Communications and Signal Processing Lab (ComSP)

Title	Caracterización y modelado de canales inalámbricos con variaciones temporales y diversidad espacial
State	Finished
Author	Adrián Sauco Gallardo
Director/s	Unai Fernández Plazaola ,  Luis Díez Del Río
University	Universidad de Málaga
Center	Escuela Técnica Superior de Ingeniería de Telecomunicación
Department	Ingeniería de Comunicaciones
Reading date	22/09/2023
File	PDF

Abstract

This doctoral thesis analyzes some of the characteristics of wireless channels, focusing

on the effect of their temporal variations or the use of spatial diversity techniques. We

studied both the radio channel and the underwater acoustic channel. Switched diversity

techniques and higher-order statistics were considered for the first one. These techniques

have been trending upward since the introduction of distributed cooperative diversity

in relay networks, and these statistics summarize channel’s temporal evolution. For the

second one, we obtained a shallow-water acoustic channel model that allowed for a simu-

lator capable of computing time-variant channel responses. This simulator is a very useful

tool for this kind of channels research as their unconventional characteristics hinder their

characterization via measuring campaign. We published the results for both in high im-

pact factor journals and presented them in national and international congresses. For this

reason, we have chosen to present this doctoral thesis as a compendium of publications.

The received power variability characterizes the radio channel. These temporal varia-

tions can be long and short-term. Short-term variations are caused by multipath propa-

gation interferences. In radio channels, it is usual to consider narrow-band assumptions

where a random signal known as fading models power variations. The statistical analysis

of fading has allowed for channel models with outstanding significance on radio commu-

nications development. Reception diversity is one of the techniques that is widely used

to mitigate the effects of fading. This technique consists in receiving different versions

of the transmitted signal, which is achievable in various manners. One of the most com-

mon is spatial diversity, which has traditionally consisted of equipping the receivers with

multiple antennas. Nonetheless, this kind of diversity can also be obtained nowadays by

establishing communication through different relay stations. Among the existing diver-

sity combination techniques, this work focuses on the switched ones, which consist in

switching conveniently, when the received power change, to a different diversity source.

Some relaying techniques, and switched diversity techniques also, cause discontinuities

on the received power, which complicates the higher-order statistics analysis that models

its temporal variability. On this report we employ a pioneering framework to study these

statistics and obtain closed-form expressions for them. This analysis resolves how fading

behaves temporally in a switched diversity system and admits all kind of statistical distri-

butions and correlation models for random signal. In line with this work we also present

i

ii

an analysis for the higher-order statistics under high SNR conditions, which concludes

that they are asymptotically independent of temporal correlation.

The underwater acoustic channel, and especially the shallow water channel, is con-

sidered among the most hostile communication channels. They behave like a low-pass

filter with high propagation loss, and a low speed of propagation depicts them. The lat-

ter causes significant signal Doppler spread not only in mobile communications, but also

in quasi-static conditions, where the only variations derive from weather conditions as

swell and tides that disturb the channel geometry. In addition, the shallow water channel

presents a strong multipath effect that enhances temporal variations as the water surface

and the ocean bottom act as wave reflectors in these channels. The multipath propaga-

tion, coupled with the slow speed of propagation, results in a channel impulse response

with an extraordinarily long delay profile. Channels with significant frequency and time

spread are known as overspread channels. This kind of channel responses last longer than

the time it takes for them to vary significantly. This characteristic makes it extremely

difficult to obtain a meaningful characterization of them by sounding in measurement

campaigns. Therefore, this work also presents the mathematical model for a simulator

of mobile acoustic channels in shallow waters. The basis for this model is a geometrical

approach based on ray tracing between transceivers. Each of the propagation paths is

modelled by a frequency response constructed from the expressions for absorption and

reflection coefficients found in the literature. Considering the different path delays the

frequency responses get superimposed, from which we obtain the static channel impulse

response by means of the inverse Fourier transform. We use this to propose different sys-

tem structures to obtain time-varying responses from different static channel invariant

responses. This model was corroborated by comparing it with a measurement campaign

of a real channel in terms of the scattering function. This function shows the channel

power distribution in frequency and time. This research confirms that these channels can

result overspread even at low transceivers speed. Moreover, our in-depth analysis of the

channel variant response has led to the conclusion that underwater acoustic channels

are not reciprocal in terms of transceiver mobility. In other words, the channel response

cannot be expressed in terms of the relative speed between transmitter and receiver, but

it is necessary to take into account the absolute speed of them.