Information-Driven Adaptive Sensing Based on Deep Reinforcement Learning

TL;DR

This paper introduces a Fisher information-based reward function for deep reinforcement learning, enabling resource-efficient sensing policies for IoT devices that adaptively balance information gain and energy consumption.

Contribution

It proposes a novel reward function for deep reinforcement learning that improves sensing efficiency in resource-constrained IoT environments.

Findings

Outperforms uniform sampling strategies in noise monitoring

Approaches near-optimal oracle performance

Demonstrates general applicability without extensive tuning

Abstract

In order to make better use of deep reinforcement learning in the creation of sensing policies for resource-constrained IoT devices, we present and study a novel reward function based on the Fisher information value. This reward function enables IoT sensor devices to learn to spend available energy on measurements at otherwise unpredictable moments, while conserving energy at times when measurements would provide little new information. This is a highly general approach, which allows for a wide range of use cases without significant human design effort or hyper-parameter tuning. We illustrate the approach in a scenario of workplace noise monitoring, where results show that the learned behavior outperforms a uniform sampling strategy and comes close to a near-optimal oracle solution.