Video Experimental Relacionado
Updated: Feb 9, 2026
03:31
Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
Published on: December 15, 2023
1.1K
Red de Q profunda personalizada adaptativa e inteligente para la descarga de tareas energéticamente eficientes en
1School of Electronics Engineering, Vellore Institute of Technology, Vellore, 632014, Tamil Nadu, India.
Scientific reports
|February 7, 2026
Resumen
Un nuevo marco de IA, AICDQN, optimiza la descarga de tareas en sistemas de borde-nube. Reduce el retraso y las caídas de tareas al tiempo que mejora la eficiencia energética para aplicaciones IoT sensibles a la latencia.
Área de la Ciencia:
- Ciencias de la Computación
- Inteligencia Artificial
- Sistemas Distribuidos
Sus antecedentes:
- La computación de borde-nube se está expandiendo, aumentando las demandas de descarga de tareas eficiente.
- Las aplicaciones de Internet de las cosas (IoT) sensibles a la latencia requieren una programación inteligente en entornos dinámicos.
Objetivo del estudio:
- Introducir un marco novedoso de aprendizaje por refuerzo, Adaptive and Intelligent Customized Deep Q-Network (AICDQN), para la programación de tareas prioritaria.
- Mejorar la toma de decisiones en tiempo real en sistemas de computación de borde móvil.
Principales métodos:
- Se formuló la descarga de tareas como un Proceso de Decisión de Markov (MDP).
- Se integró una Unidad Recurrente Gated-Long Short-Term Memory (GRU-LSTM) híbrida para la predicción de la carga de trabajo.
- Se empleó un agente Dynamic Dueling Double Deep Q-Network para las decisiones de descarga en los niveles local, de borde y de nube.
- Se modelaron los nodos de cómputo utilizando sistemas de colas con prioridad (M/M/1, M/M/c, M/M/∞).
- Se implementó una función de puntuación de prioridad dinámica y una política de programación consciente de la energía.
Principales resultados:
- AICDQN logró hasta un 33,39% de reducción en el retraso.
- Demostró una mejora del 57,74% en la eficiencia energética.
- Redujo la tasa de caída de tareas en un 81,25% en comparación con los algoritmos existentes.
- Superó a Deep Deterministic Policy Gradient (DDPG), Distributed Dynamic Task Offloading (DDTO-DRL), Potential Game based Offloading Algorithm (PGOA) y User-Level Online Offloading Framework (ULOOF).
Conclusiones:
- AICDQN proporciona una solución escalable y adaptable para la descarga de tareas de borde-nube.
- El marco maneja de manera efectiva la programación en tiempo real, prioritaria y con restricciones de energía.
- Se validó la eficacia de la predicción híbrida GRU-LSTM y el agente Dueling Double DQN.
Videos de Conceptos Relacionados
02:26Protein Networks
4.6K
An organism can have thousands of different proteins, and these proteins must cooperate to ensure the health of an organism. Proteins bind to other proteins and form complexes to carry out their functions. Many proteins interact with multiple other proteins creating a complex network of protein interactions.
These interactions can be represented through maps depicting protein-protein interaction networks, represented as nodes and edges. Nodes are circles that are representative of a protein,...
These interactions can be represented through maps depicting protein-protein interaction networks, represented as nodes and edges. Nodes are circles that are representative of a protein,...
4.6K
01:27Intelligence
8.7K
The term "intelligence" is complex because it refers to both behavior and individuals, and its interpretation varies across cultures. European Americans tend to link intelligence with reasoning and cognitive skills, while in Kenya, it is tied to responsible participation in family and social life. In Uganda, intelligence is seen as the ability to know the right actions and carry them out effectively, while the Iatmul people of Papua New Guinea associate it with the capacity to remember...
8.7K
01:26Activation Energy
86.8K
Activation energy is the minimum amount of energy necessary for a chemical reaction to move forward. The higher the activation energy, the slower the rate of the reaction. However, adding heat to the reaction will increase the rate, since it causes molecules to move faster and increase the likelihood that molecules will collide. The collision and breaking of bonds represents the uphill phase of a reaction and generates the transition state. The transition state is an unstable high-energy state...
86.8K
04:10What is Energy?
59.2K
The universe is composed of matter in different forms, and all forms of matter contain energy. The different forms of energy on Earth originate from the Sun — the ultimate energy source. Plants capture light energy from the Sun, and, via the process of photosynthesis, convert it into chemical energy. This stored energy from plants can be harnessed in many ways. For example, eating plant products as food provides energy for our body to function, and burning wood or coal (fossilized...
59.2K
01:21Free Energy
52.2K
Free energy—abbreviated as G for the scientist Gibbs who discovered it—is a measurement of useful energy that can be extracted from a reaction to do work. It is the energy in a chemical reaction that is available after entropy is accounted for. Reactions that take in energy are considered endergonic and reactions that release energy are exergonic. Plants carry out endergonic reactions by taking in sunlight and carbon dioxide to produce glucose and oxygen. Animals, in turn, break...
52.2K
02:18Network Covalent Solids
16.2K
Network covalent solids contain a three-dimensional network of covalently bonded atoms as found in the crystal structures of nonmetals like diamond, graphite, silicon, and some covalent compounds, such as silicon dioxide (sand) and silicon carbide (carborundum, the abrasive on sandpaper). Many minerals have networks of covalent bonds.
To break or to melt a covalent network solid, covalent bonds must be broken. Because covalent bonds are relatively strong, covalent network solids are typically...
To break or to melt a covalent network solid, covalent bonds must be broken. Because covalent bonds are relatively strong, covalent network solids are typically...
16.2K