How does GRIM™ relate to what the manufacturers are doing in their
boxes?
This program is very similar and could be placed in a GPS receiver box instead
of a computer. The computer has some advantages because it is very flexible
and offers users many choices¾there are 70 compression choices in GRIM™.
To place it in a box, users would need to select a subset of these choices.
It seems like it takes a long time to get to the centimeter level of
accuracy. Is that correct?
This (Hydra) is a static example with the troposphere states on. The troposphere
takes a lot of information and slows the process down. It is possible to do
a faster one; it has been done in five minutes. This is not integer fixing.
This is for monitoring a volcano. Once it gets to the millimeter level, it
will stay there for the next 10 years. If users run a single program in integer
fixing mode, the program will fix the integers in a few minutes if the baseline
is not too long while maintaining a centimeter solution.
The dynamic mode (DynaPos) can converge in two minutes. If, for example, there
is a short baseline, users can begin in the narrow lane, switch to the wide
lane, then switch to an ionosphere free to maintain accuracy without interference
from the ionosphere.
If this program was in a GPS receiver "box," a user might have to
select a subset of the solutions. The computer technique offers a choice of
solutions, depending on the user's need.
One more thing: there are many levels of compression, to 800 - 1000 bps. It
is possible to do 12 satellites inside of 1000 bits per second. In fact, there
is a compression option which allows users to squeeze dual carrier and code
observations every second within about 700 bits. Altogether, there are about
70 compression choices in the system. These include things like L1 only, L1
carrier, etc. A technique has been developed that will allow the L5 measurements
to fit in the same bandwidth (12 satellites; 1000 bits) when L5 is introduced
in a few years.
Within Phase I there are two other techniques. NGS is developing other ideas
and RTCM is also working on an approach. At some point over the next year,
those signals will be tested on the HA-NDGPS system.
The plan is for Hawk Run to be operating sometime this fall.
How are the available tools used?
There are a number of ways to use these tools. Users can load GRIM on a laptop
and connect to the receiver. They can compare the data with what is available
from a Web site. The data will be raw prebroadcast data from Hagerstown, MD.
Within GRIM there are real-time libraries that help to interface to users'
applications. Users can simply collect data and compare it to what was broadcast
(e.g., obtained from Web site). The next step is using the libraries available
from GRIM and developing applications.
How far from Hagerstown was the rover data collected?
The rover data was collected about 45.8 km from the reference station at Hagerstown,
MD.
Are buildings and other structures a problem with this broadcast? Are
additional sites needed?
Estimates to achieve the number of additional sites beyond the current NDGPS
plans appear to be between 5 and 50. The cost of a new site is approximately
$750,000. At medium frequencies, terrain masking is not an issue. Propagation
goes a long way. Atmospheric noise is what impacts the range. This occurs
less than 1 percent of the time.
What is your source of orbits?
Broadcast orbits were used, although there are some rapid or predicted orbits
that also could have been used. Broadcast orbits were used for real time,
but precise rapid orbits can be used for post mission.
Note: Broadcast orbits are not as accurate as precise orbits; if users can
get centimeter results from broadcast orbits today they will get better results
either from planned improved broadcast orbits or by using precise rapid orbits
in real time.
What is the format that goes out of these sites?
This is not a universal format. There is not a standard format that could
squeeze data within such a small bandwidth. The goal was to get the information
out in the bit rate available. Most current systems use short-range high data
rate link. This demonstration showed that it is possible to achieve centimeter
level over long ranges. The data at this point is fully described in the broadcast.
After it is demodulated and out put from a GRIM, it can be used by any manufacturer.
Compression has a tremendous impact, especially when L5 becomes available
with a very wide lane. It will be possible to send out information on 9 -
12 navigation satellites. This is still in the conceptual stage, and there
is a lot of work yet to do.
What is the cost of the receiver?
The current receiver design is built on an L1 GPS receiver. These receivers
cost approximately $2,200, but in production, it is likely that they could
be less expensive. The demodulator is actually very straightforward. A demodulator
can be built out of a single field programmable gate array and an analog to
digital (A/D) converter.
What is the cost of a GPS receiver to participate in this test?
Participation is possible with either a single frequency receiver or a dual
frequency receiver. A single receiver, antenna, cables, etc. might cost $4,000.
A dual frequency receiver would likely cost around $10,000 to $15,000 after
antenna, cables, etc. are considered.
