Deniz Gunduz
Imperial College London
596 Papers
2.3K Citations
Deniz Gunduz is an academic researcher from Imperial College London. The author has contributed to research in topics: Computer science & Communication channel. The author has an hindex of 52, co-authored 505 publications. Previous affiliations of Deniz Gunduz include Princeton University & Norwegian University of Science and Technology.
Chat about Author
Papers
Distortion Minimization in Gaussian Layered Broadcast Coding with Successive Refinement
TL;DR: A transmitter without channel state information wishes to send a delay-limited Gaussian source over a slowly fading channel, and the power distribution that minimizes expected distortion converges to the one that maximizes expected capacity.
75
Management and Orchestration of Virtual Network Functions via Deep Reinforcement Learning
TL;DR: This work presents a novel RL approach, called parameterized action twin (PAT) deterministic policy gradient, which leverages an actor-critic architecture to learn to provision resources to the VNFs in an online manner, and presents numerical performance results, and map them to 5G key performance indicators (KPIs).
75
Blind Federated Edge Learning
TL;DR: An analog ‘over-the-air’ aggregation scheme, in which the devices transmit their local updates in an uncoded fashion, and the proposed algorithm becomes deterministic despite the lack of perfect CSI when the PS has a sufficiently large number of antennas.
74
Update Aware Device Scheduling for Federated Learning at the Wireless Edge
Mohammad Mohammadi Amiri,Deniz Gunduz,Sanjeev R. Kulkarni,H. Vincent Poor +3 more
- 21 Jun 2020
TL;DR: In this article, the authors study federated learning at the wireless edge, where power-limited devices with local datasets train a joint model with the help of a remote parameter server (PS).
71
•Posted Content
Energy-Aware Analog Aggregation for Federated Learning with Redundant Data
TL;DR: In this article, the authors proposed an online energy-aware dynamic worker scheduling policy, which maximizes the average number of workers scheduled for gradient update at each iteration under a long-term energy constraint.
70