Processor for Measuring Radio Network Design Quality

Juan A Gomez-Pulido,Paulo J Cordeiro,Silvio Priem Mendes,Juan M Sanchez-Perez,Miguel A Vega-Rodriguez

doi:10.4236/wet.2011.23028

Juan A Gomez-Pulido, Paulo J Cordeiro + Show 3 more

Open Access

https://doi.org/10.4236/wet.2011.23028

Copy DOI

Abstract

In this paper we present the design and prototyping of an arithmetic processor based on reconfigurable technology, whose purpose is to determine in a parallel manner the quality of the solution in a radio network design optimization problem. This problem consists in the search for an optimal set of locations in which to place radio antennas in order to obtain the maximum possible coverage, for a given terrain and antenna characteristics. The original computational contribution of this work is to use programmable logic devices to avoid the high cost of computing the evolutionary algorithms required to tackle this optimization problem. This is achieved by means of reconfigurable processors working in parallel. On the basis of the results obtained from the prototype, it may be considered a parallel architecture capable of achieving a great acceleration in the calculations.

Highlights

The Radio Network Design Problem (RND) originated in the context of wireless communication technologies
In this paper we present the design and prototyping of an arithmetic processor based on reconfigurable technology, whose purpose is to determine in a parallel manner the quality of the solution in a radio network design optimization problem
The RND problem is an optimization problem belonging to NP-Hard class: there are a great number of possible solutions, prohibiting the determination of the optimal one through their sequential evaluation

Summary

Introduction

The Radio Network Design Problem (RND) originated in the context of wireless communication technologies. The RND problem consists of minimizing the number of transmitting base stations (referred to from here on as antennas) and establishing their optimal locations, with the goal of obtaining the maximum coverage area and providing services to a larger number of terminals. Of a set of antennas placed in any given manner in the network This fitness function can be obtained from the coverage rate and the number of antennas [1,2], as shown in Equation (1):

F Coverage2

Fitness Processor Prototype

GB RAM

Performance Study

Increasing the Performance

Conclusions and Future Works