Abstract
The recent development of Global Navigation Satellite Systems (GNSS)
software receivers has brought a new perspective to receiver design.
The first step necessary to use them is to have access to sampled
data at an Intermediate Frequency (IF). These sampled data can be
provided through two different ways. First, an RF frontend that would
sample data coming from either the real satellites or from a hardware
signal generator, if the latter exists at all for the signals of
interest. Real data might not be suitable when a specific aspect
of GNSS is studied, especially in a research and development context,
or might not yet be available as in the case of GPS L5 and Galileo
signals. Hardware signal simulators were created to fill that gap,
and are now widely used. However, they are very expensive, may not
be available for the signals of interest and consequently might not
be always suitable for specialized research due to their lack of
flexibility. Sampled data can also be provided by a software IF signal
generator. Such a tool includes signal simulation as well as front-end
filtering and sampling, fully software-based, and can fed directly
into a software receiver. Such programs are already used and are
often created by the research groups themselves to fit their research
programs. Although they offer a total control of all the parameters
for simulation purpose, they are often highly specialized and, as
such tend to model only a part of the error sources of interest,
e.g., jamming or multipath. The Position, Location And Navigation
(PLAN) research group of the Department of Geomatics Engineering
has developed a complete IF signal generator that can model the new
GPS and Galileo signals propagation channels while being versatile
and customizable. This signal generator allows for the modeling of
pre-defined environments for specific applications. It offers a direct
access to IF sampled data that can be directly used by a software
receiver. From an ephemeris file, an entire constellation of satellites
can be easily simulated. The entire GNSS propagation channel is then
modeled. Finally, the front-end filter and the Analog-to-Digital
Converter (ADC) (1-bit quantization) are simulated. A first Matlab/C,
GPS-only version of this IF signal generator modeled thermal noise,
satellite clock errors, atmospheric errors, the front-end filter,
and an ADC. The current enhanced version, written in C++ to improve
processing speed, now includes multipath modeling,
Links and resources
Tags