Noise-assisted concurrent multipath traffic distribution in ad hoc networks

Area of Science:

• Computer Networking
• Network Control Systems
• Biologically Inspired Computing

Background:

• Traditional network control methods struggle with unpredictable conditions in wireless ad hoc networks.
• Existing solutions often require extensive knowledge of network components and pre-tuned parameters.
• Need for robust and adaptive control methods in dynamic network environments.

Purpose of the Study:

• To introduce a novel control concept, attractor perturbation (AP), for network management.
• To develop a multipath traffic distribution method based on AP.
• To reduce average end-to-end delay in computer networks using end-to-end information.

Main Methods:

• Utilized the attractor perturbation (AP) concept for network control.
• Proposed a concurrent multipath traffic distribution strategy.
• Adjusted transmission rates on individual paths based on AP principles.

Main Results:

• The proposed AP-based method effectively controls network performance using only end-to-end information.
• Demonstrated a lower average end-to-end delay compared to methods not accounting for fluctuations.
• Simulations confirmed the efficacy of the attractor perturbation relationship in improving network delay.

Conclusions:

• Attractor perturbation offers a viable approach for robust and adaptive network control.
• The concurrent multipath traffic distribution method successfully lowers network delay.
• This method provides an effective solution for dynamic network environments without requiring detailed component data.