The something3 points to the next start of the SDP
entry. The four bytes in front of the " SDP" are not
known and just discarded. Prepare to be able to
recursively parse the SDP header...
I must have picked in the wrong section of these
files... There are some kind of header entries
that are all 138 byte long and this is the total
length...
In many cases we actually want a name / unique ID for the lchan,
not just for the on-air timeslot... especially in SDCCH/8 case,
where 8 SDCCHs share one timeslot...
When we allocate a channel, we send the RSL CHAN ACT REQ and wait until we get
a CHAN ACT ACK. Only the ACK will change the state, so there is a race where
we allocate that same channel to a different channel request before we get
the ACT ACK.
Introducing a new ACT_REQ state resolves this issue.
When we have received the End Ack we are just doing nothing as we
are done. This means rc remains -1 and we will print a warning but
there is no need to have a warning...
* The FOM header needs to be different. We need to address the
base station transceiver, bts, trx set to 0 and ts to 255
* We need to transfer the the \0 of 'id' and 'version'
* We need to issue a NM_ATT_SW_DESCR (just the value)
* We need to use 16bit length for the other two ids..
* After this our Software Load Init is getting an Ack.
Strictly speaking we would only need to start the Site Manager
and could probably start flashing afterwards but it is more easy
to have one config path...
This will mostly work like the downloading in bs11_config
and is based on the bs11_config state machine as well. Once
it is working we can see how to unite both implementations.
See GSM 04.11 Chapter 5.4 for details. The idea is that when
multi-SMS are mobile originated, it's possible the CP-ACK of
the previous transaction to be lost and the reception of a
new CP-DATA for a new transaction should close previous transaction
"as-if" we had received the CP-ACK ...
Note that testing is hard since it's an exceptional condition that's
hard to create. I tested by temporarly disabling CP-ACK processing
and checked it worked as expected.
Signed-off-by: Sylvain Munaut <tnt@246tNt.com>
The current code used the variable bitmap format, but
that's not possible since in this format the base ARFCN is
part of the set. That lead to a neighbor list containing ARFCN 0.
Signed-off-by: Sylvain Munaut <tnt@246tNt.com>
We need transaction_id to be a int (as returned by trans_assign_trans_id)
to detect the error condition -1.
Signed-off-by: Sylvain Munaut <tnt@246tNt.com>
The idea is to find the highest used id and try to get the
next. This way when there are transactions back to back with
an overlap, we go 0 1 2 3 4 5 6 0 1 2 ...
Signed-off-by: Sylvain Munaut <tnt@246tNt.com>
This marks the departure from printing all the debug messages to the console by
default. We only print NOTICE and WARNING level messages by default
If you're interested in more details, you need to enable it via command
line options or the VTY
The VTY code makes so many allocations that a full report is
simply too long to provide any useful information. So we sub-divide
it in multiple contexts, and report only one level deep at SIGURS1.
We also introduce SIGUSR2 for the full detailed VTY report.
In case we have multiple TRX configured, but not all of them are
actually active/operational, we should not try to allocate channels
from such transceivers.
This proxy allows us to restart OpenBSC while the BTS's are kept
running with their CCCH/BCCH alive. This is very useful to
make sure the phones don't roam to other networks while restarting
OpenBSC.
The proxy also intrduces UDP sockets for injecting UDP packets
into the A-bis data stream.
So far I have not much idea about the format. It is starting with
the magic byte and the header is spanning until the next occurence
of the " SDP" marker.
The magic " SDP" is occuring twice in the file. The first time
seems to be the file header and the second time it is with the
payload. We will need to parse this somehow...
* The two version strings are not in an easy to parse header
and from my trace it appears like the whole file is sent
to the BTS. So just open the firmware file..
Change the counters_store_db function to be a generic for_each
function taking a function pointer and data. Use that in bsc_hack
to store it to the DB.
This is removing the DB requirement and will allow to handle
the counter values in different ways without making the counter
list public.
I verified that the syncing is still taking place.
This is the new logging architecture, including
* support for multiuple logging targets like stderr and vty
* log levels in addition to categories/subsystems
* filtering based on imsi, i.e. only see events for one subscriber
* dynamically change log level for each category for each vty
This has the advantage that counters can be added all over the code
very easily, while having only one routine that stores all of the
current counter values to the database. The counters are synced
every 60 seconds, providing relatively fine grained statistics
about the network usage as time passes by.
Tweaking theses can be useful especially tx-integer that influence
both the spread of rach attemps and the delay between two attemps.
Looking up GSM 04.08 3.3.1.1.2 & 10.5.2.29 can help determine good
values. The default are choosed with a wide spacing between attemps
(tx integer = 9 -> T=12 & S=217 (non-combined CCCH/SDCCH) or 115 (for
combined CCCH/SDCCH)). This alleviates the problem of responding to
several RACH attempts by a same MS, allocating several RF channels when
only 1 is needed.
