Added implementations for entering/exiting STARTUP, DRAIN, PROBE_BW, PROBE_RTT modes, also updated M

Added implementations for entering/exiting STARTUP, DRAIN, PROBE_BW, PROBE_RTT modes, also updated MaxBandwidthFilter class, with the filter implementation which stores three best estimates for the filter window.

BUG=webrtc:7713
Review-Url: https://codereview.webrtc.org/2982233002
Cr-Commit-Position: refs/heads/master@{#19196}
Committed: https://chromium.googlesource.com/external/webrtc/+/d339dbc7d4758e62b02740feda85c82dd6b2eba5

[gnish1, 2017-07-19 15:21:10]

Description was changed from

==========
Merge branch 'master' of https://chromium.googlesource.com/external/webrtc into
bla


Disable flaky NoBandwidthDropAfterDtx test.

BUG=chromium:744695

Review-Url: https://codereview.webrtc.org/2978323002
Cr-Commit-Position: refs/heads/master@{#19092}

Roll chromium_revision 6ada1228ca..8779e5365c (487809:487834)

Change log:
https://chromium.googlesource.com/chromium/src/+log/6ada1228ca..8779e5365c
Full diff:
https://chromium.googlesource.com/chromium/src/+/6ada1228ca..8779e5365c

Changed dependencies:
* src/base:
https://chromium.googlesource.com/chromium/src/base/+log/acd6f21afd..301d8613c5
* src/build:
https://chromium.googlesource.com/chromium/src/build/+log/ddbb32b392..53c0c1a75f
* src/ios:
https://chromium.googlesource.com/chromium/src/ios/+log/0f76340cb8..e9391264b9
* src/third_party:
https://chromium.googlesource.com/chromium/src/third_party/+log/2087aacb81..f...
* src/tools:
https://chromium.googlesource.com/chromium/src/tools/+log/2aab8e9626..9f8398f59c
DEPS diff:
https://chromium.googlesource.com/chromium/src/+/6ada1228ca..8779e5365c/DEPS

No update to Clang.

TBR=
BUG=None
CQ_INCLUDE_TRYBOTS=master.internal.tryserver.corp.webrtc:linux_internal

Review-Url: https://codereview.webrtc.org/2984623002
Cr-Commit-Position: refs/heads/master@{#19091}

Logic for modes, new bw filter.


nothing.


nothing.
==========

to

==========
BUG=webrtc:7713

Review-Url: https://codereview.webrtc.org/2978323002
Cr-Commit-Position: refs/heads/master@{#19092}

Roll chromium_revision 6ada1228ca..8779e5365c (487809:487834)

Change log:
https://chromium.googlesource.com/chromium/src/+log/6ada1228ca..8779e5365c
Full diff:
https://chromium.googlesource.com/chromium/src/+/6ada1228ca..8779e5365c

Changed dependencies:
* src/base:
https://chromium.googlesource.com/chromium/src/base/+log/acd6f21afd..301d8613c5
* src/build:
https://chromium.googlesource.com/chromium/src/build/+log/ddbb32b392..53c0c1a75f
* src/ios:
https://chromium.googlesource.com/chromium/src/ios/+log/0f76340cb8..e9391264b9
* src/third_party:
https://chromium.googlesource.com/chromium/src/third_party/+log/2087aacb81..f...
* src/tools:
https://chromium.googlesource.com/chromium/src/tools/+log/2aab8e9626..9f8398f59c
DEPS diff:
https://chromium.googlesource.com/chromium/src/+/6ada1228ca..8779e5365c/DEPS

No update to Clang.

TBR=
BUG=None
CQ_INCLUDE_TRYBOTS=master.internal.tryserver.corp.webrtc:linux_internal

Review-Url: https://codereview.webrtc.org/2984623002
Cr-Commit-Position: refs/heads/master@{#19091}

Logic for modes, new bw filter.


nothing.


nothing.
==========

[gnish1, 2017-07-19 15:22:07]

gnish@webrtc.org changed reviewers:
+ philipel@webrtc.org, stefan@webrtc.org, terelius@webrtc.org

[terelius, 2017-07-19 16:31:23]

CL descriptions seems to be incorrect. Please verify that the CL contains
exactly the changes you intended to upload, and update the description.

