This is merely a historical archive of years 2008-2021, before the migration to mailman3.
A maintained and still updated list archive can be found at https://lists.osmocom.org/hyperkitty/list/gerrit-log@lists.osmocom.org/.
Harald Welte gerrit-no-reply at lists.osmocom.orgHarald Welte has submitted this change and it was merged. ( https://gerrit.osmocom.org/c/osmo-bts/+/14334 ) Change subject: trx: Update documentation for fn-advance and rts-advance ...................................................................... trx: Update documentation for fn-advance and rts-advance Change-Id: I83eff296bfb7acff9d8560b5346bd9791fe6e7d9 Related: OS#4036 --- M doc/manuals/chapters/bts-models.adoc 1 file changed, 45 insertions(+), 2 deletions(-) Approvals: Harald Welte: Looks good to me, approved Jenkins Builder: Verified diff --git a/doc/manuals/chapters/bts-models.adoc b/doc/manuals/chapters/bts-models.adoc index 4ae73c0..e1f8d9a 100644 --- a/doc/manuals/chapters/bts-models.adoc +++ b/doc/manuals/chapters/bts-models.adoc @@ -245,10 +245,53 @@ Set the number of frames to be transmitted to transceiver in advance of current GSM frame number. +GSM is a TDMA (time division multiple access) system on the radio +interface. OsmoTRX is the "clock master" of that in the Osmocom +implementation. It informs OsmoBTS of the current GSM frame +number. However, as there is non-zero delays (UDP packet trnsmission +delay, operating system scheduler delay on both OsmoTRX and OsmoBTS +side, ...), OsmoBTS must compensate for that delay by "advancing" +the clock a certain amount of time. + +In other words, if OsmoTRX informs us that the current frame number is N, +we advance it by `fn-advance` and transmit burst data for +`N + fn-advance` towards OsmoTRX. + +The fn-advance should be kept as low as possible to avoid additional +delays to the user voice plane as well as to improve the performance +of the control plane (LAPDm) as well as GPRS. + +However, fn-advance must be kept sufficiently high to ensure no +underruns on the OsmoTRX side. + +The detailed value will depend on your underlying computer systems, +operating system and related tuning parameters. Running OsmoTRX +on a remote host will inevitably require a higher fn-advance then +running it on the same machine, where the UDP packetes are just passed +over the loopback device. + +The default value for `fn-advance` is 20 (corresponding to 92 +milliseconds). + ===== `osmotrx rts-advance <0-30>` -Set the number of frames to be requested from PCU in advance of current -frame number. Do not change this unless you have a good reason! +Set the number of frames to be requested from L1SAP in advance of current +frame number and fn-advance. + +The value specified as `rts-advance` is added to the current GSM frame +number as reported by OsmoTRX *and* the `osmotrx fn-advance` in order +to generate the PH-RTS.ind (ready to send indications) across the L1SAP +interface inside osmo-bts. This will trigger the Layer 2 (LAPDm for +the ocntrol plane, RTP for the voice plane, and OsmoPCU for GPRS) to +generate a MAC block and input it into the osmo-bts-trx TDMA scheduler. + +If OsmoTRX reported N as the current frame number, the actual frame number +reported on L1SAP to higher layes will be computed as follows: + + N + fn-advance + rts-advance + +The default value of `rts-advance` is 5 (corresponding to 23 milliseconds). +Do not change this unless you have a good reason! ===== `osmotrx rx-gain <0-50>` -- To view, visit https://gerrit.osmocom.org/c/osmo-bts/+/14334 To unsubscribe, or for help writing mail filters, visit https://gerrit.osmocom.org/settings Gerrit-Project: osmo-bts Gerrit-Branch: master Gerrit-Change-Id: I83eff296bfb7acff9d8560b5346bd9791fe6e7d9 Gerrit-Change-Number: 14334 Gerrit-PatchSet: 3 Gerrit-Owner: Harald Welte <laforge at gnumonks.org> Gerrit-Reviewer: Harald Welte <laforge at gnumonks.org> Gerrit-Reviewer: Jenkins Builder Gerrit-Reviewer: fixeria <axilirator at gmail.com> Gerrit-MessageType: merged -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.osmocom.org/pipermail/gerrit-log/attachments/20190603/425b0217/attachment.htm>