Rajkummar Buyya

FedAD: An Adaptive Block Dropping Approach for Federated Learning on Resource-Constrained Devices

—In a federated learning environment with heterogeneous, limited and time-varying computational resource capabilities, it is challenging to update deep neural network in real-time. To address this challenge, we propose FedAD, an adaptive block-dropping approach for federated learning on resource-constrained devices. Fe...

Future Generation Computer Systems

Future Generation Computer Systems

Darpan: A Digital Twin Framework for the Next-Generation Computing Continuum

Computing-continuum applications distribute work across devices, edge systems, fog resources, and clouds. While a placement, scheduling, or recovery decision is being made, resource availability, network conditions, and application progress may change, so the decision can be invalid by the time it is executed. Existing...

Kafila: Serving Large Language Models on a Trusted Set of Heterogeneous Commodity Machines

Kafila is proposed, whose protocol assembles a ring from behind NATs, preferring direct paths and relaying where traversal fails, while its planner measures each device's memory bandwidth, capacity and reachability, divides the model exactly for a fixed ring order, and places the head together with that division rather...

Stability-Aware Proactive Autoscaling Using a Double Deep Q-Network in Cloud Computing Environments

A double deep Q-Network-based proactive autoscaling approach (DDQN-Proactive) along with Resource Removal Strategy (RRS) that enhances decision-making by decoupling action selection from value evaluation, enabling more stable and adaptive scaling.

Rollout Efficiency in Reinforcement Learning for Reasoning Large Language Models: A Taxonomy and Future Directions

This survey provides a systematic taxonomy of recent research on rollout efficiency for reasoning-oriented reinforcement learning, classifying existing approaches from both mechanism and bottleneck perspectives.

CELLServe: An SLO-Aware and Cost Efficient LLMs Serving System for Serverless Computing Environments

CELLServe formalizes SLO-constrained joint resource provisioning as an optimization problem with a dedicated algorithm, and introduces an opportunistic instance merging strategy for decode phase functions to reclaim fragmented resources.

AgentR A Stateful and Recovery-Aware Software Architecture for LLM-based Auditable Workflows

A stateful architecture for LLM workflow systems that enables improved observability, recoverability, and operational accountability in LLM workflow systems, and preliminary proof-of-concept that persistent state machine design, asynchronous orchestration, and cost-aware usage logging can enable improved observability,...

CLASP: Chained-Request-Aware Scaling and Operator Placement for Serverless Stream Processing

CLASP, a scaling and scheduling strategy for stream processing in stateful serverless environments, which improves throughput by up to 3.3x and reduces median end-to-end latency by up to 76% compared with state-of-the-art scaling strategies.

Optimizing resource allocation for federated LLM training via workload forecasting and autoscaling in edge environments

An integrated framework that combines LLM fine-tuning using federated adaptive local low-rank adaptation (FedALoRA) with forecasting-driven autoscaling policies is proposed, making it well-suited for federated-LLM Autoscaling in dynamic edge environments.

Trust-Aware Topology Learning for Dynamic Decentralized Federated Learning under Adversaries

DMTT (Dynamic MURMURA with Trusted Topology), a decentralized personalized FL protocol built on MURMURA, which uses evidential deep learning to down-weight distribution-mismatched peers, extended here to time-varying graphs under topology-manipulation attacks.