Antonio Díaz Estrella

Performance Analysis of a Cellular Slotted CDMA System with Imperfect Power Control over a Rayleigh Fading Channel

An analytical approach for the calculation of throughput and delay in a cellular slotted CDMA system with imperfect power control is presented. Power control is essential for such systems for good performance, however, perfect power control is hard to achieve in real systems. Usually, imperfections in power control after despreading are modelled as a lognormal random variable. It has been shown that the standard deviation of this variable is a function of the mobiles velocity and of whether or not the mobile is communicating with the base station where the power is measured. In this work, we show that when this fact is not taken into account the performance analysis results obtained are too optimistic. We use the error statistics to model the intercell interference and evaluate the impact of different mobiles velocity and number of resolvable paths on the system performance.

Intelligent Policing Function for ATM Networks

In this work we present an intelligent system implemented with a hybrid technique that aims to resolve the problem of traffic control in Asynchronous Transfer Mode (ATM) networks. In this system, fuzzy and neural techniques are combined to obtain a policing mechanism simple enough to be cost-effective and with a performance hugely better than the traditional ones. The UPC proposed is very fast, has a high selectivity and presents an immediate hardware implementation.

Simulation of MPEG Video Traffic Using Neural Networks

A new model for the simulation of MPEG video traffic is presented. The model, implemented on neural networks, is capable of accurately adjusting the autocorrelation and probability distribution functions of a given video traffic. This adjustment is performed by capturing the projected conditioned histogram of the real traffic, so that the neural model will be able to yield a simulated traffic at its output, just as a function of an input white noise. Moreover, using neural networks we benefit from their inherent capacities for working in real time, because of their parallelism, and interpolating unknown functions. These interpolations avoid the need of searching in transition matrices of other histogram-based methods as well as they reduce the amount of stored information. Results are presented for a real MPEG video source.

Neural Network Touting Controller for Communication Parallel Multistage Interconnection Networks

In this paper we propose a centralized neural network contention controller for Asynchronous Transfer Mode (ATM) input buffered switches fabrics. These switches are self-routing type, and suffer of head of line (input), internal and output blocking possibilities. Messages (cells) through the switch can collide at these three points. The neural controller will be able to choose within each celt slot, the maximum possible number of connection-requesting cells from the inputs queues with legal routes. The switch will be able of connecting these selected cells to desired output addresses without internal or output blocking possibility. Increased throughput performance will result. We propose and analyze new high performance architectures of parallel connected switches and associated controllers, which, in addition, are fault-tolerant. Simulation results show low cell loss ratio for small size of the input queues.