S
Sandeep Chinchali
Researcher at Stanford University
Publications - 47
Citations - 699
Sandeep Chinchali is an academic researcher from Stanford University. The author has contributed to research in topics: Computer science & Engineering. The author has an hindex of 11, co-authored 27 publications receiving 467 citations. Previous affiliations of Sandeep Chinchali include University of North Carolina at Chapel Hill & California Institute of Technology.
Papers
More filters
Proceedings Article
Cellular Network Traffic Scheduling With Deep Reinforcement Learning
Sandeep Chinchali,Pan Hu,Tianshu Chu,Manu Sharma,Manu Bansal,Rakesh Misra,Marco Pavone,Sachin Katti +7 more
TL;DR: This work presents a reinforcement learning (RL) based scheduler that can dynamically adapt to traffic variation, and to various reward functions set by network operators, to optimally schedule IoT traffic and can enable mobile networks to carry 14.7% more data with minimal impact on existing traffic.
NUMFabric: Fast and Flexible Bandwidth Allocation in Datacenters
Kanthi Nagaraj,Sandeep Chinchali,Dinesh Bharadia,Hongzi Mao,Mohammadreza Alizadeh Attar,Sachin Katti +5 more
TL;DR: In this paper, the authors present xFabric, a datacenter transport design that provides flexible and fast bandwidth allocation control, which enables operators to specify how bandwidth is allocated among contending flows to optimize for different service level objectives such as minimizing flow completion times, weighted allocations, different notions of fairness, etc.
Proceedings ArticleDOI
NUMFabric: Fast and Flexible Bandwidth Allocation in Datacenters
TL;DR: In this article, the authors present xFabric, a datacenter transport design that provides flexible and fast bandwidth allocation control, which enables operators to specify how bandwidth is allocated among contending flows to optimize for different service level objectives such as minimizing flow completion times, weighted allocations, different notions of fairness, etc.
Proceedings ArticleDOI
Neural Networks Meet Physical Networks: Distributed Inference Between Edge Devices and the Cloud
TL;DR: This work proposes a distributed DNN architecture that learns end-to-end how to represent the raw sensor data and send it over the network such that it meets the eventual sensing task's needs.
Journal ArticleDOI
Network offloading policies for cloud robotics: a learning-based approach
Sandeep Chinchali,Apoorva Sharma,James Harrison,Amine Elhafsi,Daniel Kang,Evgenya Pergament,Eyal Cidon,Sachin Katti,Marco Pavone +8 more
TL;DR: This paper formulates offloading as a sequential decision making problem for robots, and proposes a solution using deep reinforcement learning, which improves vision task performance and allows robots the potential to significantly transcend their on-board sensing accuracy but with limited cost of cloud communication.