Vertical Handover Prediction Based on Hidden Markov Model in Heterogeneous VLC-WiFi System

Area of Science:

• Wireless communication networks
• Data science and machine learning

Background:

• Visible Light Communication (VLC) offers high bandwidth but suffers from Line-of-Sight (LoS) dependency, leading to transmission interruptions.
• Integrating VLC with Radio Frequency (RF) networks like WiFi enhances Quality of Experience (QoE) by ensuring continuous connectivity.
• Efficient Vertical Handover (VHO) is crucial for hybrid VLC-WiFi systems to manage short-duration VLC blockages and optimize user experience.

Purpose of the Study:

• To develop an automated VHO algorithm for hybrid VLC-WiFi systems.
• To optimize switching costs and dwelling time in heterogeneous networks.
• To enhance the reliability and seamlessness of wireless connectivity.

Main Methods:

• A Hidden Markov Model (HMM) is employed for VHO prediction.
• Principal Component Analysis (PCA) is used to extract effective Received Signal Strength (RSS) values from large datasets, reducing computational complexity.
• The HMM-VHO scheme utilizes channel characterization, specifically measured RSS at different locations.

Main Results:

• The proposed HMM-based VHO scheme demonstrates high prediction accuracy for the next Access Point (AP).
• It significantly reduces the Mixed Absolute Percentage Error (MAPE) compared to existing methods.
• The algorithm improves network dwell time and minimizes handover events, mitigating ping-pong effects.

Conclusions:

• The HMM-based VHO algorithm provides a robust solution for seamless connectivity in hybrid VLC-WiFi networks.
• It effectively addresses the limitations of VLC by leveraging WiFi integration and intelligent handover prediction.
• This approach enhances user experience by ensuring stable and efficient wireless communication.