Learn More
Intrusion detection systems (IDS's) ensure the security of computer networks by monitoring traffic and generating alerts, or taking actions, when suspicious activities are detected. This paper proposes a network-based IDS supporting an intuitive visualization of the time evolution of network traffic. The system is designed to assist the network manager in(More)
The paper presents an efficient hardware realization of circular back propagation (CBP) networks on digital signal processors (DSP). The resulting neural-system is aimed at enhancing "smart" TV displays and supports the estimation of perceptual quality of visual signals. The DSP-based neuro-platform operates on raw digital signals and yields the associate(More)
The paper investigates the representation properties of a new neural network architecture based on adaptive activation function, called Generalized Sigmoidal Neural Network (GSNN) and analyzes the advantages of this model in the application domain of image compression. As a consequence, the basic network scheme is most similar to classical structures with(More)
The Montgomery multiplication algorithm is one of the most efficient techniques used to speed up the modular multiplications required during exponentiation processes in public-key cryptography. In view of the expanding demand of embedded security services, this paper tackles the problem of efficiently supporting Montgomery's method on inexpensive digital(More)
In this paper an implementation of the back propagation (BP) algorithm (Rumelhart86) on a cluster of workstations is presented. In a previous paper (Anguita94), we proposed an efficient implementation of BP on RISC architectures (Matrix Back-Propagation MBP): now we extend that work to clusters of workstations using PVM. Several methods have been proposed(More)
A modular design methodology supports CMOS circuits for the analog implementation of the Soft-Max function. An optimization-based strategy allows the designer to fit VLSI-technology requirements to soft max mapping accuracy. Specific circuit solutions to both the approximation of the expO function and the normalizing ratio enhance overall effectiveness by:(More)
A dichotomy between 'analogue' modeling and 'digital' implementation is often encountered when designing vector quantizers. In the case of digital systems, the requirement of optimality can bring about NP-hard problems. The paper discusses the possibility of using advanced paradigms such as quantum computing for digital optimization processes in order to(More)
  • 1