Osmocom New Splits stacks for voice is SOLVED and stable with LimeSDR!
Voice is working actually since back then, my fault is using only C117/118 for test call all the time. Now I just test using newer phone both Caller and Callee and voice is work perfectly!
I tested both osmo-trx-lms and osmo-trx legacy. using osmo-trx legacy 0.20 also have better performance using the firmware/gateware version I mention here : firmware/gateware version: LimeSDR-USB_HW_1.3_r3.0.img and LimeSDR-USB_HW_1.4_r2.9.rbf from LimeSuite .
There is trace in wireshark when using old phone as C117/118, the BTS always freeze and send Measurement Indication all the time when calling is made and answered, but with newer phone, the Indication is normal.
Sandi Suhendro djks74@gmail.com wrote:
Voice is working actually since back then, my fault is using only C117/118 for test call all the time. Now I just test using newer phone both Caller and Callee and voice is work perfectly! [...] There is trace in wireshark when using old phone as C117/118, the BTS always freeze and send Measurement Indication all the time when calling is made and answered, but with newer phone, the Indication is normal.
In your tests with C117/118 phones that produced failing results, were those phones running OsmocomBB or Motorola's original firmware? If Motorola phones running their original firmware (which presumably had passed official certification and type approval tests back in the day) don't work with Osmocom networks (be it LimeSDR or other Osmocom- compatible BTS hardware, SDR-based or otherwise) but "newer" (meaning more proprietary and more thoroughly locked down) phones do work, then it is a very grave problem from the perspective of Freedom, and needs to be taken very seriously.
Sysmocom has two of my FreeCalypso development boards, only one of which has been calibrated with a CMU200 instrument and fc-rfcal-tools before leaving my shop - the other board did not receive calibration because Harald claimed it early in the bring-up process, before I had climbed the calibration learning curve and developed my calibration programs. Looking at the photo in issue #2704:
https://osmocom.org/issues/2704
The board you've got set up there is the one that did get properly calibrated before being sold to Sysmocom. Like all boards from that batch, it was shipped with an official FreeCalypso firmware image in its flash - based on the date it was shipped, it was probably the 20170831 build. Do you still have that firmware image in the flash, or have you erased it? Where is the RF from that board connected to - does it go to a sysmoBTS unit? Would it be possible for someone from Sysmocom (or someone from the community who have been set up with remote access to that board) to test and see if the FCDEV3B can successfully connect to Osmocom network using the official FreeCalypso firmware on the board, driven via AT commands?
If Sysmocom folks are philosophically against doing any tests with FreeCalypso firmware, would someone else in the community be more willing? I am hoping to have the next batch of my boards (which will also be at hardware revision V2) made in the next few months (more precisely, as soon as I can afford the ~3kUSD cost of the new PCB fab run for V2), and when I get those boards made, I plan on giving away a limited number of them on a subsidized basis, free of cost to the recipient. If anyone here is interested in pairing the source-enabled (source freely published for all to see and to tinker with) network side GSM implementation from Osmocom with the equally source-enabled mobile side implementation from FreeCalypso (which is much more robust and feature-complete than OsmocomBB), you may very well be eligible to receive a subsidized FCDEV3B V2 board when they become available.
Warm regards, Mychaela Falconia, Mother of FreeCalypso
P.S. Regarding Sysmocom's other FCDEV3B board which you got before it was calibrated, if you have a CMU200 instrument which is itself in good calibration standing, you can calibrate the board yourself. The needed software is here:
https://bitbucket.org/falconian/fc-rfcal-tools
Given your level of expertise with GSM RF, I trust that you should have absolutely no difficulty with figuring out how to use my calibration sw if you have the needed RF test equipment.
I tested using original firmware. On Fri, Sep 14, 2018 at 6:53 AM Mychaela Falconia < mychaela.falconia@gmail.com> wrote:
Sandi Suhendro djks74@gmail.com wrote:
Voice is working actually since back then, my fault is using only C117/118 for test call all the time. Now I just test using newer phone both Caller and Callee and voice is work perfectly! [...] There is trace in wireshark when using old phone as C117/118, the BTS always freeze and send Measurement Indication all the time when calling is made and answered, but with newer phone, the Indication is normal.
In your tests with C117/118 phones that produced failing results, were those phones running OsmocomBB or Motorola's original firmware? If Motorola phones running their original firmware (which presumably had passed official certification and type approval tests back in the day) don't work with Osmocom networks (be it LimeSDR or other Osmocom- compatible BTS hardware, SDR-based or otherwise) but "newer" (meaning more proprietary and more thoroughly locked down) phones do work, then it is a very grave problem from the perspective of Freedom, and needs to be taken very seriously.
