Effectiveness of Data Augmentation for Localization in WSNs Using Deep Learning for the Internet of Things

Area of Science:

• Computer Science
• Electrical Engineering
• Networking

Background:

• Wireless Sensor Networks (WSNs) are crucial for IoT applications, but accurate node localization remains a challenge.
• The Dv-hop algorithm offers a simple, range-free localization method for WSNs, yet it has limitations in accuracy.
• Deep Learning (DL) shows promise for improving localization but requires substantial training data.

Purpose of the Study:

• To develop an accurate, Deep Learning (DL)-based range-free localization technique for WSNs in IoT environments.
• To address the data scarcity issue in DL for WSN localization.
• To improve the localization accuracy of the Dv-hop algorithm.

Main Methods:

• A Deep Neural Network (DNN) was employed to refine distance estimations between unknown and anchor nodes.
• A Data Augmentation Strategy (DAS) was proposed, creating virtual anchors to generate more training data.
• The enhanced Dv-hop algorithm integrated DNN correction for improved localization.

Main Results:

• The proposed Data Augmentation Strategy (DAS) effectively increases the training dataset size for DNNs.
• The integration of DNN correction significantly enhances the localization accuracy compared to the standard Dv-hop algorithm.
• The DL-based approach provides a feasible solution for low-cost, accurate localization in WSNs and IoT.

Conclusions:

• Deep Learning, augmented with a Data Augmentation Strategy, offers a powerful method to overcome the accuracy limitations of range-free localization algorithms like Dv-hop in WSNs.
• This approach makes advanced DL-aided localization more practical and cost-effective for Internet of Things (IoT) deployments.
• The developed technique demonstrates superior performance, paving the way for more precise location awareness in wireless sensor networks.