Hello
Firstly, I wonder if this is even a viable idea for GSoC, since it is quite specific and there might be no mentors for it.
Anyway, given that all ADSL2+ modems as far as I know, are either or both implemented in hardware and proprietary, I wish to create a (almost) complete open source implementation in software, much like software defined radio.
From reading the ADSL2+ standard so far, the only bottleneck I see that might be a problem in the modem is an 1024-point IFFT that needs to be run at ~4 kHz.
Anyway, this was written mainly just to see if there is anyone that can mentor me on this project.
Thanks in advance
Hello
Firstly, I wonder if this is even a viable idea for GSoC, since it is
quite specific and there might be no mentors for it.
Anyway, given that all ADSL2+ modems as far as I know, are either or both
implemented in hardware and proprietary, I wish to create a (almost)
complete open source implementation in software, much like software defined
radio.
"Almost" open source?? There's no such thing, either it's open source or it
ain't. To be viable for GSoC it has to be open source.
From reading the ADSL2+ standard so far, the only bottleneck I see that
might be a problem in the modem is an 1024-point IFFT that needs to be run
at ~4 kHz.
Anyway, this was written mainly just to see if there is anyone that can
mentor me on this project.
What's the use case for a BeagleBone based ADSL2+ modem? And how will the
BeagleBoard.org community benefit from this project? If you can answer that
question then it may be a viable GSoC project.
“Almost” open source in the sense that maybe some hardware components used, such as the ADC and DAC, might not be available open source.
However, all the software and schematics will be open source. That is, anything that I end up writing myself will be open sourced.
The goal is to create something that can easily be ported to multiple platforms but easily hackable - think OpenWRT.
Making it a completely custom board would make the project quite inaccessible to most users of OpenWRT. If it is designed around an existing open platform with the addition of only a few extra hardware components (ADC, DAC), then it would be much more accessible.
Indeed I could use the Raspberry Pi instead, and possibly the higher speed Arduinos, but neither of them are mentors for this year’s GSoC.
Porting the modem to them would also be a goal, however.