Sysmocom has two of my FreeCalypso development boards, only one of which has been calibrated with a CMU200 instrument and fc-rfcal-tools before leaving my shop - the other board did not receive calibration because Harald claimed it early in the bring-up process, before I had climbed the calibration learning curve and developed my calibration programs. Looking at the photo in issue #2704:
https://osmocom.org/issues/2704
The board you've got set up there is the one that did get properly calibrated before being sold to Sysmocom. Like all boards from that batch, it was shipped with an official FreeCalypso firmware image in its flash - based on the date it was shipped, it was probably the 20170831 build. Do you still have that firmware image in the flash, or have you erased it? Where is the RF from that board connected to - does it go to a sysmoBTS unit? Would it be possible for someone from Sysmocom (or someone from the community who have been set up with remote access to that board) to test and see if the FCDEV3B can successfully connect to Osmocom network using the official FreeCalypso firmware on the board, driven via AT commands?
If Sysmocom folks are philosophically against doing any tests with FreeCalypso firmware, would someone else in the community be more willing? I am hoping to have the next batch of my boards (which will also be at hardware revision V2) made in the next few months (more precisely, as soon as I can afford the ~3kUSD cost of the new PCB fab run for V2), and when I get those boards made, I plan on giving away a limited number of them on a subsidized basis, free of cost to the recipient. If anyone here is interested in pairing the source-enabled (source freely published for all to see and to tinker with) network side GSM implementation from Osmocom with the equally source-enabled mobile side implementation from FreeCalypso (which is much more robust and feature-complete than OsmocomBB), you may very well be eligible to receive a subsidized FCDEV3B V2 board when they become available.
Warm regards, Mychaela Falconia, Mother of FreeCalypso
P.S. Regarding Sysmocom's other FCDEV3B board which you got before it was calibrated, if you have a CMU200 instrument which is itself in good calibration standing, you can calibrate the board yourself. The needed software is here:
https://bitbucket.org/falconian/fc-rfcal-tools
Given your level of expertise with GSM RF, I trust that you should have absolutely no difficulty with figuring out how to use my calibration sw if you have the needed RF test equipment.
Hi Mychaela,
On Thu, Sep 13, 2018 at 03:53:22PM -0800, Mychaela Falconia wrote:
Sysmocom has two of my FreeCalypso development boards,
To clarify: sysmocom != Harald.
I personally have bought one, which I have at home, and sysmocom has another one (at the sysmocom office/lab).
The board you've got set up there is the one that did get properly calibrated before being sold to Sysmocom. Like all boards from that batch, it was shipped with an official FreeCalypso firmware image in its flash - based on the date it was shipped, it was probably the 20170831 build. Do you still have that firmware image in the flash, or have you erased it?
Where is the RF from that board connected to - does it go to a sysmoBTS unit?
the way it looks, it has been connected to our osmo-gsm-tester setup, which basically is a coaxial RF distribution network to which a variety of BTSs (nanoBTS, sysmoBTS, USRP B2xx, UmTRX, ...) are connected, and where we run automatic continous tests with a set of modems against our latest "master" of the Osmocom stack.
Would it be possible for someone from Sysmocom (or someone from the community who have been set up with remote access to that board) to test and see if the FCDEV3B can successfully connect to Osmocom network using the official FreeCalypso firmware on the board, driven via AT commands?
This should all be possible. I doubt anyone has removed/erased the firmware on it.
In fact, I think that connecting the FCDEV to the osmo-gsm-tester setup might not be the best idea. It might be better to have a dedicated test setup with a single BTS and the FCDEV, to which we can then provide remote access. But let's have that discussion at https://osmocom.org/issues/2704
If Sysmocom folks are philosophically against doing any tests with FreeCalypso firmware,
I'm not sure where you would get such an idea from. We're happy to test whatever makes sense technically, and if there are any [suspected] interoperability problems with the Osmocom network-side stack, then we want to resolve those, no matter what the MS side implementation.
Hi Harald!
If Sysmocom folks are philosophically against doing any tests with FreeCalypso firmware,
I'm not sure where you would get such an idea from. We're happy to test whatever makes sense technically, and if there are any [suspected] interoperability problems with the Osmocom network-side stack, then we want to resolve those, no matter what the MS side implementation.
Good to hear. I was afraid that you would see my firmware as being untouchable for you because you are located in a country that deems it as being in violation of copyright. Now if you *are* willing to test FC firmware against Osmocom network-side gear, does that willingness extend only to the dated fw version already in the flash, or would you also be willing to test newer FC fw versions which you would need to download from ftp.freecalypso.org and flash yourself? Or would the latter be unacceptable to you because your country deems all of these FC firmwares to be in violation of copyright?
Just wondering what your stance is.
M~
-
Harald Welte -
Mychaela Falconia -
Sandi Suhendro