Newest papers

Injection of Space Radiation Errors on Low-power NVM using Generative Deep Learning

Eleonora Giudici, Antonio Miele, Luca Mottola · 2026 · arXiv (Cornell University)

DOI: 10.48550/arxiv.2610.26724 · cited by 0 works (OpenAlex) · open-access PDF

German

Wir entwickeln wahrscheinlich ein Fehlermodell für ein Non-Volatile Memory (NVM) Gerät, das die Auswirkungen von physikalischen Fehlern in einer CubeSat-Mission in Low Earth Orbit (LEO) darstellt. Das CubeSat trägt ein Off-the-Shelf Low-Power Ferroelectric RAM (FRAM) Chip und führt wiederholte Lesen/Schreiben-Tests nach einem vordefinierten Muster. Wir vergleichen jeden Speicher-Snapshot, der von dem Satelliten empfangen wird, mit einer Referenz-Bitmap, um Bitfehler zu identifizieren. Wir analysieren visuell fehlerhafte Werteverteilung in den Speicher-Snapshots und wir verstehen, dass sie häufig nach regelmäßigen räumlichen Mustern verteilt werden. Wir definieren eine

Machine translation (Meta M2M100); the English original follows.

Abstract

We develop an error model for a Non-Volatile Memory (NVM) device that represents the effects of physical faults experienced in a CubeSat mission in Low Earth Orbit (LEO). The CubeSat carries an off-the-shelf low-power Ferroelectric RAM (FRAM) chip and runs repeated read/write tests according to a predefined pattern. We compare every memory snapshot received from the satellite with a reference bitmap to identify bit errors. We visually analyze erroneous values distributions in the memory snapshots and we understand that they are frequently distributed according to regular spatial patterns. We define a generative deep learning technique to automatically learn these spatial patterns and to synthetically generate additional error patterns mimicking the observed ones. The model is capable at producing memory errors for general application-level experimentation. Our preliminary validation indicates that when comparing the model output with the data we collect, 0.92 of the generated bit flips correspond to bit flips in the memory snapshots and the model captures 0.91 of the observed bit flips. A comparison of the spatial distribution of bit flips in 10,000 generated error patterns also provides a Wasserstein distance of 0.073, indicating that the two distributions differ by roughly 7% of the maximum possible displacement of their probability masses.

Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese

Topics: Computer science, Artificial intelligence, Deep learning, Bitmap, Algorithm, CubeSat

Sources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange