[LIGHT] A new ERC Proof of Concept at EURECOM brings new wireless communications standards for VoD
In the context of communication networks, the use of caching is a fundamental ingredient for handling the exponential growth in data volumes [1]. In computing, a cache is a storage layer typically used in a transient fashion, mainly aiming to access and serve requested data much faster than with the use of the initial storage. LIGHT, the new ERC PoC of EURECOM’s Professor Petros Elia, aspires to provide a cache-aided solution, in order to serve better the increasing number of users of Video On Demand (VoD) services. Through this grant, his solution could potentially reach our phones; or at least their cache memories.
Q. Congratulations on your second ERC! What were the steps that lead to this success?
PE. Thank you! Well, 6 years ago I applied for my first ERC Consolidator grant for the project DUALITY, which was mainly about caching. In my opinion, it was a successful project during which, we were able to solve some fundamental theoretical problems, which may have an impact in practice. Therefore, I decided that it is a good idea to apply for an ERC PoC, which is designed for this type of scenarios; namely solidifying the commercial potential of previous ERC research results. In order to obtain a PoC, you have to create a business plan and a short market survey, as well as a structure of the team that you envision.
So, I received this ERC PoC in August 2022 and it may start in the beginning of next year, with a 18-month duration. The allocated sum (150k) can be used on demonstrating results, hire consultancy and any other essential components that could ensure the success of this next step. In our case, we hope to be able to achieve a result close to commercialisation, but of course it is easier said than done.
Q. Could you briefly describe the context of LIGHT project?
PE. LIGHT project is about using the cache memory in your phone or laptop to speed up wireless communications or even wired ones in some cases. We are dealing with network load reduction, mainly for Videos on Demand (VoD) by using cache-aided folding techniques. Equivalently, our approach allows ultra-fast massive downloads, thanks to cached storage, regardless of the increase in network load. No user coordination is needed. Our techniques allow us to elevate even the most advanced, cutting edge technologies. Note that our results can be applied to any cachable content, like movies and audio, but not your zoom calls!
Now, imagine that there is a Netflix server having to potentially serve 1000 people, each wanting to see their own separate movie of 1 Gigabyte. This means that normally, the amount of data to be transmitted in this scenario would be 1000 Gb. What LIGHT and my previous ERC aim, is to reduce such load on the server, by intervening both at the placement and delivery phase of caching, using mathematical tools from graph and information theory. For example, if you allow me to store a very small fraction of the movie in the cache memory on your phone, the transmission load of the server could drop by a factor of 3x-5x, namely to about 200–300Gb. For the market, this would mean an extreme cost reduction, especially in network configurations that have massive synchronous download or streaming requirements, like student dormitories or hotels.
Q. What are the next steps and your vision for LIGHT project?
PE. In order to be successful, we would need to carefully construct a team of collaborators, with the relevant expertise to guide us through the whole process; from understanding the project’s potential to concluding a proof of concept. It is quite rewarding to create useful research, whether this is for reducing costs, or for making users’ lives easier.
REFERENCES
- Federico Brunero, Petros Elia (2022). Unselfish Coded Caching Can Yield Unbounded Gains Over Selfish Caching. IEEE Transactions on Information Theory.
- H. Zhao, A. Bazco-Nogueras and P. Elia, “Vector Coded Caching Multiplicatively Increases the Throughput of Realistic Downlink Systems,” in IEEE Transactions on Wireless Communications, 2022, doi: 10.1109/TWC.2022.3213475
- E. Parrinello, A. Ünsal and P. Elia, “Fundamental Limits of Coded Caching With Multiple Antennas, Shared Caches and Uncoded Prefetching,” in IEEE Transactions on Information Theory, vol. 66, no. 4, pp. 2252–2268, April 2020, doi: 10.1109/TIT.2019.2955384.
- https://erc.europa.eu/news/second-round-recipients-2022-proof-concept-call-2-funding
- https://erc.europa.eu/sites/default/files/document/file/erc-2022-poc-2-results-list.pdf
- https://erc.europa.eu/news-events/magazine/survey-poc-academic-entrepreneurship
