Enhancing service availability and resource deployment in IoT using a shared service replication method

Abstract

In many real-world contexts, the Internet of Things (IoT) is valued for its capacity to facilitate the smooth operation of interoperable applications and services. It is critical to ensure the accessibility and replication of IoT resources to improve the agility of these applications. As a solution, the Network Function Virtualization (NFV) paradigm is embedded into the IoT design to leverage information from various endpoint applications better and maximize resource utilization. In this study, the Shared Replication Augmenting Method (SRAM) is proposed to increase resource usage in underutilized NFVs and maintain service availability simultaneously. The regressive decision-making learning used by SRAM enables the detection of NFV needs for data and application portability across various real-time use cases. This regression method can uncover data needs and their causes, allowing for prompt answers and more efficient use of available resources. The suggested SRAM technique dynamically modifies the procedure while considering computation-less function allocations, making it suitable for various interoperable applications. It distributes root-to-service virtualization and availability based on historical use and data replication. Therefore, SRAM improves resource usage by 7.09 % with no increase in latency or delays. It also increases service availability by 10.4 %, reduces latency by 11.89 %, eliminates backlogs by 11.1 %, and reduces data repetition by 8.97 %. This study enhances resource consumption and productivity in IoT settings by offering SRAM as a viable solution. The study's results prove its potential to reduce the occurrence of replication, delay, and queues while raising the availability of services.

Keywords

• Data availability
• IoT
• NFV
• Regression learning
• Resource utilization
• Service availability