We want to avoid system clock dependencies in congestion
controllers as it makes it harder to test them. This CL removes
a rtc::TimeMillis() call from the AlrDetector class and removes
dependencies on rtc_base_approved as it exposes time_utils.h.
Bug: None
Change-Id: Ie50a27399c05a0c50cdc17ad142db884b94ee918
Reviewed-on: https://webrtc-review.googlesource.com/c/124491
Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
Reviewed-by: Christoffer Rodbro <crodbro@webrtc.org>
Commit-Queue: Sebastian Jansson <srte@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26879}
In a previous refactor, the ALR probe timestamp update was moved
after a return statement by accident. This CL fixes this.
The impact of this bug is limited as there are several other criteria
that has to be fulfilled for sending ALR probes.
Bug: None
Change-Id: Ia85e6ff9d782c1c4722a3df7e01ed803cf86b11d
Reviewed-on: https://webrtc-review.googlesource.com/c/124489
Reviewed-by: Erik Språng <sprang@webrtc.org>
Commit-Queue: Sebastian Jansson <srte@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26861}
There are two RTT values reported to GoogCC. They come from the same
source initially but one is calculated and smoothed in the video call stats.
However, there's not really any technical reasons why this value should
be received via the stats, this has just been maintained for legacy reasons.
Experiments shows no real difference between the modes, therefore the
stats-reported RTT is removed in this CL as a cleanup.
Bug: None
Change-Id: If1462d6c91570ffb883ecef2ba034f04a571c9b5
Reviewed-on: https://webrtc-review.googlesource.com/c/123883
Reviewed-by: Christoffer Rodbro <crodbro@webrtc.org>
Commit-Queue: Sebastian Jansson <srte@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26833}
Originally RtcEventProbeClusterCreated was logged in bitrate prober. This means that anyone who was using GoogCcNetworkControl wasn't logging it, and the NetworkControl wasn't self-contained.
This changes moves the responsibility for logging ProbeClusterCreated to ProbeController (where the probe is created), it also moves the responsibility for assigning probe ids to the probe controller.
Bug: None
Change-Id: If0433cc6d311b5483ea3980749b03ddbcd2bf041
Reviewed-on: https://webrtc-review.googlesource.com/c/122927
Commit-Queue: Peter Slatala <psla@webrtc.org>
Reviewed-by: Sebastian Jansson <srte@webrtc.org>
Reviewed-by: Björn Terelius <terelius@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26713}
This will be used to investigate the effect of congestion window
pushback on bandwidth esimation. There is currently no data available
in event logs to analyze this in test runs.
Bug: None
Change-Id: I2397842e90fd4acab6306b03d1ee9daf62469ee3
Reviewed-on: https://webrtc-review.googlesource.com/c/121765
Reviewed-by: Konrad Hofbauer <hofbauer@webrtc.org>
Reviewed-by: Sebastian Jansson <srte@webrtc.org>
Commit-Queue: Evan Shrubsole <eshr@google.com>
Cr-Commit-Position: refs/heads/master@{#26681}
With the current implementation, whenever we are toggling between
sending/not sending retransmissions, the BitrateAllocator will
toggle the total requested max bitrate that is signalled to the
probing mechanism. The result is that spurious probes are sent
mid-call, at |max_bitrate| and |2*max_bitrate|. This behaviour
is undesirable, and thus removed in this CL. Instead, whenever
the allocation limits actually change, we produce a single
set of probes at |max_bitrate| and |2*max_bitrate|.
This CL does not change how the BitrateAllocator hysteresis is
accounting for protection, since it does not relate to the
spurious probes.
Bug: webrtc:10275
TBR: sprang@webrtc.org
Change-Id: Iab3a690a500372c74772a8ad6217fb838af15ade
Reviewed-on: https://webrtc-review.googlesource.com/c/120808
Commit-Queue: Rasmus Brandt <brandtr@webrtc.org>
Reviewed-by: Sebastian Jansson <srte@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26544}
Will focus on delivering model based controller instead.
Bug: webrtc:9718
Change-Id: I5df82424469c577f3c170758e0db64e3e1aa7705
Reviewed-on: https://webrtc-review.googlesource.com/c/120607
Reviewed-by: Christoffer Rodbro <crodbro@webrtc.org>
Commit-Queue: Sebastian Jansson <srte@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26478}
It didn't have proper time to stabilize, making it sensitive to small
changes. This CL increases the stabilization period from 20 to 30s.
Also fixing some minor test suite bug found during investigation.
Bug: webrtc:9718
Change-Id: If56dba5383251ad3d3efe304eebcd880522afabe
Reviewed-on: https://webrtc-review.googlesource.com/c/119943
Reviewed-by: Christoffer Rodbro <crodbro@webrtc.org>
Commit-Queue: Sebastian Jansson <srte@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26408}
This class collects the field trial based configuration of audio
allocation and bandwidth in one place. This makes it easier
overview and prepares for future cleanup of the trials.
