Regional
Primary Audience Industry/Business
Results and Impact In an invited talk, the team discussed its latest research outcomes on human factors and audience understanding in VR multimedia. A demo showcased to the audience how the technology can be used to improve business intelligence, improve user experience, and develop new applications.
Year(s) Of Engagement Activity 2019
Description Project review for Science Foundation Ireland 
Form Of Engagement Activity A formal working group, expert panel or dialogue
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Third sector organisations
Results and Impact Join the expert panel for Science Foundation Ireland to review a research project.
Year(s) Of Engagement Activity 2017
Description Research visit at BBC R&D, London 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Professional Practitioners
Results and Impact Deliver a talk on media synchronization and network intelligence to BBC R&D researchers and engineers, which sparked discussions on topics for future collaboration.
Year(s) Of Engagement Activity 2017