[gnish1, 2017-07-20 07:45:53]

Description was changed from

==========
BUG=webrtc:7713

Review-Url: https://codereview.webrtc.org/2978323002
Cr-Commit-Position: refs/heads/master@{#19092}

Roll chromium_revision 6ada1228ca..8779e5365c (487809:487834)

Change log:
https://chromium.googlesource.com/chromium/src/+log/6ada1228ca..8779e5365c
Full diff:
https://chromium.googlesource.com/chromium/src/+/6ada1228ca..8779e5365c

Changed dependencies:
* src/base:
https://chromium.googlesource.com/chromium/src/base/+log/acd6f21afd..301d8613c5
* src/build:
https://chromium.googlesource.com/chromium/src/build/+log/ddbb32b392..53c0c1a75f
* src/ios:
https://chromium.googlesource.com/chromium/src/ios/+log/0f76340cb8..e9391264b9
* src/third_party:
https://chromium.googlesource.com/chromium/src/third_party/+log/2087aacb81..f...
* src/tools:
https://chromium.googlesource.com/chromium/src/tools/+log/2aab8e9626..9f8398f59c
DEPS diff:
https://chromium.googlesource.com/chromium/src/+/6ada1228ca..8779e5365c/DEPS

No update to Clang.

TBR=
BUG=None
CQ_INCLUDE_TRYBOTS=master.internal.tryserver.corp.webrtc:linux_internal

Review-Url: https://codereview.webrtc.org/2984623002
Cr-Commit-Position: refs/heads/master@{#19091}

Logic for modes, new bw filter.


nothing.


nothing.
==========

to

==========
Added implementations for entering/exiting all modes, also updated
MaxBandwidthFilter class, where newer implementation for bandwidth filter is
added.

BUG=webrtc:7713
==========

[philipel, 2017-07-26 11:37:17]

https://codereview.webrtc.org/2982233002/diff/1/webrtc/modules/remote_bitrate...
File webrtc/modules/remote_bitrate_estimator/test/estimators/bbr.cc (right):

https://codereview.webrtc.org/2982233002/diff/1/webrtc/modules/remote_bitrate...
webrtc/modules/remote_bitrate_estimator/test/estimators/bbr.cc:38: const int
kMinimumCongestionWindow = 4000;
Is this in bytes? Why are there two kMinimumCongestionWindow (the other one
being in congestion_window.cc) and why are they of different sizes?

https://codereview.webrtc.org/2982233002/diff/1/webrtc/modules/remote_bitrate...
webrtc/modules/remote_bitrate_estimator/test/estimators/bbr.cc:39: const int
kProbeRttDurationRounds = 1;
Comment what this value represent.

https://codereview.webrtc.org/2982233002/diff/1/webrtc/modules/remote_bitrate...
webrtc/modules/remote_bitrate_estimator/test/estimators/bbr.cc:40: const int
kProbeRttDurationMs = 200;
Comment what this value represent.

https://codereview.webrtc.org/2982233002/diff/1/webrtc/modules/remote_bitrate...
webrtc/modules/remote_bitrate_estimator/test/estimators/bbr.cc:137: if
(congestion_window_->data_inflight() <= TargetCongestionWindow(1))
Create a variable for 1 which describes what it is.

https://codereview.webrtc.org/2982233002/diff/1/webrtc/modules/remote_bitrate...
webrtc/modules/remote_bitrate_estimator/test/estimators/bbr.cc:146:
congestion_window_gain_ = 1.5f;
Create a variable for 1.5 which describes what it is.

https://codereview.webrtc.org/2982233002/diff/1/webrtc/modules/remote_bitrate...
webrtc/modules/remote_bitrate_estimator/test/estimators/bbr.cc:147: unsigned int
seed = time(NULL);
Use webrtc::Random instead (in rc_base/random.h).