Bug: webrtc:9718
Change-Id: I34a441c0165b423f1e2ee63894337484684146ac
Reviewed-on: https://webrtc-review.googlesource.com/c/118282
Commit-Queue: Sebastian Jansson <srte@webrtc.org>
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Reviewed-by: Oskar Sundbom <ossu@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26370}
By enabling this trial, we can also remove reporting of packet
feedback status from send streams that was used before.
Bug: webrtc:9718
Change-Id: I3e7c4656b0ac6592a834617e044f23a072454181
Reviewed-on: https://webrtc-review.googlesource.com/c/118281
Reviewed-by: Oskar Sundbom <ossu@webrtc.org>
Reviewed-by: Erik Språng <sprang@webrtc.org>
Reviewed-by: Christoffer Rodbro <crodbro@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26363}
This cl is part of work to move several experiments into a joint
experiment group. Most of them vill be ralted to video, hence the name.
Bug: webrtc:10223
Change-Id: I8767c43abb6aa910ab51710eeb908e0f9df1e296
Reviewed-on: https://webrtc-review.googlesource.com/c/118361
Commit-Queue: Erik Språng <sprang@webrtc.org>
Reviewed-by: Sebastian Jansson <srte@webrtc.org>
Reviewed-by: Stefan Holmer <stefan@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26346}
Fix has 2 parts:
1. Fix for the LossBasedControl being at much lower levels than
DelayBased in StartUpPhase.
2. Explicitly fix state machine problem leading to toggling between
the two estimates.
Bug: webrtc:10222
Change-Id: Ieaaaec6c9233da61a86b69d936c4979c79645686
Reviewed-on: https://webrtc-review.googlesource.com/c/118280
Commit-Queue: Christoffer Rodbro <crodbro@webrtc.org>
Reviewed-by: Sebastian Jansson <srte@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26327}
The WebRtcKeyValueConfig interface allows providing custom key value
configurations that changes per instance of GoogCcNetworkController.
Bug: webrtc:10009
Change-Id: I520fff030d1c3c755455ec8f67896fe8a6b4d970
Reviewed-on: https://webrtc-review.googlesource.com/c/116989
Commit-Queue: Sebastian Jansson <srte@webrtc.org>
Reviewed-by: Per Kjellander <perkj@webrtc.org>
Reviewed-by: Björn Terelius <terelius@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26312}
This makes it possible to save log outputs from scenario tests to
either files or memory.
Bug: webrtc:9510
Change-Id: I883bd8240ab712d31d54118adf979041bd83481a
Reviewed-on: https://webrtc-review.googlesource.com/c/116321
Commit-Queue: Sebastian Jansson <srte@webrtc.org>
Reviewed-by: Per Kjellander <perkj@webrtc.org>
Reviewed-by: Niels Moller <nisse@webrtc.org>
Reviewed-by: Björn Terelius <terelius@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26284}
That means it does not have to be set on every update of StreamsConfig.
BUG=webrtc:9586
Change-Id: I6a348160e209042857c4475323466e2aa92adef8
Reviewed-on: https://webrtc-review.googlesource.com/c/116690
Reviewed-by: Sebastian Jansson <srte@webrtc.org>
Commit-Queue: Per Kjellander <perkj@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26184}
While this class is deprecated, it's needed as a stop-gap solution.
Other methods to configure the max probe rate all effect the current
estimate and/or trigger new probes to be sent, and we need a way to
configure the max without affecting other behavior.
Bug: webrtc:10070
Change-Id: I2b0ba2fef42d0bab6e5ea7f7c921681557802b4b
Reviewed-on: https://webrtc-review.googlesource.com/c/114880
Reviewed-by: Sebastian Jansson <srte@webrtc.org>
Commit-Queue: Erik Språng <sprang@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26162}
The new controller behaves mostly like before, but increases the target
rate on timer update rather than when feedback is received. This makes
the behavior easier to predict. It also uses a duration parameter to
track the increase, removing the meed for the minimum rate increase
constants that exists in the previous solution.
Bug: webrtc:9718
Change-Id: Iae31a9ba2d6474a8236f8eb72f86ff434f1d1fc6
Reviewed-on: https://webrtc-review.googlesource.com/c/114681
Commit-Queue: Sebastian Jansson <srte@webrtc.org>
Reviewed-by: Christoffer Rodbro <crodbro@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26088}
If we're in ALR, the acked rate is going to be significantly lower than
the current estimate for the link capacity. If we need to back off in
this situation (usually caused by latency spikes), this CL makes us back
off relative to current estimate if. We then immediately send a new
probe just in case the network did actually change.
