Research Publications

Now showing 1 - 3 of 3

Embargo
Intelligent Adaptive Lighting Control: Reinforcement Learning-Based Optimization for Smart Home Energy Efficiency
(Institute of Electrical and Electronics Engineers Inc., 2025) Hewakapuge M.M; Gamage W.G.T; Surendra D.M.B.G.D; Thejan K.G.T; Rajapaksha, S; Rajendran, K
This study introduces a novel research paper outlining a behavioral-based adaptive lighting system that aims to revolutionise smart home lighting by integrating user behavior tracking to enhance energy efficiency and user comfort. Unlike traditional motion-sensor-based lighting, the novelty of this approach is the ability to adapt dynamically to evolving user behaviors through reinforcement learning. The system utilises Wi-Fi-based positioning, GPS and accelerometer data to monitor user movements and classify different areas of the house. Users initially calibrate the home layout through a mobile application, marking room locations and lighting configurations. The system then collects movement data over time to predict optimal lighting schedules based on user routines and refines the predictions and updates lighting adjustments accordingly, minimising energy wastage while maximising user convenience. A serverless backend architecture ensures scalability, cost-effectiveness, and seamless data processing. The adaptive framework continuously refines lighting automation, responding to evolving behavioral patterns.
Embargo
'xīnl' The Social Media App to Replenish Mental Health with the Aid of an Egocentric Network
(Institute of Electrical and Electronics Engineers, 2022-11-03) Kasthurirathna, D; Kalansooriya, S; Kaluarachchi, A; Weerawickrama, C; Nanayakkara, D; Adeepa, D
The impact of social groups on one's emotional health is a crucial issue that must be addressed correctly. Emotions and social groups play significant roles in human mental and physical activities. It is difficult to detect and maintain track of changing emotional states. The main goal of this study is to build a social media app called Xinli, that proposes an aggregated method to predict emotions using a multimodal approach and to predict personalized activities based on the user's mental state, and to further track the improvement of emotional state with the impact of recommended activities and social support groups. The results suggest that the aggregated modalities method is more accurate in recognizing emotions, and activity prediction using reinforcement learning is a clean way to personalize activities based on the emotional state from user to user, which is the novelty of the proposed study.
Embargo
Improving trusted routing by identifying malicious nodes in a MANET using reinforcement learning
(IEEE, 2017-09-06) Mayadunna, H; De Silva, S. L; Wedage, L; Pabasara, S; Rupasinghe, L; Liyanapathirana, C; Kesavan, K; Nawarathna, C; Sampath, K. K
Mobile ad-hoc networks (MANETs) are decentralized and self-organizing communication systems. They have become pervasive in the current technological framework. MANETs have become a vital solution to the services that need flexible establishments, dynamic and wireless connections such as military operations, healthcare systems, vehicular networks, mobile conferences, etc. Hence it is more important to estimate the trustworthiness of moving devices. In this research, we have proposed a model to improve a trusted routing in mobile ad-hoc networks by identifying malicious nodes. The proposed system uses Reinforcement Learning (RL) agent that learns to detect malicious nodes. The work focuses on a MANET with Ad-hoc On-demand Distance Vector (AODV) Protocol. Most of the systems were developed with the assumption of a small network with limited number of neighbours. But with the introduction of reinforcement learning concepts this work tries to minimize those limitations. The main objective of the research is to introduce a new model which has the capability to detect malicious nodes that decrease the performance of a MANET significantly. The malicious behaviour is simulated with black holes that move randomly across the network. After identifying the technology stack and concepts of RL, system design was designed and the implementation was carried out. Then tests were performed and defects and further improvements were identified. The research deliverables concluded that the proposed model arranges for highly accurate and reliable trust improvement by detecting malicious nodes in a dynamic MANET environment.

Research Publications

Browse

Filters

Advanced Search

Filter by

Settings

Sort By

Results per page

Search Results