https://codereview.webrtc.org/2982233002/diff/1/webrtc/modules/remote_bitrate...
webrtc/modules/remote_bitrate_estimator/test/estimators/bbr.cc:182:
minimum_congestion_window_start_time_ms_ = -1;
I think using an rtc::Optional might be cleaner in this case.

https://codereview.webrtc.org/2982233002/diff/1/webrtc/modules/remote_bitrate...
webrtc/modules/remote_bitrate_estimator/test/estimators/bbr.cc:191: bool
min_rtt_expired) {
|min_rtt_expired| is not used.

https://codereview.webrtc.org/2982233002/diff/1/webrtc/modules/remote_bitrate...
File webrtc/modules/remote_bitrate_estimator/test/estimators/bbr.h (right):

https://codereview.webrtc.org/2982233002/diff/1/webrtc/modules/remote_bitrate...
webrtc/modules/remote_bitrate_estimator/test/estimators/bbr.h:80: int64_t
cycle_start_time_ms_;
Add comments for these variables.

https://codereview.webrtc.org/2982233002/diff/1/webrtc/modules/remote_bitrate...
File
webrtc/modules/remote_bitrate_estimator/test/estimators/max_bandwidth_filter.cc
(right):

https://codereview.webrtc.org/2982233002/diff/1/webrtc/modules/remote_bitrate...
webrtc/modules/remote_bitrate_estimator/test/estimators/max_bandwidth_filter.cc:29:
if (bandwidth_samples_[0].first == 0 ||
What is the motivation for implementing the max bandwidth filter this way? This
does not look like BBR.

https://codereview.webrtc.org/2982233002/diff/1/webrtc/modules/remote_bitrate...
File
webrtc/modules/remote_bitrate_estimator/test/estimators/max_bandwidth_filter.h
(right):

https://codereview.webrtc.org/2982233002/diff/1/webrtc/modules/remote_bitrate...
webrtc/modules/remote_bitrate_estimator/test/estimators/max_bandwidth_filter.h:51:
std::pair<int64_t, size_t> bandwidth_samples_[3];
Why 3? And what about the comment on line 36-38

[gnish1, 2017-07-27 08:38:37]

https://codereview.webrtc.org/2982233002/diff/1/webrtc/modules/remote_bitrate...
File webrtc/modules/remote_bitrate_estimator/test/estimators/bbr.cc (right):

https://codereview.webrtc.org/2982233002/diff/1/webrtc/modules/remote_bitrate...
webrtc/modules/remote_bitrate_estimator/test/estimators/bbr.cc:38: const int
kMinimumCongestionWindow = 4000;
On 2017/07/26 11:37:16, philipel wrote:
> Is this in bytes? Why are there two kMinimumCongestionWindow (the other one
> being in congestion_window.cc) and why are they of different sizes?

Done.

https://codereview.webrtc.org/2982233002/diff/1/webrtc/modules/remote_bitrate...
webrtc/modules/remote_bitrate_estimator/test/estimators/bbr.cc:39: const int
kProbeRttDurationRounds = 1;
On 2017/07/26 11:37:16, philipel wrote:
> Comment what this value represent.

Done.

https://codereview.webrtc.org/2982233002/diff/1/webrtc/modules/remote_bitrate...
webrtc/modules/remote_bitrate_estimator/test/estimators/bbr.cc:40: const int
kProbeRttDurationMs = 200;
On 2017/07/26 11:37:16, philipel wrote:
> Comment what this value represent.

Done.

https://codereview.webrtc.org/2982233002/diff/1/webrtc/modules/remote_bitrate...
webrtc/modules/remote_bitrate_estimator/test/estimators/bbr.cc:137: if
(congestion_window_->data_inflight() <= TargetCongestionWindow(1))
On 2017/07/26 11:37:16, philipel wrote:
> Create a variable for 1 which describes what it is.

Done.

https://codereview.webrtc.org/2982233002/diff/1/webrtc/modules/remote_bitrate...
webrtc/modules/remote_bitrate_estimator/test/estimators/bbr.cc:146:
congestion_window_gain_ = 1.5f;
On 2017/07/26 11:37:16, philipel wrote:
> Create a variable for 1.5 which describes what it is.

