Turbocell/WORP status

[tom at ceisystems.com]

I'd actually suggest using something like WICCP or FROTTLE.  They
implement polling-based communication, and are already implementable.

Thomas Cameron
CEI Systems


> -----Original Message-----
> From: hostap-bounces+tom=ceisystems.com at shmoo.com 
> [mailto:hostap-bounces+tom=ceisystems.com at shmoo.com] On 
> Behalf Of Jouni Malinen
> Sent: Thursday, October 02, 2003 12:12 AM
> To: hostap at shmoo.com
> Subject: Re: Turbocell/WORP status
> 
> 
> On Wed, Oct 01, 2003 at 01:43:29PM +0200, Maxent s.r.o. wrote:
> 
> > I've searched mailing list and found latest question about 
> Turbocell a
> > half year old with answer of non probability of 
> implementing due to hw 
> > problems. But because Turbocell/WORP and similar 
> technologies can have 
> > quite big impact on wireless network performance, I'm 
> asking again: Is 
> > there any developement in this area, or are there any 
> recomendations, 
> > especially on buying hw for hostap which could have the 
> possibility of 
> > supporting of these technologies in some future release 
> (again big maybe) ?
> 
> Well, one thing about Turbocell/WORP is that at least I'm not 
> very fond of such proprietary solutions.. I just took a quick 
> look at a WORP technology overview and can't believe that 
> someone is still claiming something like "the protocol is not 
> publicized or standardized, which makes it less vulnerable 
> for hackers ..."
> 
> Anyway, something similar to these (i.e., yet another proprietary
> solution) could probably be implemented on top of any Prism2 
> driver. One could either use IEEE 802.11 ACKs and 
> re-transmits by selecting Host AP mode or alternatively, try 
> using pseudo-adhoc mode without ACKs and do all the needed 
> retransmits in the driver. Polling would also be implemented 
> on top of IEEE 802.11 with special data frames (etc.). In 
> addition, the driver could combine short packets into one 
> IEEE 802.11 frame to improve performance.
> 
> This does not seem to be horribly complex, but would probably 
> require quite a bit of work to get it working with optimal 
> performance. Simpler proof-of-concept type of polling system 
> should be doable relatively quickly.
> 
> Apart from the technical issues, the remaining main issue is 
> to get someone interested in implementing this.. Getting an 
> open source alternative for such a protocol might be useful; 
> especially, if the design would be well documented and 
> interoperable with devices from different IEEE 802.11 vendors.
> 
> -- 
> Jouni Malinen                                            PGP 
> id EFC895FA
> _______________________________________________
> HostAP mailing list
> HostAP at shmoo.com
> http://lists.shmoo.com/mailman/listinfo/hostap
>