Session WMLC-Opening

Opening Remarks

Conference
4:45 PM — 5:00 PM EEST
Local
Jun 15 Mon, 9:45 AM — 10:00 AM EDT

Opening Remarks

0
This talk does not have an abstract.

Session Chair

Enter Zoom
Session WMLC-Session-1

Session 1

Conference
5:00 PM — 6:00 PM EEST
Local
Jun 15 Mon, 10:00 AM — 11:00 AM EDT

Invited Paper: Meta-Scheduling for the Wireless Downlink through Learning with Bandit Feedback

Gustavo De Veciana (The University of Texas at Austin)

1
In this paper, we study learning-assisted multi-user scheduling for the wireless downlink. There have been many scheduling algorithms developed that optimize for a plethora of performance metrics; however a systematic approach across diverse performance metrics and deployment scenarios is still lacking. We address this by developing a meta-scheduler – given a diverse collection of schedulers, we develop a learning-based overlay algorithm (meta-scheduler) that selects that “best” scheduler from amongst these for each deployment scenario. More formally, we develop a multi-armed bandit (MAB) framework for meta-scheduling that assigns and adapts a score for each scheduler to maximize reward (e.g., mean delay, timely throughput etc.). The meta-scheduler is based on a variant of the Upper Confidence Bound algorithm (UCB), but adapted to interrupt the queuing dynamics at the base-station so as to filter out schedulers that might render the system unstable. We show that the algorithm has a poly-logarithmic regret in the expected reward with respect to a genie that chooses the optimal scheduler for each scenario. Finally through simulation, we show that the meta-scheduler learns the choice of the scheduler to best adapt to the deployment scenario (e.g. load conditions, performance metrics).

Invited Talk: Auto-Tuning and Capacity Region Learning for Downlink Scheduling in Cellular Systems

Sanjay Shakkottai (The University of Texas at Austin)

2
This talk does not have an abstract.

Session Chair

Srinivas Shakkottai

Enter Zoom
Session WMLC-Session-2

Session 2

Conference
6:30 PM — 8:00 PM EEST
Local
Jun 15 Mon, 11:30 AM — 1:00 PM EDT

Invited Talk: Physical Layer Communications via Deep Learning

Hyeji Kim (Samsung AI Research)

0
This talk does not have an abstract.

Invited Talk: Deploying AI enabled network control planes in commercial mobile networks

Sachin Katti (Stanford University)

0
This talk does not have an abstract.

Invited Paper: Designing an ML-Friendly Wireless Physical Layer for Low-Power IoT

Swarun Kumar (Carnegie Mellon University)

0
With the advent of low-power Internet of Things (IoT), there is an increase in interest for designing inference systems in the cloud that aggregate and perform machine learning tasks from the low-power sensor data. Yet, unlike traditional mobile devices, low-power clients are too battery constrained to transmit large amounts of data within short time spans, as needed for many complex inference models. In this paper, we present a vision for bridging the gap between the power-starved low-power clients and the data-starved inference engines in the cloud. We present a mechanism that takes into account the battery life of these clients and how it affects the traditional inference models.

Session Chair

Mohammad Alizadeh

Enter Zoom
Session WMLC-Session-3

Session 3

Conference
9:30 PM — 11:00 PM EEST
Local
Jun 15 Mon, 2:30 PM — 4:00 PM EDT

Invited Paper: Thompson-Sampling-Based Wireless Transmission for Panoramic Video Streaming

R. Srikant (University of Illinois at Urbana-Champaign)

0
Panoramic video streaming has received great at-tention recently due to its immersive experience. Different from traditional video streaming, it typically consumes 4 ∼ 6× larger bandwidth with the same resolution. Fortunately, users can only see a portion (roughly 20%) of 360° scenes at each time and thus it is sufficient to deliver such a portion, namely Field of View (FoV), if we can accurately predict user’s motion. In practice, we usually deliver a portion larger than FoV to tolerate inaccurate prediction. Intuitively, the larger the delivered portion, the higher the prediction accuracy. This however leads to a lower transmission success probability. The goal is to select an appropriate delivered portion to maximize system throughput, which can be formulated as a multi-armed bandit problem, where each arm represents the delivered portion. Different from traditional bandit problems with single feedback information, we have two-level feedback information (i.e., both prediction and transmission outcomes) after each decision on the selected portion. As such, we propose a Thompson Sampling algorithm based on two-level feedback information, and demonstrate its superior performance than its traditional counterpart via simulations.

Invited Talk: Keeping Up with the Line Rate: Fast Implementations of Algorithms Using Neural Networks

Balaji Prabhakar (Stanford University)

0
This talk does not have an abstract.

Invited Paper: Reinforcement Learning for Multi-Hop Scheduling and Routing of Real-Time Flows

Srinivas Shakkottai (Texas A&M University)

0
We consider the problem of serving real-time flows over a multi-hop wireless network. Each flow is composed of packets that have strict deadlines, and the goal is to maximize the weighted timely throughput of the system. Consistent with recent developments using mm-wave communications, we assume that the links are directional, but are lossy, and have unknown probabilities of successful packet transmission. An average link utilization budget (similar to a power constraint) constrains the system. We pose the problem in the form of a Constrained Markov Decision Process (CMDP) with an unknown transition kernel. We use a duality approach to decompose the problem into an inner unconstrained MDP with link usage costs, and an outer link-cost update step. For the inner MDP, we develop model-based reinforcement learning algorithms that sample links by sending packets to learn the link statistics. While the first algorithm type samples links at will at the beginning and constructs the model, the second type is an online approach that can only use packets from flows to sample links that they traverse. The approach to the outer problem follows gradient descent. We characterize the sample complexity (number of packets transmitted) to obtain near-optimal policies, to show that a basic online approach has a poorer sample complexity bound, it can be modified to obtain an online algorithm that has excellent empirical performance.

Session Chair

Sanjay Shakkottai

Enter Zoom
Session WMLC-Session-4

Session 4

Conference
11:30 PM — 12:30 AM EEST
Local
Jun 15 Mon, 4:30 PM — 5:30 PM EDT

Invited Talk: Machine Learning for Real-Time Interactive Video: Measurement and Deployment in an Operational Video Telephony System

Xinyu Zhang (The University of California at San Diego)

0
This talk does not have an abstract.

Invited Talk: Aligning Network Resource with User Attention

Junchen Jiang (The University of Chicago)

0
This talk does not have an abstract.

Session Chair

Romit Roy Choudhury

Enter Zoom
Session WMLC-Closing

Closing Remarks

Conference
11:45 PM — 12:40 AM EEST
Local
Jun 15 Mon, 4:45 PM — 5:40 PM EDT

Closing Remarks

0
This talk does not have an abstract.

Session Chair

Enter Zoom

Made with in Toronto · Privacy Policy · © 2020 Duetone Corp.