All of this is behind experiment flags for now.
Bug: webrtc:10144
Change-Id: I062a259c36417eea2211d44592ef7fc979aa22b7
Reviewed-on: https://webrtc-review.googlesource.com/c/113880
Commit-Queue: Erik Språng <sprang@webrtc.org>
Reviewed-by: Sebastian Jansson <srte@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26045}
The removed coded causes problems if the same RTCP packet is forwarded
to the congestion controller multiple times.
Bug: webrtc:10125
Change-Id: I659d8f8f3ce3c643710156fa81176ceeaedd714a
Reviewed-on: https://webrtc-review.googlesource.com/c/114165
Commit-Queue: Björn Terelius <terelius@webrtc.org>
Reviewed-by: Sebastian Jansson <srte@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26016}
Some tests had to be updated due to this change.
Bug: webrtc:9510
Change-Id: I79c4c0166d8ba5e8190a607d5d35b67dc30a3c14
Reviewed-on: https://webrtc-review.googlesource.com/c/113522
Commit-Queue: Sebastian Jansson <srte@webrtc.org>
Reviewed-by: Christoffer Rodbro <crodbro@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#25952}
The limit we put on probing is a bit too conservative now. If an
allocation limit is set, this CL allows probing up to 2x the current
max allocation limit.
This better handles overshooting when networks actually have the
capacity to allow bursts.
Bug: webrtc:10070
Change-Id: I0003f6b22512c13b6a83c1934952a2c3a2b70b48
Reviewed-on: https://webrtc-review.googlesource.com/c/112905
Commit-Queue: Erik Språng <sprang@webrtc.org>
Reviewed-by: Sebastian Jansson <srte@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#25888}
Conversion to kbps will fail if the estimate is lower than the deviation
estimate * 3, since that would produce a negative value.
Bug: webrtc:9718
Change-Id: I83b52acd476d90b1f22c9db9894fa26c9a3e8e17
Reviewed-on: https://webrtc-review.googlesource.com/c/112560
Reviewed-by: Niels Moller <nisse@webrtc.org>
Reviewed-by: Sebastian Jansson <srte@webrtc.org>
Commit-Queue: Sebastian Jansson <srte@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#25854}
This prepares for future refactoring of rate controller.
Bug: webrtc:9718
Change-Id: I425c8c547399bda98b4271a0d24a0bb7ee06bc13
Reviewed-on: https://webrtc-review.googlesource.com/c/112420
Commit-Queue: Sebastian Jansson <srte@webrtc.org>
Reviewed-by: Björn Terelius <terelius@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#25846}
The intention is to provide a bandwidth estimate that only updates if
the actual available bandwidth is known to have changed. This will be
used in media streams to avoid changing the configuration (such as
frame size, audio frame length etc), just because the control target
rate changed.
Bug: webrtc:9718
Change-Id: I17ba5a2f9e5bd408a71f89c690d45541655a68e2
Reviewed-on: https://webrtc-review.googlesource.com/c/107726
Commit-Queue: Sebastian Jansson <srte@webrtc.org>
Reviewed-by: Björn Terelius <terelius@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#25772}
This prepares for providing an additional implementation of delay based
rate control. By moving the probe controller, less code will have to be
added in the upcoming CL.
Bug: webrtc:9718
Change-Id: I64eb2c8f5f7950b6e9d209f110dc0a757c710b4b
Reviewed-on: https://webrtc-review.googlesource.com/c/111860
Commit-Queue: Sebastian Jansson <srte@webrtc.org>
Reviewed-by: Christoffer Rodbro <crodbro@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#25770}
This CL decouples //rtc_base:rtc_base_tests_utils from gunit by
moving gunit helpers (rtc_base/gunit.h) and rtc_base/testclient.h
(which depends on gunit helpers) to their own build target.
It also removes some unused dependencies in the WebRTC build graph.
Bug: None
Change-Id: Ia9820e84ff697da39b351eef73c45f6e4bdf2623
Reviewed-on: https://webrtc-review.googlesource.com/c/111861
Commit-Queue: Mirko Bonadei <mbonadei@webrtc.org>
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#25769}
- Rename avg_max_bitrate_kbps to link_capacity_estimate_kbps and change
the type to optional.
- Remove the RateControlRegion enum. The old code seems to have the invariant
that the region is kRcMaxUnknown iff avg_max_bitrate_kbps is uninitialized.
- Change floats to double.
Bug: webrtc:9942
Change-Id: Ic071a11ec4950053ec92beaa06f28f43192521d7
Reviewed-on: https://webrtc-review.googlesource.com/c/111247
Commit-Queue: Björn Terelius <terelius@webrtc.org>
Reviewed-by: Stefan Holmer <stefan@webrtc.org>
Reviewed-by: Sebastian Jansson <srte@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#25752}