Signed-off-by: Sylvain Munaut <tnt@246tNt.com>
The previous implementation had some shortcomings:
- If the MIN ID given was not the exact id of the first unsent SMS,
it would try to submit the same sms several time until id++ finally
made id go to the next one.
- If a subscriber had several SMS pending it would try to submit
them individually (only to get rejected because a paging for that
subscriber was already in progress)
Signed-off-by: Sylvain Munaut <tnt@246tNt.com>
we need to set newbfd->priv_nr to 2+trx_id, rather than keeping
it '2' all the time, as it is used to look-up the e1i_ts when
we receive a packet. A constant '2' would always match to TRX 0.
we also need to keep one separate bit for each TRX state in order
to properly generate the EVT_E1_TEI_UP event for trx > 0.
This implements the handover algorithm (and associated parameters)
as described in Chapter 8 of the book "Performance Enhancements in
a Frequency |Hopping GSM Network" by Thomas Toftegard Nielsen and Jeroen
Wigard.
The parameters such as averaging windows are configured in struct
gsm_network. We keep some state to trakc up to 10 neighbors as
they are being reported from the MS.
This has so far only been tested in a network with two BTS that
have each other as neighbor. Networks with morge neighbors might
encounter bugs.
This allows us to block packets that we have received after the channel
is no longer being used. This is visible during handover, where we still
receive a measurement report after the MS has switched to the new channel.
This leftover measurement report then attempts to trigger another handover,
which si bogus and will fail - and thus only consumes resources.
With the new LCHAN_S_ACTIVE state, we can check for this when processing
the measurement report.
This provides two functions: get_meas_rep_avg() to obtain the sliding
window average of one particular field, and meas_rep_n_out_of_m_be()
to check if at least N out of M measurments are >= BE.
If a RF channel is assigned but no response is ever heard from
the phone, we will receive a CONNECTION FAIL from the BTS,
triggering a RF release freeing the channel. Then sometime later,
T3101 will expire as well and free the channel again ...
Signed-off-by: Sylvain Munaut <tnt@246tNt.com>
- Need to use sms.id for the ORDER BY since 'subscriber' also has 'id'
- Need to add the join clause between 'SMS' and 'subscriber'
- Add a LIMIT 1 (probably no impact for the db size we're dealing with
here, but with large DB and mysql/postgresql this can help the planner)
- (fix a wrong comment in passing ...)
Signed-off-by: Sylvain Munaut <tnt@246tNt.com>
This patch takes care of handling the RTP streams / sockets during
an in-call handover from one BTS to another BTS.
It only works in combination with rtp_proxy mode.
This is not really nice, but we will soon have multiple users of
the CRCX / MDCX / DLCX signals, and we cannot guarantee the ordering
of them. So as a workaround, we move the RTP socket creation and
deletion into the core abis_rsl codebase.
We cannot support in-call handover of calls without a RTP proxy,
since at the time of the handover the SSRC, sequence number and
timestamp of the RTP frames change.
Since the MNCC API can now send and receive frames to/from the MNCC
application, we can also implement a proxy this way. Not at the RTP/UDP packet
level, but at the 'TCH speech frame' level.
Especially for handover, we need this mode as the receiver in the BTS needs a
persistent SSRC and monotonic frame numbers / timestamps.
During handover, we will not send RTP frames for quite some time. However,
the way the rtp_send code is structured, it will increment the timestamp
with a fixed amount every time we send a frame, independent how much wallclock
time has actually passed.
This code is a hack to update the sequence number and timestamp in case it
seems to be wrong. It makes handover much more reliable.
Instead of passing TRAU frames down the MNCC API to the call control
application like MNCC, we now decode the TRAU frame into the actual codec
frame. We do the same with the RTP packets in case of ip.access and
thus have a unified format of passing codec data from the BTS to
an application, independent of the BTS type.
This is only implemented for V1 full-rate at the moment, and needs
to be fixed.
It seems that depending on the manufacturer, there is a need to include
the L2 pseudo-length in the SI5+SI6 messasges (SACCH FILLING)
Thanks to Dieter for pointing this out.
Multiple CM SERVICE REQUEST can happen on a single RR
connection, in this case, since the subscr reference is
tracked through lchan->subscr and will only be put'd once
on lchan_free, we need to make sure we don't get several
reference ....
Signed-off-by: Sylvain Munaut <tnt@246tNt.com>
'unknown' has a negative connotation for a case that's totally
normal so refer to it as 'new' so it doesn't sound like a problem.
Signed-off-by: Sylvain Munaut <tnt@246tNt.com>
This introduces a new LOGP() macro together with LOGL_* definition to
support multiple log levels (severities) throughout the codebase.
