Olga Kondrateva

Dr. Olga Kondrateva is a researcher specializing in small satellite networks, with a focus on machine learning applications for data transmission optimization. Her work addresses critical challenges in low-Earth-orbit (LEO) satellite communication, including short contact times, bandwidth limitations, and high packet loss probabilities. Key contributions include novel techniques for joint source-and-channel coding, incremental neural network updates, and mathematical optimization frameworks like Benders decomposition for scalable satellite network planning.

• Education: PhD (implied by "Dr." prefix)

Research Interests
Olga’s research bridges machine learning and network optimization in constrained satellite environments. She develops methods to enhance reliability and efficiency of onboard neural networks, prioritizes model parameter updates, and designs communication protocols for Earth observation missions. Her work often leverages vector quantization and linear programming to address unique challenges like intermittent connectivity and dynamic mission requirements.

Publication Trends
Her recent publications (2022–2024) emphasize progressive neural network updates, fault tolerance, and throughput optimization in LEO satellite networks. Themes include bandwidth-efficient transmission, model compression, and incremental learning to adapt to changing mission needs.