arXiv subjects

Praveen Kumar Donta

Publications and source records attributed to Praveen Kumar Donta.

39 records · Page 3Linked to original sources

arXiv2023

Active Inference on the Edge: A Design Study

Machine Learning (ML) is a common tool to interpret and predict the behavior of distributed computing systems, e.g., to optimize the task distribution between devices. As more and more data is created by Internet of Things (IoT) devices, data processing and ML training are carried out by edge devices in close proximity. To ensure Quality of Service (QoS) throughout these operations, systems are supervised and dynamically adapted with the help of ML. However, as long as ML models are not retrained, they fail to capture gradual shifts in the variable distribution, leading to an inaccurate view of the system state. Moreover, as the prediction accuracy decreases, the reporting device should actively resolve uncertainties to improve the model's precision. Such a level of self-determination could be provided by Active Inference (ACI) -- a concept from neuroscience that describes how the brain constantly predicts and evaluates sensory information to decrease long-term surprise. We encompassed these concepts in a single action-perception cycle, which we implemented for distributed agents in a smart manufacturing use case. As a result, we showed how our ACI agent was able to quickly and traceably solve an optimization problem while fulfilling QoS requirements.

eess.SY↗

arXiv2023

Collaborative Inference in DNN-based Satellite Systems with Dynamic Task Streams

As a driving force in the advancement of intelligent in-orbit applications, DNN models have been gradually integrated into satellites, producing daily latency-constraint and computation-intensive tasks. However, the substantial computation capability of DNN models, coupled with the instability of the satellite-ground link, pose significant challenges, hindering timely completion of tasks. It becomes necessary to adapt to task stream changes when dealing with tasks requiring latency guarantees, such as dynamic observation tasks on the satellites. To this end, we consider a system model for a collaborative inference system with latency constraints, leveraging the multi-exit and model partition technology. To address this, we propose an algorithm, which is tailored to effectively address the trade-off between task completion and maintaining satisfactory task accuracy by dynamically choosing early-exit and partition points. Simulation evaluations show that our proposed algorithm significantly outperforms baseline algorithms across the task stream with strict latency constraints.

cs.DC↗

arXiv2021

EDGF: Empirical dataset generation framework for wireless network networks

In wireless sensor networks (WSNs), simulation practices, system models, algorithms, and protocols have been published worldwide based on the assumption of randomness. The applied statistics used for randomness in WSNs are broad in nature, e.g., random deployment, activity tracking, packet generation, etc. Even though with adequate formal and informal information provided and pledge by authors, validation of the proposal became a challenging issue. The minuscule information alteration in implementation and validation can reflect the enormous effect on eventual results. In this proposal, we show how the results are affected by the generalized assumption made on randomness. In sensor node deployment, ambiguity arises due to node error-value ($ε$), and it's upper bound in the relative position is estimated to understand the delicacy of diminutives changes. Moreover, the effect of uniformity in the traffic and contribution of scheduling position of nodes also generalized. We propose an algorithm to generate the unified dataset for the general and some specific applications system models in WSNs. The results produced by our algorithm reflects the pseudo-randomness and can efficiently regenerate through seed value for validation.

cs.NI↗