An Electronic Reconfigurable Neural Architecture for Intrusion Detection

The explosive growth of the traffic in computer systems has made it clear that traditional control techniques are not adequate to provide the system users fast access to network resources and prevent unfair uses. In this paper, we present a reconfigurable digital hardware implementation of a specific neural model for intrusion detection. It uses a specific vector of characterization of the network packages (intrusion vector) which is starting from information obtained during the access intent. This vector will be treated by the system. Our approach is adaptative and to detecting these intrusions by using a complex artificial intelligence method known as multilayer perceptron. The implementation have been developed and tested into a reconfigurable hardware (FPGA) for embedded systems. Finally, the Intrusion detection system was tested in a real-world simulation to gauge its effectiveness and real-time response.

Simulation of the secondary electrons energy deposition produced by proton beams in PMMA: influence of the target electronic excitation description

We have studied the radial dependence of the energy deposition of the secondary electron generated by swift proton beams incident with energies T = 50 keV–5 MeV on poly(methylmethacrylate) (PMMA). Two different approaches have been used to model the electronic excitation spectrum of PMMA through its energy loss function (ELF), namely the extended-Drude ELF and the Mermin ELF. The singly differential cross section and the total cross section for ionization, as well as the average energy of the generated secondary electrons, show sizeable differences at T ⩽ 0.1 MeV when evaluated with these two ELF models. In order to know the radial distribution around the proton track of the energy deposited by the cascade of secondary electrons, a simulation has been performed that follows the motion of the electrons through the target taking into account both the inelastic interactions (via electronic ionizations and excitations as well as electron-phonon and electron trapping by polaron creation) and the elastic interactions. The radial distribution of the energy deposited by the secondary electrons around the proton track shows notable differences between the simulations performed with the extended-Drude ELF or the Mermin ELF, being the former more spread out (and, therefore, less peaked) than the latter. The highest intensity and sharpness of the deposited energy distributions takes place for proton beams incident with T ~ 0.1–1 MeV. We have also studied the influence in the radial distribution of deposited energy of using a full energy distribution of secondary electrons generated by proton impact or using a single value (namely, the average value of the distribution); our results show that differences between both simulations become important for proton energies larger than ~0.1 MeV. The results presented in this work have potential applications in materials science, as well as hadron therapy (due to the use of PMMA as a tissue phantom) in order to properly consider the generation of electrons by proton beams and their subsequent transport and energy deposition through the target in nanometric scales.

Can model Hamiltonians describe the electron–electron interaction in π-conjugated systems?: PAH and graphene

Model Hamiltonians have been, and still are, a valuable tool for investigating the electronic structure of systems for which mean field theories work poorly. This review will concentrate on the application of Pariser–Parr–Pople (PPP) and Hubbard Hamiltonians to investigate some relevant properties of polycyclic aromatic hydrocarbons (PAH) and graphene. When presenting these two Hamiltonians we will resort to second quantisation which, although not the way chosen in its original proposal of the former, is much clearer. We will not attempt to be comprehensive, but rather our objective will be to try to provide the reader with information on what kinds of problems they will encounter and what tools they will need to solve them. One of the key issues concerning model Hamiltonians that will be treated in detail is the choice of model parameters. Although model Hamiltonians reduce the complexity of the original Hamiltonian, they cannot be solved in most cases exactly. So, we shall first consider the Hartree–Fock approximation, still the only tool for handling large systems, besides density functional theory (DFT) approaches. We proceed by discussing to what extent one may exactly solve model Hamiltonians and the Lanczos approach. We shall describe the configuration interaction (CI) method, a common technology in quantum chemistry but one rarely used to solve model Hamiltonians. In particular, we propose a variant of the Lanczos method, inspired by CI, that has the novelty of using as the seed of the Lanczos process a mean field (Hartree–Fock) determinant (the method will be named LCI). Two questions of interest related to model Hamiltonians will be discussed: (i) when including long-range interactions, how crucial is including in the Hamiltonian the electronic charge that compensates ion charges? (ii) Is it possible to reduce a Hamiltonian incorporating Coulomb interactions (PPP) to an 'effective' Hamiltonian including only on-site interactions (Hubbard)? The performance of CI will be checked on small molecules. The electronic structure of azulene and fused azulene will be used to illustrate several aspects of the method. As regards graphene, several questions will be considered: (i) paramagnetic versus antiferromagnetic solutions, (ii) forbidden gap versus dot size, (iii) graphene nano-ribbons, and (iv) optical properties.

Electronic properties of transition metal atoms on Cu2N/Cu(100)

We study the nature of spin excitations of individual transition metal atoms (Ti, V, Cr, Mn, Fe, Co, and Ni) deposited on a Cu2N/Cu(100) surface using both spin-polarized density functional theory (DFT) and exact diagonalization of an Anderson model derived from DFT. We use DFT to compare the structural, electronic, and magnetic properties of different transition metal adatoms on the surface. We find that the average occupation of the transition metal d shell, main contributor to the magnetic moment, is not quantized, in contrast with the quantized spin in the model Hamiltonians that successfully describe spin excitations in this system. In order to reconcile these two pictures, we build a zero bandwidth multi-orbital Anderson Hamiltonian for the d shell of the transition metal hybridized with the p orbitals of the adjacent nitrogen atoms, by means of maximally localized Wannier function representation of the DFT Hamiltonian. The exact solutions of this model have quantized total spin, without quantized charge at the d shell. We propose that the quantized spin of the models actually belongs to many-body states with two different charge configurations in the d shell, hybridized with the p orbital of the adjacent nitrogen atoms. This scenario implies that the measured spin excitations are not fully localized at the transition metal.