Mathematical Models of Intrusion Detection by an Intelligent Immunochip

Abstract

Based on mathematical models of immunocomputing, this paper proposes an approach to intrusion detection that allows both low-level signal processing (feature extraction) and high-level ("intelligent") pattern recognition.The key model is the formal immune network (FIN) including apoptosis (programmed cell death) and immunization both controlled by cytokines (messenger proteins). FIN can be formed from the raw signal using discrete tree transform, singular value decomposition, and the proposed index of inseparability in comparison with the Renyi entropy. The speed and the accuracy of the approach probably mean a further step toward placing more of the intelligent functions on the chip.

Immunocomputing for intelligent signal processing

“…These advances of the IC approach together with its biologic nature probably mean a further step toward placing on the chip more of the functions of intelligent signal processing [34,45]. For example, this chip could be selfadapted to different types of signals (e.g., audio, video, sonar, radar, etc.)…”

“…Then FIN processes the cells using apoptosis (programmed cell death) and immunization both controlled by cytokines (messenger proteins) [36]. The result of such feature extraction by FIN is estimated by the index of inseparability which has been proposed in [34] in comparison with the Renyi entropy [37,38]. In the recognition mode, current signal (antigen) is pre-processed by DTT, mapped to the cells of FIN, and recognized by cytokine class of the nearest cell of the FIN.…”

“…A connection of IC with cellular automata (CA) [23] looks encouraging for threedimensional (3D) computer graphics [24] and identification of CA [25]. The IC is also connected with advances in applied self-organizing systems [26] including intelligent reconstruction and simulation of hydro-physical fields [27,28], signal processing [29][30][31] and intrusion detection [32][33][34].…”

“…In the recognition mode, current signal (antigen) is pre-processed by DTT, mapped to the cells of FIN, and recognized by cytokine class of the nearest cell of the FIN. This approach is implemented as a computing scheme (architecture) of an intelligent immunochip [34] and compared with neural computing and support vector machines using a fragment of UCI KDD archive [39].…”