Automata-Driven Intrusion Detection with Malware Detection Secure Communication Framework for Wireless Sensor Networks

Abstract

Security vulnerabilities or weaknesses result from the characteristics of WSNs which include: open communication mediums, a decentralized topology (the WSN is not located in a central point, but rather in many different locations), and limited processing power and battery life. This study proposes an Automata-Driven Intrusion Detection with Malware Detection Secure Communication Framework for Wireless Sensor Networks to enhance network security, reliability, and real-time threat mitigation. The results of the JFLAP simulation indicate that the DFA based architecture can effectively detect malicious packets using very simple computational operation requirements. The use of state transition processes and minimal memory requirements make the proposed approach suitable for use within resource-constrained wireless sensor networks. Further, the NS-3 simulation thus indicates that the use of DFA based intrusion detection methods increase the reliability of communications, maintain network efficiency and lower the amount of computational overhead. Finally, The results of the OMNeT++ simulation provide supporting evidence for the use of deterministic finite automata to provide stable routing and reliable communication within networks under attack conditions.