An Enhanced Security Algorithm for Wireless Sensor Networks

In recent years, the significance of Wireless Sensor Networks (WSNs) has increased due to their cost-effective deployment and widespread internet accessibility. WSNs have found applications in various domains such as military, aviation, healthcare, home automation, transportation, fitness, and forestry. However, the finite energy source of these sensors necessitates the replacement of depleted energy. Moreover, WSNs face numerous challenges including high bandwidth requirements, power consumption, Quality of Service (QoS) delivery, data processing and compression techniques, and cross-layer design. Denial-of-Sleep (DoSL) attacks further exacerbate these challenges. These attacks exploit the sleep mode, which is crucial for minimizing power consumption in sensor nodes. By flooding the targeted nodes with unnecessary fake packets, DoSL attacks prevent them from entering sleep mode, ultimately depleting their power supply and adversely affecting network performance and lifespan. Hence, practical and effective solutions are necessary to tackle these issues. This paper addresses the emerging problems and challenges related to security in WSNs. The proposed solution, namely the DSD-RSA algorithm, integrates existing models such as the S-MAC protocol, AES algorithm, and RSA algorithm. DSD-RSA aims to minimize power consumption and end-to-end delay while maximizing network throughput in WSNs. To evaluate the effectiveness of DSD-RSA, we conducted a comparative analysis with three traditional algorithms, namely WSN-FAHN, GA-DoSLD, and GSHMAC, using the Network Simulator-2 (NS-2) tool. Computer simulation results demonstrated that the proposed DSD-RSA algorithm outperformed the traditional algorithms in terms of average throughput, average delay, and energy consumption. Specifically, the DSD-RSA algorithm achieved an average throughput of 95%, an average end-to-end delay of 19%, and an average energy consumption of 30%. In contrast, WSN-FAHN, GA-DoSLD, and GSHMAC algorithms yielded an average throughput of 60%, 89%, and 45%, an average end-to-end delay of 40%, 35%, and 60%, and an average energy consumption of 50%, 35%, and 45%, respectively.