Done.

https://codereview.webrtc.org/2982233002/diff/1/webrtc/modules/remote_bitrate...
webrtc/modules/remote_bitrate_estimator/test/estimators/bbr.cc:147: unsigned int
seed = time(NULL);
On 2017/07/26 11:37:16, philipel wrote:
> Use webrtc::Random instead (in rc_base/random.h).

Done.

https://codereview.webrtc.org/2982233002/diff/1/webrtc/modules/remote_bitrate...
webrtc/modules/remote_bitrate_estimator/test/estimators/bbr.cc:182:
minimum_congestion_window_start_time_ms_ = -1;
On 2017/07/26 11:37:16, philipel wrote:
> I think using an rtc::Optional might be cleaner in this case.

Done.

https://codereview.webrtc.org/2982233002/diff/1/webrtc/modules/remote_bitrate...
webrtc/modules/remote_bitrate_estimator/test/estimators/bbr.cc:191: bool
min_rtt_expired) {
On 2017/07/26 11:37:16, philipel wrote:
> |min_rtt_expired| is not used.

Done.

https://codereview.webrtc.org/2982233002/diff/1/webrtc/modules/remote_bitrate...
File webrtc/modules/remote_bitrate_estimator/test/estimators/bbr.h (right):

https://codereview.webrtc.org/2982233002/diff/1/webrtc/modules/remote_bitrate...
webrtc/modules/remote_bitrate_estimator/test/estimators/bbr.h:80: int64_t
cycle_start_time_ms_;
On 2017/07/26 11:37:16, philipel wrote:
> Add comments for these variables.

Done.

https://codereview.webrtc.org/2982233002/diff/1/webrtc/modules/remote_bitrate...
File
webrtc/modules/remote_bitrate_estimator/test/estimators/max_bandwidth_filter.cc
(right):

https://codereview.webrtc.org/2982233002/diff/1/webrtc/modules/remote_bitrate...
webrtc/modules/remote_bitrate_estimator/test/estimators/max_bandwidth_filter.cc:29:
if (bandwidth_samples_[0].first == 0 ||
On 2017/07/26 11:37:16, philipel wrote:
> What is the motivation for implementing the max bandwidth filter this way?
This
> does not look like BBR.

Done.

https://codereview.webrtc.org/2982233002/diff/1/webrtc/modules/remote_bitrate...
File
webrtc/modules/remote_bitrate_estimator/test/estimators/max_bandwidth_filter.h
(right):

https://codereview.webrtc.org/2982233002/diff/1/webrtc/modules/remote_bitrate...
webrtc/modules/remote_bitrate_estimator/test/estimators/max_bandwidth_filter.h:51:
std::pair<int64_t, size_t> bandwidth_samples_[3];
On 2017/07/26 11:37:16, philipel wrote:
> Why 3? And what about the comment on line 36-38

Done.

[philipel, 2017-07-27 09:15:28]

https://codereview.webrtc.org/2982233002/diff/20001/webrtc/modules/remote_bit...
File webrtc/modules/remote_bitrate_estimator/test/estimators/bbr.cc (right):

https://codereview.webrtc.org/2982233002/diff/20001/webrtc/modules/remote_bit...
webrtc/modules/remote_bitrate_estimator/test/estimators/bbr.cc:123:
min_rtt_filter_->min_rtt_ms(), gain);
Before calling this function you need to determine what the min rtt is, it is
not always set.

https://codereview.webrtc.org/2982233002/diff/20001/webrtc/modules/remote_bit...
webrtc/modules/remote_bitrate_estimator/test/estimators/bbr.cc:157: Random* rand
= new Random(time(NULL));
Every time you call this function you will leak one webrtc::Random object. Also,
random should only be initialized once, so make it into a member variable of
this class.

https://codereview.webrtc.org/2982233002/diff/20001/webrtc/modules/remote_bit...
webrtc/modules/remote_bitrate_estimator/test/estimators/bbr.cc:172: if
(pacing_gain_ > 1.0 &&
Is this according to the BBR design or is this some tweak you have made? If it
is a tweak please update the comment with that info.

