Evaluating the Impact of Edge Computing on IoT Data Processing Efficiency

Thesis Abstract

The rapid proliferation of Internet of Things (IoT) devices has amplified the demand for efficient data processing architectures capable of managing vast volumes of real-time information. Traditional centralized cloud computing models often encounter limitations related to latency, network bandwidth, and energy consumption, thereby impeding the timely and reliable delivery of IoT services. Edge computing has emerged as a promising paradigm to address these challenges by decentralizing data processing closer to data sources, yet the empirical evaluation of its impact on IoT data processing efficiency remains limited. This study aims to evaluate the influence of edge computing deployment on the efficiency of IoT data processing systems, with specific objectives to quantify latency reductions, assess bandwidth utilization, and analyze energy consumption before and after the implementation of edge computing architectures within IoT ecosystems. The research adopts a mixed-methods approach, combining quantitative measurement and qualitative insights to provide a comprehensive evaluation. The quantitative component employs a quasi-experimental design involving a sample of 150 IoT devices across smart manufacturing settings, selected via stratified random sampling to ensure diversity in device types and data loads. Data collection instruments include network monitoring tools to record latency, bandwidth, and energy metrics over six months, supplemented by structured questionnaires administered to technicians and system administrators to capture perceptions of system performance and reliability. The qualitative facet entails thematic analysis of interview transcripts from 20 key informants involved in system deployment and management, guided by the Diffusion of Innovation theory and the Technology Acceptance Model (TAM) to interpret user acceptance and operational integration practices. Data analysis will utilize multiple regression analysis to quantify the relationship between the deployment of edge computing and data processing efficiency metrics, with ANOVA employed to compare system performance before and after edge architecture implementation. Additionally, time-series analysis will be applied to observe trends and variations across measurement periods, while thematic analysis will interpret qualitative data regarding user experiences and perceived benefits or challenges associated with edge computing integration. To ensure validity and reliability, triangulation of quantitative and qualitative findings will be performed, alongside pilot testing of data collection instruments and peer review of analytical procedures. Anticipated findings suggest that incorporating edge computing within IoT systems will significantly reduce data processing latency by an estimated 35%, improve bandwidth efficiency by approximately 20%, and decrease energy consumption by around 15%. It is expected that these improvements will be corroborated by positive perceptions among system users, indicating increased operational reliability and user satisfaction. The study aims to contribute to theoretical understanding by validating the applicability of the Diffusion of Innovation theory and TAM within IoT-Edge integration contexts, as well as filling empirical gaps concerning the quantifiable benefits of edge computing. This research provides actionable insights for practitioners, policymakers, and researchers by demonstrating measurable efficiencies attributable to edge computing adoption, thus informing design strategies and deployment policies for IoT infrastructures. The conclusion will emphasize the transformative potential of edge architectures for enhancing IoT operational performance and recommend best practices for scalable and secure integration of edge computing solutions. Future research suggestions involve longitudinal studies to examine long-term impacts, exploration of security implications, and expansion to diverse IoT sectors to generalize findings across different application domains.