A Q-Learning-Based Delay-Aware Routing Algorithm to Extend the Lifetime of Underwater Sensor Networks

Area of Science:

• Computer Science
• Networking
• Marine Technology

Background:

• Underwater sensor networks (UWSNs) are crucial for aquatic applications.
• Limited battery life and difficult replacement of underwater sensor nodes necessitate extending network lifetime.
• The variable speed of sound in water presents routing challenges for UWSNs.

Purpose of the Study:

• To propose a Q-learning based delay-aware routing (QDAR) algorithm for UWSNs.
• To enhance the network lifetime and reduce end-to-end delay in UWSNs.
• To address the challenges of reliable routing in dynamic underwater environments.

Main Methods:

• Developed a Q-learning based delay-aware routing (QDAR) algorithm.
• Incorporated a data collection phase for dynamic environment adaptation.
• Defined an action-utility function considering residual energy and propagation delay for routing decisions.

Main Results:

• QDAR extends network lifetime through uniform residual energy distribution.
• QDAR achieves lower end-to-end data transmission delays.
• Simulations show QDAR yields comparable network lifetime and reduces end-to-end delay by 20-25% versus QELAR.

Conclusions:

• QDAR effectively balances energy distribution and minimizes delay in UWSNs.
• The Q-learning approach enables optimal next-hop selection for improved network performance.
• QDAR offers a promising solution for enhancing the efficiency and longevity of underwater sensor networks.