This paper addresses the formal conditions underlying the autonomous emergence of a network hierarchy in large mobile networks under multihop, broadcast communication. The presence of a hierarchical structure is demonstrated as a viable means to support node cooperation, while keeping controlled the communication overhead. The proposed strategy for the creation of a network hierarchy relies on a multilevel approach, where each node broadcasts information over a distance (in hops) related to its hierarchical level. The hierarchical relationships among nodes were purposely chosen to be loose, so as to be reactive to the dynamical changes imposed by mobility to the network topology. A theoretical framework for the hierarchical network formation/maintenance is presented, while a fully distributed procedure for its emergence is proposed. In addition, simulation results on the convergence properties of the procedure are provided.

This paper presents the design of a novel architectural implementation of the LOCO-I compression scheme, the lossless/near-lossless algorithm used inside the JPEG-LS standard. Differently from what previously reported in literature, the proposed design fully exploits the sequential nature of the algorithm by means of a pipelined architecture, without modifications to the original compression scheme. The result is a good performance circuit well fitted for field-programmable gate-array realization, thus devised for application in the wearable computers and remote sensing domains.

In this paper the indoor wireless localization problem is addressed both from the theoretical and application standpoints. The main result of the paper is on the theoretical side: the topological definition of regular and irregular nodes is introduced, and formal results are presented to support regularity as a desirable network property for the attainment of precise node localization. In force of this definition, a mixed convex/non-convex optimization approach has been derived for the solution of the positioning problem. The two procedures, suitably combined, allow the achievement of better convergence towards the best positioning of a multitude of blind wireless nodes. A completely decentralized, partially asynchronous algorithm is presented, which proceeds locally on each node based on the sole knowledge of the distances measured from, and of the estimated positions of the connected nodes only. Its repeated asynchronous application on each nodes guarantees the convergence of the algorithm to the positioning of the whole network, even in presence of a limited number of peripheral reference points. Indeed, no global information is required for the proper functioning of the algorithm. Simulations of relevant case studies have been performed to qualify the proposed scheme in realistic conditions, and the results are presented.

This paper analyzes the criteria for the direct correspondence between a deterministic neural network and its stochastic counterpart, and presents the guidelines that have been derived to establish such a correspondence during the design of a neural network application. In particular, the role of the slope and bias of the neuron activation function and that of the noise of its output have been addressed, thus filling a specific literature gap. This paper presents the results that have been theoretically derived in this regard, together with the simulations of few relevant application examples that have been performed to support them.

This paper presents a new kind of architecture for artificial digital neurons based on the voting circuit, which may be considered an improved version of those presented in literature. Stochastic pulse modulation has been used, where the values of the neuron's inputs are coded in terms of bit probabilities. The resulting activation function closely resembles the logistic sigmoid, with a transition slope that can be selected at the architectural level with no additional hardware requirements. The proposed neuron architecture has been simulated in software. Simulation results confirm that the neuron features a sigmoid transfer characteristic similar to that of conventional voting circuits. The resource occupation of the neuron, as obtained from implementation on reconfigurable platforms, has been estimated to be significantly lower than previous implementations. The theoretical analysis of the neuron's behavior is also presented.

A numerical method for the determination of isotropic band models has been developed and applied to silicon. The resulting model accurately approximates both density of states and group velocity of the corresponding anisotropic band structure, thus providing an excellent agreement with both the collision and nonhomogeneous terms of the Boltzmann transport equation. The model, represented by a simple set of energy-wave vector tables, has been implemented in a Monte Carlo device simulator, but can also be extended to alternative methods for solving the Boltzmann equation. Simulation of homogeneous silicon shows a very good agreement with available experimental data. Comparison with results obtained using the complete anisotropic band structure, both in homogeneous and nonhomogeneous silicon devices, confirms the validity of the model.>