Innovative Spectrum Handoff Process Using a Machine Learning-Based Metaheuristic Algorithm

Area of Science:

• Computer Science
• Electrical Engineering
• Telecommunications

Background:

• Cognitive Radio Networks (CRNs) address spectrum scarcity by enabling dynamic spectrum access.
• Spectrum handoff (SHO) is crucial for CRNs but causes communication delays and power consumption.
• Reducing SHO events is vital for efficient CRN operation and secondary user (SU) connectivity.

Purpose of the Study:

• To propose a novel metaheuristic algorithm for efficient spectrum handoff management in CRNs.
• To minimize spectrum handoff occurrences and associated delays for improved network performance.
• To enhance the utilization of unallocated spectrum for secondary users.

Main Methods:

• A machine learning-based metaheuristic algorithm, the Support Vector Machine-Red Deer Algorithm (SVM-RDA), is proposed.
• Dynamic Spectrum Access (DSA) is employed to identify available channels during handoff.
• The algorithm's performance is evaluated through simulations measuring handoffs, delay, throughput, and SNR.

Main Results:

• The SVM-RDA algorithm demonstrates resilience and low complexity in simulation.
• The proposed method effectively predicts handoff delay and significantly reduces the number of handoffs.
• Experimental results show improved system performance, including higher throughput and better SNR.

Conclusions:

• The SVM-RDA offers an effective solution for spectrum handoff management in CRNs.
• The algorithm enhances system performance by minimizing handoffs and predicting delays.
• This approach contributes to more reliable and efficient spectrum utilization for secondary users.