Timely Execution on Intermittently Powered Batteryless Sensors

Hester, Josiah; Storer, Kevin; Sorber, Jacob

doi:10.1145/3131672.3131673

Cited by 144 publications

(128 citation statements)

References 25 publications

Supporting

Mentioning

128

Contrasting

Order By: Relevance

“…We survey these works in this section. [105] Task-based checkpointing-versioning, × × manual idempotency analysis Ratchet [106] Task-based checkpointing, × × automatic idempotency analysis Chain [107] Idempotency task programming, × channel-based data exchange HarvOS [108] CFG-based checkpoint placement, × × × advanced ADC interrupt Clank [109] Dynamic idempotency task decomposition, × checkpointing and versioning Alpaca [110] Idempotency task programming, × privatization data exchange Mayfly [111] Idempotency task programming, remanence timekeeping A. Checkpointing Optimizations 1) Checkpoint placement and activation: refers to the scheme of inserting potential checkpoints to the program at compile-time and activate checkpointing processes at run-time. Inserting checkpoints at different locations of the program leads to distinct checkpointing overhead (including energy and memory) as the number of variables to be saved varies.…”

Section: Computing Optimizations For Batterylessmentioning

confidence: 99%

Sensing, Computing, and Communications for Energy Harvesting IoTs: A Survey

Lan

Hassan

et al. 2020

IEEE Commun. Surv. Tutorials

256

View full text Add to dashboard Cite

Sensing Optimizations Computing Optimizations Communications Optimizations  Commercial products and services  Standards for wireless protocol and transducer test method  Checkpointing optimization  Timekeeping across power failures  Packet-less communication  Reinforcement learning based communication optimization  Backscatter communication  Kinetic EH sensing  Thermal EH sensing  Solar EH sensing  RF EH sensingTABLE I: Vendors specializing in energy harvesting components and solutions suitable for IoTs. Company Source Energy Harvester EH Solution Application Foundation Year Piezo Systems Kinetic × piezoelectric energy harvesting 1988 MIDE Technology Kinetic × piezoelectric energy harvesting

show abstract

Section: Computing Optimizations For Batterylessmentioning

confidence: 99%

Sensing, Computing, and Communications for Energy Harvesting IoTs: A Survey

Lan

Hassan

et al. 2020

IEEE Commun. Surv. Tutorials

256

View full text Add to dashboard Cite

show abstract

“…If we employ existing intermittent computing frameworks like MayFly [35] to execute machine learning tasks, the system would blindly use every incoming training example to update the model parameters and thus drain the harvested energy much faster than needed. Although it considers the staleness of data to increase the system lifetime, it does not help a learner as the data can be fresh, yet their utility toward an application's high-level goal can be null.…”

Section: The Scope Of Intermittent Learningmentioning

confidence: 99%

“…• Action-based Programming. Similar to the task-based intermittent computing platforms [14,15,35,56,57,87], an action in the proposed intermittent learning framework is a user-defined block of code. An action, given sufficient energy to execute to completion, is guaranteed to have memory-consistency and control-flow that can be equivalently achieved with a continuously-powered execution.…”

Section: Intermittent Learning Programming Modelmentioning

confidence: 99%

“…Programs that run on these platforms follow the so-called intermittent computing paradigm [55,57,81,84] where a system pauses and resumes its code execution based on the availability of harvested energy. Over the past decade, the efficiency of batteryless computing platforms has been improved by reducing their energy waste through hardware provisioning, through check-pointing [71] to avoid restarting code execution from the beginning at each powerup [4], and through discarding stale sensor data [35] which are no longer useful. Despite these advancements, the capability of batteryless computing platforms has remained limited to simple sensing applications only.…”

Section: Introductionmentioning

confidence: 99%

“…We provide a programming model that allows a programmer to develop an intermittent learning application that executes correctly on an intermittent system. Like many existing proposals, we adopt a task-based -which we call action-based -intermittent programming model [14,15,35,56,57,87]. We provide application programmers with an energy pre-inspection tool that helps them split an existing application code into sub-modules called actions that atomically run to completion on intermittently-powered systems.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Intermittent Learning

Lee

Islam

Luo

et al. 2019

Proc. ACM Interact. Mob. Wearable Ubiquitous Technol.

View full text Add to dashboard Cite

This paper introduces intermittent learning -the goal of which is to enable energy harvested computing platforms capable of executing certain classes of machine learning tasks effectively and efficiently. We identify unique challenges to intermittent learning relating to the data and application semantics of machine learning tasks, and to address these challenges, we devise 1) an algorithm that determines a sequence of actions to achieve the desired learning objective under tight energy constraints, and 2) propose three heuristics that help an intermittent learner decide whether to learn or discard training examples at run-time which increases the energy efficiency of the system. We implement and evaluate three intermittent learning applications that learn the 1) air quality, 2) human presence, and 3) vibration using solar, RF, and kinetic energy harvesters, respectively. We demonstrate that the proposed framework improves the energy efficiency of a learner by up to 100% and cuts down the number of learning examples by up to 50% when compared to state-of-the-art intermittent computing systems that do not implement the proposed intermittent learning framework.

show abstract

Integrating cloud and mist computing to lower latency in IoT topologies

Herrero

2023

Trans Emerging Tel Tech

View full text Add to dashboard Cite

Traditional IoT topologies involve access and core networks that share a common edge. On this edge, border routers and gateways are responsible for converting protocols at different layers of the stack. Devices like sensors and actuators sit on the access network while applications are located on the core network. The application performs predictions that trigger actuation based on received sensor readouts. Prediction, in turn, is the result of machine learning (ML) algorithms that are typically executed on the cloud. An alternative to this approach consists of performing the prediction on constrained devices on the IoT access network. This leads to Tiny ML (TinyML) and mist computing. In this context, there is a trade‐off between latency and computational power that becomes a deciding factor when choosing the application to carry on predictions. This paper introduces an algorithm that can be used to dynamically select the right application based on network layer parameters.

show abstract

Timely Execution on Intermittently Powered Batteryless Sensors

Cited by 144 publications

References 25 publications

Sensing, Computing, and Communications for Energy Harvesting IoTs: A Survey

Sensing, Computing, and Communications for Energy Harvesting IoTs: A Survey

Intermittent Learning

Integrating cloud and mist computing to lower latency in IoT topologies

Contact Info

Product

Resources

About