Adaptive golf-optimized multi-layer perceptron framework for energy-efficient geographic routing and extended network...

Area of Science:

• Computer Science
• Network Engineering
• Artificial Intelligence

Background:

• Mobile ad-hoc networks (MANETs) face challenges in routing and network lifetime due to dynamic topology and node mobility.
• Geographic routing improves resource utilization but often lacks optimal performance and longevity.
• Existing methods struggle to balance routing efficiency with sustained network operation.

Purpose of the Study:

• To propose an improved multi-layer perceptron-based golf optimization algorithm for geographic routing (IMLP-GOA-GR) in MANETs.
• To enhance routing efficiency, network lifetime, and overall performance in MANETs.
• To reduce energy expenditure and end-to-end delay through optimized routing decisions.

Main Methods:

• Developed an improved multi-layer perceptron-based golf optimization algorithm (IMLP-GOA-GR).
• Utilized mobile node's local position for guiding routing, movement, and communication decisions.
• Conducted simulations across varying node densities to evaluate performance.

Main Results:

• Improved network lifetime by 18-25% compared to baseline methods.
• Reduced total energy consumption by 20-27%.
• Lowered average end-to-end delay by 44.44% and improved packet delivery ratio (PDR) by 37.5%.

Conclusions:

• The IMLP-GOA-GR framework significantly enhances routing efficiency in MANETs.
• The proposed method effectively extends network lifetime and reduces energy consumption.
• IMLP-GOA-GR demonstrates superior performance in terms of delay and packet delivery ratio.