Intelligent resource allocation in internet of things using random forest and clustering techniques

Area of Science:

• Computer Science
• Electrical Engineering
• Network Engineering

Background:

• The proliferation of Internet of Things (IoT) devices presents significant resource management challenges.
• Constraints like limited energy, bandwidth, and computation power necessitate efficient resource allocation strategies.
• Existing methods (e.g., evolutionary algorithms, multi-agent reinforcement learning) struggle with the dynamic nature of IoT networks due to complexity and cost.

Purpose of the Study:

• To propose an intelligent resource allocation approach for Internet of Things (IoT) networks.
• To address the inefficiencies of current methods in dynamic and scalable IoT environments.
• To improve prediction accuracy, reduce energy consumption, and decrease response times.

Main Methods:

• Integration of clustering (K-Means) and machine learning (Random Forest) techniques.
• Clustering IoT devices based on energy consumption and bandwidth requirements.
• Training a Random Forest model to predict resource needs for optimal allocation.

Main Results:

• Achieved a prediction accuracy of 94% for resource needs.
• Reduced energy consumption by 20% compared to existing methods.
• Decreased response time by 10% in dynamic IoT environments.

Conclusions:

• The proposed approach effectively manages resources in dynamic and scalable IoT networks.
• Combining K-Means clustering and Random Forest machine learning offers a superior solution for IoT resource allocation.
• Demonstrated significant improvements in accuracy, energy efficiency, and response time.