Hybrid-based optimization of wireless channel characterization for health services in medical complex environments

Abstract

Nowadays, plenty of simulation techniques based on deterministic methods like Ray Tracing and Ray Launching, combining Geometric Optics and Uniform Theory of Diffraction, are used to predict waves' behaviour within a defined environment and to make better decisions on the deployment of antennas for particular purposes. However, simulations' quality largely depends on the number of rays, providing sufficient angular resolution in the ray launching process. The computational cost of simulations in high definition (HD) prevents their use in complex environments and their low definition (LD) counterparts are used. In this article we propose a technique based on collaborative filtering (CF) to lessen the low quality problems of LD simulations. Such technique outperforms previous works in this field in terms of accuracy, precision and computational cost. Moreover, we show that our approach obtains results very similar to those of HD in much less time. Therefore, using combined deterministic-CF techniques allows the estimation of radio-planning tasks in large, complex medical scenarios with a potentially large amount of transceivers.