[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

*To*: <starship-design@lists.uoregon.edu>*Subject*: RE: starship-design: Massively Distributed Computing for SETI*From*: "L. Parker" <lparker@cacaphony.net>*Date*: Sun, 18 Mar 2001 11:36:28 -0600*Importance*: Normal*In-Reply-To*: <3AB2F67B.D153684A@jetnet.ab.ca>*Reply-To*: "L. Parker" <lparker@cacaphony.net>*Sender*: owner-starship-design@lists.uoregon.edu

> This may be true but the beam's does not need to updated that > much, I guess > once a minute and that is mostly with the feq of the beam. As Zenon's post points out, this is all dependent upon the beam aperture and distance to target. I posted the relevant equations several years ago, they should be in the archives. Without actually sitting down and figuring it out, I would say that Zenon's guess at ten minutes would be the outside limit for retargeting in 3 dimensions. The frequency shift depends upon two things, the acceleration rate and the nature of the sail and how forgiving it is. Probably not less than every ten minutes though. A typical non-mobile phased array in use today has around 100,000 emitters. Each of which requires a separate solution. There is a limited ability to "group" them to reduce the computation load to manageable levels, which results in some tradeoffs in other areas, but since we aren't talking about using this as a sensor, the tradeoffs aren't relevant. Zenon's best case was a 100 km antenna, which is probably the one we should go with given the power densities we need. Since even relatively small airborne phased array antennas of approximately one meter square contain between 2000 and 3000 emitters, we can use the smaller number to estimate the number of emitters and therefore the computing power necessary. A 100 km antenna contains 10,000,000,000 m^2 * 2,000 emitters or 20,000,000,000,000 emitters each of which requires several Fourier transforms to update its targeting. The computers controlling the one meter square airborne radars are state of the art digital signal processors and we would need the equivalent of 20 billion DSPs to perform these computations for the emitter processing alone. Now factor in the overall targeting solution using the three body solution problem I began with and you begin to see the size of the problem. We haven't even touched on several other of Zenon's objections such as downrange beam degradation, etc. Lee

**Follow-Ups**:**Re: starship-design: Massively Distributed Computing for SETI***From:*Ben Franchuk <bfranchuk@jetnet.ab.ca>

**References**:**Re: starship-design: Massively Distributed Computing for SETI***From:*Ben Franchuk <bfranchuk@jetnet.ab.ca>

- Prev by Date:
**Re: starship-design: Massively Distributed Computing for SETI** - Next by Date:
**Re: starship-design: Massively Distributed Computing for SETI** - Prev by thread:
**Re: starship-design: Massively Distributed Computing for SETI** - Next by thread:
**Re: starship-design: Massively Distributed Computing for SETI** - Index(es):