Scholarlink

<p>Nowadays, Wireless Sensor Networks (WSNs) play a significant role in data collection and dissemination in various applications. In hierarchically clustered WSN models, the cluster heads (CHs) consume more energy due to the additional workload of receiving, aggregating, and transmitting data from their member nodes to the sink. The CH selection is important in extending the lifetime of WSNs by conserving energy expenditure at sensor nodes. Therefore, this paper proposes a Deep Learning based Enhanced Data Aggregation with Multi-Objective Optimization method in Wireless Sensor Network. The Gazelle Optimization Algorithm (GOA) is employed for the energy-efficient cluster-head selection with the incorporation of a well-defined fitness function constructed with intra-cluster proximity, sink proximity, and the residual energy. A single candidate optimizer (SCO) is introduced to select the most suitable routes from the CH to the sink node, considering factors such as energy and proximity. To enhance the efficiency of data aggregation (DA), a novel approach is introduced using a Self-Attention-Based Provisional Variational-Auto-Encoder Generative-Adversarial-Network (SPVAGAN). The proposed framework demonstrates its effectiveness in Energy Consumption (Ec), Packet Delivery Ratio (PDR), End-to-End Delay (E2ED), Communication Overhead (CoH), and Data Accuracy over the other models.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Deep learning based Enhanced Data Aggregation with Multi-Objective Optimization method in Wireless Sensor Networks

  • 期刊论文

  • 发表日期：

    2025年3月28日

查看全文
  • Thiyagarajan N,
  • Shanmugasundaram N

摘要

Nowadays, Wireless Sensor Networks (WSNs) play a significant role in data collection and dissemination in various applications. In hierarchically clustered WSN models, the cluster heads (CHs) consume more energy due to the additional workload of receiving, aggregating, and transmitting data from their member nodes to the sink. The CH selection is important in extending the lifetime of WSNs by conserving energy expenditure at sensor nodes. Therefore, this paper proposes a Deep Learning based Enhanced Data Aggregation with Multi-Objective Optimization method in Wireless Sensor Network. The Gazelle Optimization Algorithm (GOA) is employed for the energy-efficient cluster-head selection with the incorporation of a well-defined fitness function constructed with intra-cluster proximity, sink proximity, and the residual energy. A single candidate optimizer (SCO) is introduced to select the most suitable routes from the CH to the sink node, considering factors such as energy and proximity. To enhance the efficiency of data aggregation (DA), a novel approach is introduced using a Self-Attention-Based Provisional Variational-Auto-Encoder Generative-Adversarial-Network (SPVAGAN). The proposed framework demonstrates its effectiveness in Energy Consumption (Ec), Packet Delivery Ratio (PDR), End-to-End Delay (E2ED), Communication Overhead (CoH), and Data Accuracy over the other models.