GPS Navigation for Ship - Single Receiver or Multiple Receivers?

On 10 Jun 2004 16:58:54 GMT, Stan Gosnell <fakename@fakeaddress.com> wrote:

>Well, you certainly should know the bow heading.  Even if the ship is 
>stationary, the magnetic compass shows which way the bow is pointing.  That's 
>elementary stuff.  Anyone going out in even a relatively small boat without a 
>mounted compass is mortally stupid.  Large ships do not set sail without 
>working compasses.

We have started fitting GPS compasses to some vessels that we work with - 2
GPS units in the one plastic package about 0.7m away from each other.
"Vector" from CSI Wireless.

Seems to work very well - better than 1 degree of accuracy, easy calibration
& no magnetic deflection trials needed.

(we are using them on oil rigs when we move them to new locations - they
aren't normally required to have magnetic compasses anyway as they aren't
self-propelled).

Dave

The email address used for sending these postings is not valid. 
All replies to the group please.

Actually it is a DOA and beamforming problem. We already have a
geostationary satellite with exactly known position. This satellite is
broadcasting a beacon signal. We need to receive the beacon using an
antenna array installed on board. Due to the GPS resolution limitation
as well as the ship's minor displacement from the GPS measured
position, we need to search the satellite beacon within a very small
region. My question is:

1) Is a DOA estimation necessary in this case? Since the searching
region should be very small, maybe +/-0.5degree, I don't know whether
the existed subspace-based DOA methods, such as MUSIC and ESPRIT, can
have such a super resolution and achieve a good estimation of the
beacon coming direction.

2) If we do adaptive beamforming directly without DOA process, how to
get the true location of the signal source?

Thanks.

waters@starhub.net.sg (Yan.L) wrote in message news:<df850db4.0406090812.4e8f7865@posting.google.com>...
> Is one GPS receiver enough to get the precise position? Or we need to
> use multiple GPS receivers, for instance, four GPS receivers installed
> on bow, tail, port and starboard, respectively?
> 
> If we use four GPS receivers, how to manipulate the GPS information
> collected by these four receivers to get the precise position of the
> ship?

"Yan.L" schrieb:

> 1) Is a DOA estimation necessary in this case? Since the searching
> region should be very small, maybe +/-0.5degree, I don't know whether
> the existed subspace-based DOA methods, such as MUSIC and ESPRIT, can
> have such a super resolution and achieve a good estimation of the
> beacon coming direction.

I have no idea. If you use a high quality receiver with a good antenna,
mounted in a way that little multipath occurs, add some short term
averaging or use WAAS, you may get into the 1m range in a typical marine
environment.

Now how that computes into your antenna array adjustment procedure - I
don't know.

- Carsten


-- 
Audio Visual Systems                          fon: +49 (0)2234 601886
Carsten Kurz                                  fax: +49 (0)2234 601887
Von-Werth-Stra�e 111                   email: audiovisual@t-online.de
50259 Pulheim / Germany               WGS84:N50�57'50.2" E06�47'28.5"

In comp.dsp Tim Wescott <tim@wescottnospamdesign.com> wrote:
> Yan.L wrote:
> 
> > Is one GPS receiver enough to get the precise position? Or we need to
> > use multiple GPS receivers, for instance, four GPS receivers installed
> > on bow, tail, port and starboard, respectively?
> > 
> > If we use four GPS receivers, how to manipulate the GPS information
> > collected by these four receivers to get the precise position of the
> > ship?
> 
> Airborne and ground-based vehicles also use inertially enhanced GPS -- 
> basically you have a small inertial measurement unit (three gyros and 
> three accelerometers) and a GPS receiver feeding a computer.  The GPS 
> provides good quality low-frequency data, the inertial measurement unit 
> provides good mid-frequency data, and you implement a great big Kalman 
> filter in the computer.  These systems can provide _very_ good 
> positional data.
> 

Any online papers on doing this?

-- 
	Sander

+++ Out of cheese error +++

Tim Wescott wrote:

> Yan.L wrote:
> 
>> Is one GPS receiver enough to get the precise position? Or we need to
>> use multiple GPS receivers, for instance, four GPS receivers installed
>> on bow, tail, port and starboard, respectively?
>>
>> If we use four GPS receivers, how to manipulate the GPS information
>> collected by these four receivers to get the precise position of the
>> ship?
> 
> 
> Airborne and ground-based vehicles also use inertially enhanced GPS -- 
> basically you have a small inertial measurement unit (three gyros and 
> three accelerometers) and a GPS receiver feeding a computer.  The GPS 
> provides good quality low-frequency data, the inertial measurement unit 
> provides good mid-frequency data, and you implement a great big Kalman 
> filter in the computer.  These systems can provide _very_ good 
> positional data.
> 

Garmin's 2650/60 has a simplified one of these in addition to taking 
input from the speedometer.

Dale

-- 
     _      _     Dale DePriest
    /`) _  //     http://users.cwnet.com/dalede
  o/_/ (_(_X_(`   For GPS and GPS/PDAs

Sander Vesik wrote:

> In comp.dsp Tim Wescott <tim@wescottnospamdesign.com> wrote:
> 
>>Yan.L wrote:
>>
>>
>>>Is one GPS receiver enough to get the precise position? Or we need to
>>>use multiple GPS receivers, for instance, four GPS receivers installed
>>>on bow, tail, port and starboard, respectively?
>>>
>>>If we use four GPS receivers, how to manipulate the GPS information
>>>collected by these four receivers to get the precise position of the
>>>ship?
>>
>>Airborne and ground-based vehicles also use inertially enhanced GPS -- 
>>basically you have a small inertial measurement unit (three gyros and 
>>three accelerometers) and a GPS receiver feeding a computer.  The GPS 
>>provides good quality low-frequency data, the inertial measurement unit 
>>provides good mid-frequency data, and you implement a great big Kalman 
>>filter in the computer.  These systems can provide _very_ good 
>>positional data.
>>
> 
> 
> Any online papers on doing this?
> 

Alas, I don't know.  I have friends who do this, I've read articles 
about it in the popular press (it's been suggested for automotive 
applications), I've read a book that includes it ("Fundamentals of High 
Accuracy Inertial Navigation", Averil B. Chatfield, AIAA, 1997), I've 
even recommended consulting companies to do it for a former employer -- 
but I can't recommend any papers.

I did do a quick search on Altavista using +GPS +inertial as search 
parameters and came up with a bunch of hits.

-- 

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com