Introduction
|
|
After the
great success of wireless communications used in land
mobile radio systems, wireless data exchange in time
variant ad-hoc networks become interesting. The steadily
increasing demand for mobile multimedia and safety
applications requires looking for new concepts for the
planning of wireless systems. Time variant scenarios can
be found in several environments: Some examples are
car-to-car communication scenarios used for driving
assistance systems or Wi-Fi hotspots in railroad
stations, airports or downtown areas. The main aspect in
such applications is the time variance of these
scenarios, as the locations of either transmitter,
receiver or of obstacles in these scenarios change
continuously. These effects influence the propagation
situation and lead to time variant channel impulse
responses.
|
Example:
Car-to-car Communication
|
|
|
Application:
Car-to-Car Communication
|
|
The
development of ad-hoc networks, which are used for car-to-car
communication, is based on extensive research
activities. One import issue is the radio channel
between the cars, because all the data is tranmitted by
using this channel. Thus it is evident to have a
reliable tranmission link between the cars.
More
information about predictions in car-to-car scenarios
are available on this separate website:
http://www.awe-com.de/Automotive
|
 Computed
propagation paths in a car-to-car scenario.
|
|
|
Application:
Adaptive Cruise Control
|
|
The
development of adaptive cruise control (ACC) is based on
intensive research activities. The investigation of the
radar channel is very important for the adaptation of
the algorithms (for angle and distance estimation) to
the hardware.
More
information about predictions in car-to-car scenarios
are available on this separate website:
http://www.awe-com.de/Automotive
|
Computed
propagation paths in an ACC scenario.
|
|
|
Definition of
Time Variant Scenarios
|
|
|
The database manager WallMan was adapted to the needs of
time variant scenarios. The basis of the databases are
either planar polygons for indoor scenarios or cylinders
with polygonal ground plate for urban environments. Each
element in the database can be either stationary (not
moving) or non-stationary (dynamic). Translation and
rotation vectors, as well as scalar values for the
velocity or the acceleration are assigned to time
variant objects for the definition of their behavior in
the scenario, depending on the time. Predictions are
then accomplished for arbitrary defined timestamps or
with constant intervals for a period of time.
WallMan was also improved concerning the conversion
and handling of large databases. Especially vehilce
databases consist of thousands of polygons. Read
here
more about the vehicle databases. |
Database manager Wallman displaying a dynamic scenario
with several vehicles on a straight street.
|
|
|
Road Course
Generator for Time Variant Vehicle Scenarios
|
|
The newly developed software package
for time variant scenarios offers a tool for the simple
and fast generation of road courses. This tool is
evident when studying in car-2-car communication
scenarios. As typical road courses are not linearly, it
would be very complicated to enter a road course
manually. Therefore the road course can be defined with
different types of parts of streets, e.g. straight
streets and curves. Straight streets are only defined by
their length, whereas curves are defined by angle,
radius and the direction (left or right). Out of these
definitions the road course is created and guardrails
and jambs are automatically added to the borders of the
streets. Buildings, vegetation and other objects can
also be included in time variant scenarios. |
Example of an automatically generated road course, with
guardrails and jambs.
|
|
|
New 3D Ray
Tracing for Time Variant Scenarios
|
|
A
new ray tracing algorithm was developed for prediction
in time variant scenarios. Algorithms for the
computation of diffractions and reflections are well
known and are implemented in the new approach available
for time variant scenarios. Even more, the consideration
of the doppler shift is possible. Each time an
interaction with a moving obstacle occurs, the doppler
shift according to the following equation is determined:
The
usage of polygonal car models is possible with the new
ray tracing algorithm. However one problem is the
complexity of the models: To obtain a realistic image of
the reality very complex car models have to be used.
This leads to much computation effort and extrem long
prediction times. The other way is to use simplified
models with only some polygons. But these models do not
represent the reality, because only few reflections and
diffractions occur on such models. WinProp offers a new
approach to avoid all these problems: The usage of radar
cross sections (RCS) is supported. Complex polygonal
models can be substituted by several bistatic RCS, which
can be measured or computed. The image on the right
shows a polygonal model which was substituted by 20 RCS.
The follwing equation is used to determine the scattered
field:
The
scattering matrix depends on the incident and scattered
angles of the ray path on the scattering center.
|
Computed
Doppler shift.
Substitution
of polygonal model with radar cross sections.
|
|
|
Examples
|
The
following images show some computation results of
WinProp. Please click on the images to enlarge them:
|
|
|
Spatial
channel impulse response (CIR).
|
Doppler
shift for several snapshots.
Click to enlarge.
|
|
|
|
Some
computed propagation paths in a suburban
Car-2-Car scenario.
|
Adaptive
Cruise Control (ACC): Some computed
propagation paths.
|
|
|
Wave
propagation phenomena and propagation paths.
|
|
