An optimized deep neural network based DoS attack detection in wireless video sensor network

Abstract

Due to the technological advancements in sensor and modern devices, the Wireless Multimedia Sensor Network (WNSMs) have gained attraction among various communities. It makes it firm and remarkable achievements in wireless communication, media transfer and digital transmission by adopting portable technologies. In recent decades, in applications such as agriculture and industry the sensor nodes are intended to measure the parameters such as temperature, moisture content and other environmental factors. WNSMs have also facilitates the applications by providing the devices with self-governing access for sending and processing the data associated with suitable audio and video information.Numerous video sensor network researches focus on reducing power consumption and optimizing transmission capacity, but data reliability is the real need. The hindrance involved in the sensor nodes result into various attacks on WMSN especially Denial of Service (DoS) attacks. The core of this attack is to incapacitate proper functioning of the network. Both Attacker and attackers, focus on attacking and preventing the legitimate network nodes from using the network resources. Distributed attack is defined as the network which is being attacked by multiple attackers. This type of attack causes significantly more issues in functionality of the network than attacks involved on a single node. Redemption of the network from the threats of DoS attacks, an enhanced machine learning algorithm has to be presented. For detecting DoS in WMSN an optimized Deep Neural Network algorithm is proposed. The parameters required are chosen from adaptive particle swarm optimization algorithm. The approach’s efficiency to be measured by the ratio of packet transmission, energy consumption, latency, network length, and throughput.