[M4IF Technotes] Error resiliency test tools

[Ralph Sperschneider]

Oztan Harmanci wrote:
> Hi. I have a couple of questions about error resiliency simulations. Any
> help is greatly appreciated.
>
> It appears that the roots of the error resiliency tests in the VM go
> back to N0999 -'95 Tokyo paper. According to the test results -for
> example- resync marker spacing is advised under different channel
> conditions and bit rates. Channel is simulated using a "NTT DoCoMo"
> originated software, and if I didnt understand wrong it uses a two state
> markov model to simulate the bit errors.
>
> Does anyone know where I can find this "NTT DoCoMo" software

Oztan,
this software is part of the MPEG-4 reference software. Just check out
14496-5.2001/audio/natural/errGen/gilbert.c
in
http://www.iso.ch/iso/en/ittf/PubliclyAvailableStandards/ISO_IEC_14496-5_2001_Software_Reference/audio.zip
Best regards,
Ralph
>  and whether
> this software is applicable to 3GPP1/2 network
> simulations(cdma2000/wcdma type networks)?
>
> There is a similar channel simulation testing software for JVT, namely
> VCEG-N37. The problem with N37 is that it does not allow any
> errored packets to be passed to upper levels. In other words, bit errors
> are not allowed. This confuses me: Assuming that N37 software is a valid
> 3G simulation tool, and that N37 can be changed to simulate MPEG4
> transmissions, what is the point of having bit error resiliency if there
> wont be any? Does that mean MPEG4's bit based error resiliency tools
> are not needed in 3G networks? This means RVLC and data partitioning
> become useless! Am I missing something here?
>
> Thanks
> Oztan

-- 
Dipl.-Ing. Ralph Sperschneider  | Phone: +49 9131 776 344
FhG IIS-A                       | Fax:   +49 9131 776 398
Am Wolfsmantel 33               | mailto:sps   iis.fhg.de
D 91058 Erlangen                | http://www.iis.fhg.de/amm/