This dissertation explores an immunological model of distributed detection, called negative detection, and studies its performance in the domain of intrusion detection on computer networks. The goal of the detection system is to distinguish between illegitimate behaviour (nonself ), and legitimate behaviour (self ). The detection system consists of sets of negative detectors that detect instances of nonself; these detectors are distributed across multiple locations. The negative detection model was developed previously; this research extends that previous work in several ways. Firstly, analyses are derived for the negative detection model. In particular, a framework for explicitly incorporating distribution is developed, and is used to demonstrate that negative detection is both scalable and robust. Furthermore, it is shown that any scalable distributed detection system that requires communication (memory sharing) is always less robust than a system that does not require communication...

Natural immune systems provide a rich source of inspiration for computer security in the age of the Internet. Immune systems have many features that are desirable for the imperfect, uncontrolled, and open environments in which most computers currently exist. These include distributability, diversity, disposability, adaptability, autonomy, dynamic coverage, anomaly detection, multiple layers, identity via behavior, no trusted components, and imperfect detection. These principles suggest a wide variety of architectures for a computer immune system.

In the genetically restricted response that follows immunization with (4-hydroxy-3-nitrophenyl)acetyl coupled to protein carriers, two distinct populations of B cells are observed in the spleens of C57BL/6 mice. By 48 h postimmunization, loci of antigen-binding B cells appear along the periphery of the periarteriolar lymphoid sheaths. These loci expand to contain large numbers of antibody-forming cells that neither bind the lectin, peanut agglutinin, nor mutate the rearranged immunoglobulin variable region loci. Germinal centers containing peanut agghtininpositive B cells can be observed by 96-120 h after immunization. Although specific for the immunizing hapten, these B cells do not produce substantial amounts of antibody, but are the population that undergoes somatic hypermutation and affinity-driven selection. Both focus and germinal center populations are paucidonal, founded, on average, by three or fewer B lymphocytes. Despite the highly specialized roles of the focus (early antibody production) and germinal center (higher affinity memory cells) B cell populations, analysis of VH to D to J. joins in neighboring loci and germinal centers demonstrate that these B cell populations have a common clonal origin. D uring the course of the primary immune response, participant

Abstract not found

pie contributions, which drives the trami- tion from the unolded to the native state (Matthews, 1987). Thus the rational engineering of either structure or stability remains an nnsolved problem (Shi et al., 1993). The alternative pproach of screening random point mutations (lgoltence et al., 1988; Chen & Baldwin, 1989; Turner et at, 1992) succeeds in only one out of 10 a to 10 mutations (Risse et al., 1992; Arose et al., 1993) and screening procedures am rctrcted to proteins with some identifiable timetlon mmh as enzymatic activity. We have addressed the problem of stabilizing n immunoglohulin domain, wklx an approach that analyses the immunoglobulin sequence database (Ksbat eta!., 1991) in the conceptual framework of statistical mechanics. Clusters of sequence variability in well delineated regions of the antibody molecule provide the molecular basis for the capacity of the immune Author to whom all correspondence should be addresse& system to respond to a large variety of antlgeni

Abstract not found

We first compare the paradigm of the immunity-based system with other biological paradigm such as neural networks and genetic algorithm. We review studies of the immunity-based system dividing them into three classes: recognition/learning based on Jerne's network, adaptation based on Edelman's selection principle, and search/optimization based on Holland's genetic algorithm. We also propose that the self-identification process is yet another important aspect of the immune system, which has not been much studied. Some implications to computer networks will be discussed.

this paper we outline the biological principles we believe to be most relevant to understanding and designing the computational networks of the future. Among the principles of living systems we see as most important to the development of robust software systems are: Modularity, autonomy, redundancy, adaptability, distribution, diversity, and use of disposable components. These are not exhaustive, simply the ones that we have found most useful in our own research. We then describe a prototype network intrusion-detection system, known as LISYS, which illustrates many of these principles. Finally, we present experimental data on LISYS' performance in a live network environment

The conservative physiology of the immune system Departamentos de 1 Bioquímica e Imunologia, and 2 Morfologia,

Chair of Advisory Committee: Dr. Andreas Holzenburg Three-dimensional (3D) structural determination of biological macromolecules is not only critical to understanding their mechanisms, but also has practical applications. Combining the high resolution imaging of transmission electron microscopy (TEM) and efficient computer processing, protein structures in solution or in two-dimensional (2D) crystals can be determined. The lipid monolayer technique uses the high affinity binding of 6His-tagged proteins to a Ni-nitrilotriacetic (NTA) lipid to create high local protein concentrations, which facilitates 2D crystal formation. In this study, several proteins have been crystallized using this technique, including the SARS virus Nsp15 endonuclease and the human Toll-like receptor (TLR) 3 extracellular domain (ECD). Single particle analysis can determine protein structures in solution without the need for crystals. 3D structures of several protein complexes had been solved by the single particle method, including IniA from Mycobacterium tuberculosis, Nsp15 and TLR3 ECD. Determining the structures of these proteins is an important step toward understanding pathogenic microbes and our immune system.