EasyLB: Adaptive Load Balancing Based on Flowlet Switching for Wireless Sensor Networks.

Area of Science:

• Computer Science
• Network Engineering
• Wireless Communication

Background:

• Wireless Sensor Networks (WSNs) face network congestion and throughput issues.
• Traditional load balancing in WSNs is complex due to fluctuating wireless channels and the need for global/local congestion information.
• Existing schemes struggle with adaptive load balancing in dynamic wireless environments.

Purpose of the Study:

• To design an adaptive load balancing scheme for WSNs using flowlet switching.
• To address the challenge of determining an optimal flowlet timeout value (δ) for effective load balancing.
• To provide a theoretical framework for setting flowlet timeout values considering non-equal path probabilities and parallel paths.

Main Methods:

• Extended the OpenFlow protocol to develop a flowlet switching based load balancing scheme (EasyLB).
• Formulated the flowlet timeout setting problem using the stationary distribution of a Markov chain.
• Incorporated non-equal probability path selection and multiple parallel load balancing paths into the timeout setting analysis.

Main Results:

• A theoretical reference for setting the flowlet timeout value (δ) was established.
• EasyLB demonstrated adaptability to wireless channel condition changes.
• The proposed timeout setting method achieved fast convergence of load balancing after link failures.

Conclusions:

• The EasyLB scheme effectively improves load balancing in WSNs through flowlet switching.
• The Markov chain-based theoretical reference provides an appropriate method for setting flowlet timeout values.
• EasyLB offers adaptive and resilient load balancing, outperforming traditional methods in dynamic wireless environments.