{"generator":"Jekyll","link":[{"@attributes":{"href":"https:\/\/remyraes.com\/feed.xml","rel":"self","type":"application\/atom+xml"}},{"@attributes":{"href":"https:\/\/remyraes.com\/","rel":"alternate","type":"text\/html"}}],"updated":"2026-03-24T18:15:39+01:00","id":"https:\/\/remyraes.com\/feed.xml","title":"remyraes","subtitle":"PhD candidate based in Lille, France.","entry":[{"title":"FLInD paper submitted","link":{"@attributes":{"href":"https:\/\/remyraes.com\/flind-paper-submitted.html","rel":"alternate","type":"text\/html","title":"FLInD paper submitted"}},"published":"2026-01-22T10:44:00+01:00","updated":"2026-01-22T10:44:00+01:00","id":"https:\/\/remyraes.com\/FLInD%20paper%20submitted","content":"

I spent the second semester of 2025 working on an extension of our previous FLI<\/a> contribution.\nThis new paper presents FLInD<\/strong>, for Fast Linear Interpolation in n-Dimension<\/em> (yes we\u2019re very good at naming our algorithms), a generalization of FLI<\/strong> that supports compression of multiple dimensions within a single model.<\/p>\n\n

You can read the preprint on HAL<\/a>.<\/p>","author":{"name":{}},"category":{"@attributes":{"term":"blog"}},"summary":"I spent the second semester of 2025 working on an extension of our previous FLI contribution. This new paper presents FLInD, for Fast Linear Interpolation in n-Dimension (yes we\u2019re very good at naming our algorithms), a generalization of FLI that supports compression of multiple dimensions within a single model."},{"title":"INTACT paper accepted!","link":{"@attributes":{"href":"https:\/\/remyraes.com\/intact-paper-accepted.html","rel":"alternate","type":"text\/html","title":"INTACT paper accepted!"}},"published":"2025-06-26T17:11:00+02:00","updated":"2025-06-26T17:11:00+02:00","id":"https:\/\/remyraes.com\/INTACT%20paper%20accepted","content":"

Last year, I presented the storage part of our FLI<\/a> contribution to the DAIS conference;\nI\u2019m glad to announce the extended, full version of the paper, including the privacy section, was accepted for publication in the Journal of Internet Services and Applications<\/a>.<\/p>\n\n

You can read the preprint on HAL<\/a>.<\/p>","author":{"name":{}},"category":{"@attributes":{"term":"blog"}},"summary":"Last year, I presented the storage part of our FLI contribution to the DAIS conference; I\u2019m glad to announce the extended, full version of the paper, including the privacy section, was accepted for publication in the Journal of Internet Services and Applications."},{"title":"AndroWatts paper accepted!","link":{"@attributes":{"href":"https:\/\/remyraes.com\/androwatts-accepted.html","rel":"alternate","type":"text\/html","title":"AndroWatts paper accepted!"}},"published":"2025-02-27T12:50:00+01:00","updated":"2025-02-27T12:50:00+01:00","id":"https:\/\/remyraes.com\/%20AndroWatts%20accepted","content":"

In this paper, named AndroWatts: Unpacking the Power Consumption of Mobile Device\u2019s Components<\/strong>, we use\ncomponent-level power measurement probes to split the smartphone\u2019s energy consumption between said components,\nand build a power usage model with a high prediction accuracy.<\/p>\n\n

\n

Power efficiency is crucial for mobile devices, where software inefficiencies can rapidly drain battery\nlife and reduce device longevity. To help developers diagnose inefficiency issues and optimize power use,\nthis paper introduces a novel methodology for estimating power consumption at a hardware component level\nin mobile devices. Unlike existing approaches that rely on coarse-grained battery discharge measurements,\nour approach consists of modeling per-component power usage as an inverse problem, utilizing linear\nstatistical models and system metrics to estimate the contributions of individual components.\n
\n
In this paper, we model the total power usage using system metrics, we evaluate the accuracy of\nper-component power estimations, and we investigate the impact of dataset size on the model\u2019s performance.\nOur empirical evaluation shows that linear models achieve high predictive accuracy, with R-squared values\nexceeding 0.90 for total power usage, and robust correlations are observed between predicted and ground-truth\npower consumption for key components such as the CPU and GPU. We also observe that the amount of measures\nrequired to build such a model is of the order of magnitude of the hundred rather than the thousand.<\/p>\n<\/blockquote>\n\n

