Advocating the use of fuzzy reasoning spiking neural P system in intrusion detection

Membrane Computing (MC) with its variants has proved to be a versatile class of distributed parallel computing model. This is because despite its infancy, it has enjoyed significant application in various fields. However, much is yet to be accomplished in the area of information and network security. So, to further explore the efficacy of MC, this paper presents a new attempt in the application of SN P system and as well provides a novel idea and method for attack detection. The extension of SN P system called trapezoidal Fuzzy Reasoning Spiking Neural P (tFRSN P) system is adopted in the network intrusion prediction model. SN P system is a neural-like computing model inspired from the way spiking neurons communicate using spikes. It has a graphical modeling advantage which makes it well suited for fuzzy reasoning as well as fuzzy knowledge representation. In order to evaluate the performance of tFRSN P system in intrusion detection, the publicly available KDD Cup benchmark dataset was employed. After the experiments, our results yielded very high detection rate of 99.78% and very low false alarm rate of 0.16% for Brute Force Attack (BFA).