Please note that the actual logging system does not use them yet,
in this patch we simply introduce the new macros at the caller site.
With this commit, we can successfully hand over a channel from one cell to
another cell. We implement asynchronous intra-BSC (but inter-BTS) handover.
Changes:
* introduce new DHO log category
* extend rsl_chan_activate_lchan() with argument for HO reference
* introduce actual minimal handover decision making in handover_decision.c
* various fixes to bsc_handover_start() in handover_logic.c
We will need this for the actual handover algorithm implementation, as we will
only know the current BTS and the BCCH ARFCN of the strongest cell in the
measurement reports. Using this new function, we can resolve the matching
gsm_bts.
Since we are keeping a bitvec of the neighbor cells, we can now use
bitvec_get_nth_set_bit() to determine the ARFCN for each reported
cell in the 04.08 MEASUREMENT REPORT.
We use a 1024-bit-sized bitvec to generate the BA and neighbor frequency list.
This bitvec is still generated from the list of all BTS's inside the BSC, but
this patch is the first step to generalize this, i.e. generate arbitrary
neighbor lists.
* introduce a new bitvec_get_bit_pos() function to determine the bit value
at a given position inside a bit vector
* make sure bitvec_{get,set}_bit_pos() share code as possible
In some cases, we can try to close the lchan a bit faster than our
lchan release timeout:
* After we've sent LOC UPD ACCEPT and MM INFO
* After a phone has confirmed the RELEASE of a call
The rationale is as following:
If we have NECI=1, then the phone will request a channel with CHREQ
"0100xxxx Originating speech call from dual-rate mobile station when TCH/H is
sufficient and supported by the MS for speech calls", then we will try to
allocate a TCH/H [as it is sufficient].
However, if there are no free TCH/H slots on the BTS, we abandon and can't
handle the MO call at all :(
We might read or write on the OML link when the BTS is
reset and will get SIGPIPE interrupt and be gone... Just
ignore the SIGPIPE we will get the "exception" on the next
run of bsc_select and kill the (old) OML link.
With ip.access, in case of TCH/H, we have one RTP stream for each half-slot
(lchan), not just one per on-air timeslot. This is quite different from
a classic BTS where the TRAU frames of the two TCH/H channels would be
part of the same 16k sub-slot in a E1 timeslot.
Before this commit, OpenBSC used templates for the SYSTEM INFO
1, 2, 3, 4, 5 and 6 messages. Those templates were patched in
various places to reflect the network config like ARFCN.
Now, we actually generate those SI messages ourselves, using
values from the configuration file, and even calculating neighbor
cell lists.
All bts'es that you have configured in OpenBSC will end up in
the neighbor cell list - which should be more than sufficient for
the current small-single-site networks.
Code to implement handover control logic. A yet-to-be-implemented
handover algorithm will call bsc_handover_start(old_lchan, new_bts)
to start the handover process.
This introduces the signals S_LCHAN_ACTIVATE_{ACK,NACK} and
S_LCAN_HANDOVER_{FAIL,COMPL,DETECT} as well as code that actually issues
those signals. The signals are relevant for a yet-to-be-written handover
control logic.
This is needed by a yet-to-be-implemented handover algorithm, after
it has allocated a new lchan for the MS. Also missing: handling
the actual HANDOVER COMPLETE / FAIL messages in response.
This patch extends struct gsm_meas_rep into a complete structure containing all
information from both uplink and downlink measurement results/reports.
This is a first step to provide this complete measurement data as a C structure
into a to-be-implemented handover decision algorithm.
- Make sure that on runtime the Radio Carrier can be
locked and unlocked. The vty code calls into the
Abis NM to lock/unlock the channel and the state is
stored there.
- Make sure that on start the Radio Carries remains
offline and we are not starting it. On start the
radio carrier is either locked or unlocked. This means
the RSL will not connect until the RF is unlocked. It
will connect then. To see RSL bringup failures one
needs to parse the RSL nack message.
- When the TRX is locked on startup the RSL link will
only be established after it will be unlocked.
On cold start the RSL link will not be brought up. Wait for the
Software to be Activated before starting the RSL link. This is
working reliable on the BTS I have tested with.
This is a partial revert of 8406ec2437 and
was discussed on the mailinglist.
In file included from sms_test.c:27:
../../include/openbsc/gsm_utils.h:33: warning: `enum gsm_band' declared inside parameter list
../../include/openbsc/gsm_utils.h:33: warning: its scope is only this definition or declaration, which is probably not what you want
../../include/openbsc/gsm_utils.h:34: warning: `enum gsm_band' declared inside parameter list
Holger Hans Peter Freyther
Harald Welte
Sylvain Munaut