report_nccr_04.tex.backup

软件无线电的平台

\documentclass[a4paper,10pt]{report}
\usepackage{graphicx}
\usepackage{geometry}
\geometry{verbose,a4paper,tmargin=2cm,bmargin=3cm,lmargin=2cm,rmargin=2cm}

% Title Page
\title{Project Report}
\author{B.Rimoldi, L.Gasser}


\begin{document}
\maketitle

\section*{Radio-Platform}

 We have implemented a software-radio. We talk about Software Radio when the
two-way map between the row data (in the memory) and the data-carrying antenna
signal is completely specified by the software. Any map that conforms with the
hardware limitations (power, bandwidth, hardware imperfections) may be
implemented by means of an appropriate (software) code.

With this definition in mind, we decided that our software-radio
platform must meet the following characteristics: 1. \emph{Reconfigurability} so
that all parameters can be changed in real-time and the effects can be made
visible; 2. \emph{Modularity} based on schoolbook-definition of a radio
transmission, in order to be able to implement a new algorithm as fast as
possible; 3. \emph{User Interface} to display the internal state of the
software-radio and to reconfigure as needed; 4. \emph{Simulation} in order to
help debug newly written modules 5. \emph{Real Time} because we don't only want
to simulate the transmission, but go over the air

\begin{figure}
\begin{center}
\begin{tabular}{ccccc}
\begin{minipage}{20mm}
\includegraphics[bb=0mm 0mm 20mm 50mm,
                 width=15mm,keepaspectratio]{software-platform}
\caption{\label{cap:fig_software_platform}The architecture}
\end{minipage}
&
\hspace{10mm}
&
\begin{minipage}{30mm}
\includegraphics[width=30mm,keepaspectratio]{capture_sradio}
\caption{\label{cap:fig_gui}Screenshot of Graphical User Interface}
\end{minipage}
\hspace{10mm}
&
\begin{minipage}{40mm}
\includegraphics[bb=0mm -20mm 100mm 60mm,
                 width=40mm,keepaspectratio]{ics-rf}
\caption{\label{cap:fig_ics_rf}From the air to the computer}
\end{minipage}
\end{tabular}
\end{center}
\end{figure}


The figure \ref{cap:fig_software_platform} shows the internal
structure of our platform. As can be seen, three parts make up the
platform and communicate with each other: 1. \emph{GUI} The Graphical User
Interface, that allows the user to interact with the radio; 2. \emph{Signal
Processing} Where the modules reside and where the signals are generated and
decoded; 3. \emph{Antenna} Is the connection to the outer world, either over the
air, or through a network-connection for simulations.

In figure \ref{cap:fig_gui} you see a screen-shot of
a running session of the software-radio in simulation mode. You can
see the modular structure and the school-book like seperation of the
signal processing task. The upper part relates to the transmitter. The lower
part to the receiver. The horizontal thick bar  with thin marks represents the
time line partitioned into slots. Four slots are visible. The figure also
indicates that the transmitter uses the third slot and the receiver the
first. Furthermore two windows are opened that show
internal signals. The upper window shows the synchronisation signal,
while the lower window represents the received and filtered QPSK
signal.


\section*{MIMO-hardware}

 The counter-part to the software-framework is the hardware that does the
actual transmission and reception of the samples. In order to keep as much
flexibility as possible, we chose an architecture that is capable of
transmitting and receiving a 2MHz-window anywhere in the range of
2.4~-~2.48~GHz, which corresponds to the free ISM-band.

 We chose an architecture that is composed of two parts, as can be seen in
figure \ref{cap:fig_ics_rf}. The ICS-cards are commercially available
aquisition-cards that offer 4 inputs or 4 outputs, respectively. They are
connected to 4 RF-cards that work at an Intermediat Frequency of 70MHz and
can translate this frequency to \hbox{2.4~-~2.48~GHz}. These RF-cards are
capable of outputting 20dBm and have an input-sensitivity of -80dBm.

\section*{LDPC-codes}

 A lot of studies have been done to test and improve LDPC-codes on the
software-radio. LDPC codes are Low Density Parity Check codes that were first
described by Shannon in the fifties, and then re-discovered in the late
nineties.

 The difficulty lies in the decoding of these codes, because the so-called
"message-passing" algorithm is time-consuming and it's difficult to tell when
to stop the algorithm. Points of interest were to compare the theoretical
performance of these codes with the practical performances, as well as to
study the behaviour of these codes in a multi-antenna (MIMO) environment.

\subsection*{SISO}

 In the single-antenna case, we only have the problem of decoding the
LDPC-codes fast enough, as synchronisation is easily done.

 In figure @comparison theoretical-real@ we can see the Bit Error Rate (BER)
in function of the Signal to Noise Ratio (SNR) in both the theoretical case
as well as in a real transmission. As can be seen, the theoretical is very
close to the real curve. 

 Work has been done on optimizing the decoding, and we managed to speed up
the decoding by a factor of 10. Unfortunatly this is still not enough to
decode in real-time using standard hardware. More work needs to be done to
find a better way to do the decoding.

\subsection*{MIMO}

In a multi-antenna system, we have more than one antenna sending and more
than one antenna receiving at the same time. From a theoretical point of
view, things can work out quite nicely and one can get a much faster
transmission. Unfortunatly this is not the case anymore in a real
transmission. We're working on a model to fit better the channel we observe
between the sending and receiving antennas, so that we can take advantage of
the additional antennas.

Once this model is set up and verified, it can serve as a base for other
theories that work with mulitple antenna systems.


\section*{Conclusion}

During these three years we managed to set up a working environment for
physical layer measurements and testing of theoretical models. We also
discovered the limits on simple channel models for multiple antenna systems
and are working on a better model.

For the future we plan to use the software-radio in a bigger class of
students and to teach the principles of digital communication using this
tool.

\end{document}