https://codereview.webrtc.org/2982233002/diff/20001/webrtc/modules/remote_bit...
webrtc/modules/remote_bitrate_estimator/test/estimators/bbr.cc:187: void
BbrBweSender::TryEnteringProbeRtt(int64_t now_ms, bool min_rtt_expired) {
Remove |min_rtt_expired| and use |min_rtt_filter_| directly.

https://codereview.webrtc.org/2982233002/diff/20001/webrtc/modules/remote_bit...
File
webrtc/modules/remote_bitrate_estimator/test/estimators/congestion_window.cc
(right):

https://codereview.webrtc.org/2982233002/diff/20001/webrtc/modules/remote_bit...
webrtc/modules/remote_bitrate_estimator/test/estimators/congestion_window.cc:56:
int bdp = *min_rtt_ms * bandwidth_estimate_bps;
|min_rtt_ms| is an Optional but you assume it is always set, that doesn't look
right. I would suggest reverting |min_rtt_ms| back to just an int64_t.

https://codereview.webrtc.org/2982233002/diff/20001/webrtc/modules/remote_bit...
File
webrtc/modules/remote_bitrate_estimator/test/estimators/max_bandwidth_filter.cc
(right):

https://codereview.webrtc.org/2982233002/diff/20001/webrtc/modules/remote_bit...
webrtc/modules/remote_bitrate_estimator/test/estimators/max_bandwidth_filter.cc:25:
// "Bbr design" doc. Rounds are units for packets rtt_time, after packet has
Please update the design doc.

[gnish1, 2017-07-27 11:24:26]

https://codereview.webrtc.org/2982233002/diff/20001/webrtc/modules/remote_bit...
File webrtc/modules/remote_bitrate_estimator/test/estimators/bbr.cc (right):

https://codereview.webrtc.org/2982233002/diff/20001/webrtc/modules/remote_bit...
webrtc/modules/remote_bitrate_estimator/test/estimators/bbr.cc:123:
min_rtt_filter_->min_rtt_ms(), gain);
On 2017/07/27 09:15:27, philipel wrote:
> Before calling this function you need to determine what the min rtt is, it is
> not always set.

GetTargetCongestionWindow() checks if the value is set or not.

https://codereview.webrtc.org/2982233002/diff/20001/webrtc/modules/remote_bit...
webrtc/modules/remote_bitrate_estimator/test/estimators/bbr.cc:157: Random* rand
= new Random(time(NULL));
On 2017/07/27 09:15:27, philipel wrote:
> Every time you call this function you will leak one webrtc::Random object.
Also,
> random should only be initialized once, so make it into a member variable of
> this class.

Done.

https://codereview.webrtc.org/2982233002/diff/20001/webrtc/modules/remote_bit...
webrtc/modules/remote_bitrate_estimator/test/estimators/bbr.cc:172: if
(pacing_gain_ > 1.0 &&
On 2017/07/27 09:15:27, philipel wrote:
> Is this according to the BBR design or is this some tweak you have made? If it
> is a tweak please update the comment with that info.

Done.

https://codereview.webrtc.org/2982233002/diff/20001/webrtc/modules/remote_bit...
webrtc/modules/remote_bitrate_estimator/test/estimators/bbr.cc:187: void
BbrBweSender::TryEnteringProbeRtt(int64_t now_ms, bool min_rtt_expired) {
On 2017/07/27 09:15:27, philipel wrote:
> Remove |min_rtt_expired| and use |min_rtt_filter_| directly.

Done.

https://codereview.webrtc.org/2982233002/diff/20001/webrtc/modules/remote_bit...
File
webrtc/modules/remote_bitrate_estimator/test/estimators/congestion_window.cc
(right):

https://codereview.webrtc.org/2982233002/diff/20001/webrtc/modules/remote_bit...
webrtc/modules/remote_bitrate_estimator/test/estimators/congestion_window.cc:56:
int bdp = *min_rtt_ms * bandwidth_estimate_bps;
On 2017/07/27 09:15:28, philipel wrote:
> |min_rtt_ms| is an Optional but you assume it is always set, that doesn't look
> right. I would suggest reverting |min_rtt_ms| back to just an int64_t.

