arXiv:2510.21366v1 Announce Type: cross 
Abstract: In this work, we propose a novel framework to enable diffusion models to adapt their generation quality based on real-time network bandwidth constraints. Traditional diffusion models produce high-fidelity images by performing a fixed number of denoising steps, regardless of downstream transmission limitations. However, in practical cloud-to-device scenarios, limited bandwidth often necessitates heavy compression, leading to loss of fine textures and wasted computation. To address this, we introduce a joint end-to-end training strategy where the diffusion model is conditioned on a target quality level derived from the available bandwidth. During training, the model learns to adaptively modulate the denoising process, enabling early-stop sampling that maintains perceptual quality appropriate to the target transmission condition. Our method requires minimal architectural changes and leverages a lightweight quality embedding to guide the denoising trajectory. Experimental results demonstrate that our approach significantly improves the visual fidelity of bandwidth-adapted generations compared to naive early-stopping, offering a promising solution for efficient image delivery in bandwidth-constrained environments. Code is available at: https://github.com/xzhang9308/BADiff.

إطار BADiff يقدم نهجًا مبتكرًا لنماذج الانتشار، مما يسمح لها بضبط جودة توليد الصور بناءً على عرض النطاق الترددي للشبكة في الوقت الفعلي. هذا مهم لأن النماذج التقليدية غالبًا ما تواجه قيودًا في عرض النطاق الترددي، مما يؤدي إلى صور مضغوطة تفقد تفاصيل مهمة. من خلال التكيف مع هذه القيود، يعد BADiff بتحسين جودة الصور المرسلة عبر الشبكات، مما يجعله تقدمًا قيمًا للتطبيقات في الحوسبة السحابية ومعالجة الصور في الوقت الفعلي.

El marco BADiff presenta un enfoque innovador para los modelos de difusión, permitiéndoles ajustar la calidad de generación de imágenes según el ancho de banda de red en tiempo real. Esto es significativo porque los modelos tradicionales a menudo enfrentan limitaciones de ancho de banda, lo que resulta en imágenes comprimidas que pierden detalles importantes. Al adaptarse a estas restricciones, BADiff promete mejorar la calidad de las imágenes transmitidas a través de redes, convirtiéndose en un avance valioso para aplicaciones en computación en la nube y procesamiento de imágenes en tiempo real.

Le cadre BADiff propose une approche innovante pour les modèles de diffusion, leur permettant d'ajuster la qualité de génération d'images en fonction de la bande passante réseau en temps réel. Cela est important car les modèles traditionnels ont souvent du mal avec les limitations de bande passante, entraînant des images compressées qui perdent des détails importants. En s'adaptant à ces contraintes, BADiff promet d'améliorer la qualité des images transmises sur les réseaux, ce qui en fait une avancée précieuse pour les applications en cloud computing et le traitement d'images en temps réel.

The BADiff framework introduces an innovative approach to diffusion models, allowing them to adjust their image generation quality based on real-time network bandwidth. This is significant because traditional models often struggle with bandwidth limitations, resulting in compressed images that lose important details. By adapting to these constraints, BADiff promises to enhance the quality of images transmitted over networks, making it a valuable advancement for applications in cloud computing and real-time image processing.

BADiff: Bandwidth Adaptive Diffusion Model

arXiv:2511.19103v1 Announce Type: new 
Abstract: The rapid growth of IoT devices has led to an enormous amount of sensor data that requires transmission to cloud servers for processing, resulting in excessive network congestion, increased latency and high energy consumption. This is particularly problematic in resource-constrained and remote environments where bandwidth is limited, and battery-dependent devices further emphasize the problem. Moreover, in domains such as agriculture, consecutive sensor readings often have minimal variation, making continuous data transmission inefficient and unnecessarily resource intensive. To overcome these challenges, we propose an analytical prediction algorithm designed for edge computing environments and validated through simulation. The proposed solution utilizes a predictive filter at the network edge that forecasts the next sensor data point and triggers data transmission only when the deviation from the predicted value exceeds a predefined tolerance. A complementary cloud-based model ensures data integrity and overall system consistency. This dual-model strategy effectively reduces communication overhead and demonstrates potential for improving energy efficiency by minimizing redundant transmissions. In addition to reducing communication load, our approach leverages both in situ and satellite observations from the same locations to enhance model robustness. It also supports cross-site generalization, enabling models trained in one region to be effectively deployed elsewhere without retraining. This makes our solution highly scalable, energy-aware, and well-suited for optimizing sensor data transmission in remote and bandwidth-constrained IoT environments.

تم اقتراح خوارزمية جديدة للتنبؤ التحليلي لبيئات الحوسبة الطرفية في الزراعة، تهدف إلى تقليل النقل المفرط لبيانات المستشعرات الناتجة عن أجهزة إنترنت الأشياء. تستخدم هذه الطريقة فلترًا تنبؤيًا لتوقع قراءات المستشعر، مما يؤدي إلى تفعيل نقل البيانات فقط عند حدوث انحرافات كبيرة، وبالتالي معالجة مشكلات ازدحام الشبكة واستهلاك الطاقة في البيئات ذات الموارد المحدودة.

Se ha propuesto un nuevo algoritmo de predicción analítica para entornos de computación en el borde en la agricultura, con el objetivo de reducir la transmisión excesiva de datos de sensores generados por dispositivos IoT. Este enfoque utiliza un filtro predictivo para prever las lecturas de los sensores, activando la transmisión de datos solo cuando ocurren desviaciones significativas, abordando así problemas de congestión de red y consumo de energía en entornos con recursos limitados.

Un nouvel algorithme de prédiction analytique a été proposé pour les environnements de calcul en périphérie dans l'agriculture, visant à réduire la transmission excessive des données des capteurs générées par les dispositifs IoT. Cette approche utilise un filtre prédictif pour prévoir les lectures des capteurs, déclenchant la transmission des données uniquement lorsque des écarts significatifs se produisent, abordant ainsi les problèmes de congestion du réseau et de consommation d'énergie dans des environnements à ressources limitées.

A new analytical prediction algorithm has been proposed for edge computing environments in agriculture, aimed at reducing the excessive transmission of sensor data generated by IoT devices. This approach utilizes a predictive filter to forecast sensor readings, triggering data transmission only when significant deviations occur, thereby addressing issues of network congestion and energy consumption in resource-constrained settings.

BADiff: Bandwidth Adaptive Diffusion Model

Was this article worth reading? Share it

CompressX

Https

Red5 Cloud