MODANE: Mobile Data Analytics at the Network Edge
- Reference number
- IS14-0061
- Start and end dates
- 150101-171231
- Amount granted
- 1 875 000 SEK
- Administrative organization
- KTH - Royal Institute of Technology
- Research area
- Information, Communication and Systems Technology
Summary
We aim to develop a mobile system architecture and algorithms that pave the way for a paradigm shift, and enable the formation of spontaneous mobile clouds for solving computational tasks using the UEs’ computational power in a collaborative way, while optimizing communication and computation subject to resource availability, and to mobility. The planned work is structured into three work packages (WPs). WP1 will focus on the design and evaluation of the infrastructure required to enable mobile data analytics. WP2 will address the algorithmic needs for efficient data analytics in the targeted mobile data analytic infrastructure, including task assignment and computational issues. In order to provide a proof-of-concept for the architecture and for the algorithms develop in WP1 and WP2, WP3 is devoted to developing a prototype implementation of the solutions. The project will contribute to the technological advances in future mobile systems. Research outputs with scientific and technological novelty and contribution will be published in prestigious international journals and conferences and core technologies will be secured by IPR. In addition, the project will contribute to the training of highly qualified professionals (HQP), including master students, PhDs and post-docs, which will develop expertise in a research program with high industrial relevance linking technological enablers with algorithmic foundations.
Popular science description
We are witnessing a proliferation of smart devices, from mobile phones through smart watches and glasses, automated power meters and smart home gateways, to connected cars. Smart devices will have many application areas, including remote medical monitoring and diagnostics, virtual reality, building energy use management and vehicular traffic optimization. The applications may seem very different, but they have in common that they depend on the smart devices to collect data, often personal data, from their environment, which then have to be processed. The current best practice is to send the data to large data centers where they are organized and analyzed, and the results are communicated back. The consequence is a dramatic need for communication bandwidth in mobile networks and power hungry data centers. In this project we propose to design a system architecture that allows data to be kept local and to be processed among collaborating mobile nodes. The nodes communicate with each other and can use each others' computational resources if that is most efficient. The architecture is compatible with the existing and future mobile communication infrastructures, and leverages the infrastructure to optimize the efficiency of the communication and the distributed computation among the nodes. It allows for significant savings on communication infrastructure and on computational needs in the data centers, and at the same time it improves information availability. The project will make contributions in two important areas. First, it will develop the design of the architecture that can support cooperative data analysis and at the same time is compatible with future mobile network architectures. Second, it will develop algorithms that are needed to optimize the interaction between the nodes that need to perform the computations.