Denial of Service Attack Detection in Adhoc Wireless Network Using Transfer Learning Methods

Deep learning is designed for enhancement of security over the networks using software only. Deployment of Deep Learning technology implies implementation of software based network between two parties (or more) over the bigger physical network. Deep Learning technology can be realized in many applications such as internet (i.e. web applications), intranet (local/private network made between particular candidates for preforming a specific task where this network is separated from the public internet network). The development of internet and computer technologies have motivated the security engineers to design virtual network that can be operated over any physical network and can be used by any number of subscribers to protect the connection privacy. This paper is illustrating the Deep Learning technology enhancement overview and details. Big data can be collected from network monitoring system and hence can be used to train the transfer learning by recurrent neural network for attack detection. The recurrent neural network is reported higher accuracy of attack detection in adhoc in 111 seconds of training time with 98.2 percent of detection accuracy.