You can read the publication on HAL<\/a>.<\/p>","author":{"name":{}},"category":{"@attributes":{"term":"blog"}},"summary":"In this paper, named AndroWatts: Unpacking the Power Consumption of Mobile Device\u2019s Components, we use component-level power measurement probes to split the smartphone\u2019s energy consumption between said components, and build a power usage model with a high prediction accuracy."},{"title":"MPC paper accepted!","link":{"@attributes":{"href":"https:\/\/remyraes.com\/mpc-accepted.html","rel":"alternate","type":"text\/html","title":"MPC paper accepted!"}},"published":"2024-12-31T12:07:00+01:00","updated":"2024-12-31T12:07:00+01:00","id":"https:\/\/remyraes.com\/MPC%20accepted","content":"

In this paper, named Application of a predictive method to protect privacy of mobility data<\/strong>, we use the\nFLI technique<\/a> to predict where users are going, to forecast whether there will be privacy issues in\nthe near future, and reacting accordingly by enabling protections before<\/em> said privacy issues occur.<\/p>\n\n

\n

Users of geo-localized applications on mobile devices need protection to avoid threats to their privacy.\nSuch protection should vary in time, to cope with the dynamical nature of mobility data. We present a method\nto protect the privacy of users of location-based services, based on Model Predictive Control techniques. We\nemploy three different predictors for future movements: an exact predictor, which serves as the baseline for\nthe best expected performance, and two additional predictors allowing for online implementation. One of these\npredictors assumes the user is moving in a way that minimizes privacy,\nwhile the other is a linear predictor<\/a>.\n
The method has been applied to two datasets, Privamov and Cabspotting, which contain mobility data collected\nfrom real users when using a mobile device. The method demonstrated an improvement in privacy compared to a\nstate-of-the-art mechanism by approximately 12% increase for Privamov users and 5% for Cabspotting users, while\nmaintaining the same level of utility.<\/p>\n<\/blockquote>\n\n

You can read the publication on ScienceDirect<\/a>.<\/p>","author":{"name":{}},"category":{"@attributes":{"term":"blog"}},"summary":"In this paper, named Application of a predictive method to protect privacy of mobility data, we use the FLI technique to predict where users are going, to forecast whether there will be privacy issues in the near future, and reacting accordingly by enabling protections before said privacy issues occur."},{"title":"FLI paper accepted!","link":{"@attributes":{"href":"https:\/\/remyraes.com\/fli-paper-accepted.html","rel":"alternate","type":"text\/html","title":"FLI paper accepted!"}},"published":"2024-05-16T15:52:00+02:00","updated":"2024-05-16T15:52:00+02:00","id":"https:\/\/remyraes.com\/FLI%20paper%20accepted","content":"

Good news, everyone!<\/p>\n\n

The storage part of our FLI<\/a> paper was accepted at\nDAIS\u201924<\/a>, which will take place\nin Groningen, Netherlands, from June 17th to June 21st.<\/p>\n\n

You can read the preprint on HAL<\/a>.<\/p>","author":{"name":{}},"category":{"@attributes":{"term":"blog"}},"summary":"Good news, everyone!"},{"title":"Venice abstract accepted!","link":{"@attributes":{"href":"https:\/\/remyraes.com\/venice-abstract-accepted.html","rel":"alternate","type":"text\/html","title":"Venice abstract accepted!"}},"published":"2024-05-16T15:52:00+02:00","updated":"2024-05-16T15:52:00+02:00","id":"https:\/\/remyraes.com\/Venice%20abstract%20accepted","content":"

Our quest for more device-to-device communication continues with Venice<\/em>,\na framework that (theoretically) enables local communication between any device,\nmobile or desktop.<\/p>\n\n

This work will be presented in two locations:<\/p>\n