Done.

https://codereview.webrtc.org/2982233002/diff/20001/webrtc/modules/remote_bit...
File
webrtc/modules/remote_bitrate_estimator/test/estimators/max_bandwidth_filter.cc
(right):

https://codereview.webrtc.org/2982233002/diff/20001/webrtc/modules/remote_bit...
webrtc/modules/remote_bitrate_estimator/test/estimators/max_bandwidth_filter.cc:25:
// "Bbr design" doc. Rounds are units for packets rtt_time, after packet has
On 2017/07/27 09:15:28, philipel wrote:
> Please update the design doc.

Done.

[philipel, 2017-07-31 12:27:57]

Some more feedback, please also update the title/description, they are not very
clear right now.

https://codereview.webrtc.org/2982233002/diff/40001/webrtc/modules/remote_bit...
File
webrtc/modules/remote_bitrate_estimator/test/estimators/max_bandwidth_filter.cc
(right):

https://codereview.webrtc.org/2982233002/diff/40001/webrtc/modules/remote_bit...
webrtc/modules/remote_bitrate_estimator/test/estimators/max_bandwidth_filter.cc:26:
// been acknowledged, one round has passed from its send time.
I don't really understand the last sentence in this comment, please
reformulate/clarify.

https://codereview.webrtc.org/2982233002/diff/40001/webrtc/modules/remote_bit...
webrtc/modules/remote_bitrate_estimator/test/estimators/max_bandwidth_filter.cc:30:
if (bandwidth_samples_[0].first == 0 ||
bandwidth_samples_[0].first is never assigned, so it should never change? Please
add more unittests for this class.

[gnish1, 2017-07-31 15:22:05]

https://codereview.webrtc.org/2982233002/diff/40001/webrtc/modules/remote_bit...
File
webrtc/modules/remote_bitrate_estimator/test/estimators/max_bandwidth_filter.cc
(right):

https://codereview.webrtc.org/2982233002/diff/40001/webrtc/modules/remote_bit...
webrtc/modules/remote_bitrate_estimator/test/estimators/max_bandwidth_filter.cc:26:
// been acknowledged, one round has passed from its send time.
On 2017/07/31 12:27:56, philipel wrote:
> I don't really understand the last sentence in this comment, please
> reformulate/clarify.

Done.

https://codereview.webrtc.org/2982233002/diff/40001/webrtc/modules/remote_bit...
webrtc/modules/remote_bitrate_estimator/test/estimators/max_bandwidth_filter.cc:30:
if (bandwidth_samples_[0].first == 0 ||
On 2017/07/31 12:27:56, philipel wrote:
> bandwidth_samples_[0].first is never assigned, so it should never change?
Please
> add more unittests for this class.

Done.

[philipel, 2017-08-01 08:10:44]

lg, but please update the title/description of this CL.

[gnish1, 2017-08-01 08:16:28]

Description was changed from

==========
Added implementations for entering/exiting all modes, also updated
MaxBandwidthFilter class, where newer implementation for bandwidth filter is
added.

BUG=webrtc:7713
==========

to

==========
Added implementations for entering/exiting STARTUP, DRAIN, PROBE_BW, PROBE_RTT
modes, also updated MaxBandwidthFilter class, with the filter implementation
which stores three best estimates for the filter window.

BUG=webrtc:7713
==========

[philipel, 2017-08-01 08:17:49]

lgtm

[gnish1, 2017-08-01 08:52:11]

The CQ bit was checked by gnish@webrtc.org

[commit-bot: I haz the power, 2017-08-01 08:52:16]

CQ is trying da patch.

Follow status at:
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.webrtc.org/...

[commit-bot: I haz the power, 2017-08-01 09:04:05]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2017-08-01 09:04:06]

Try jobs failed on following builders:
  win_dbg on master.tryserver.webrtc (JOB_FAILED,
http://build.chromium.org/p/tryserver.webrtc/builders/win_dbg/builds/20972)

[gnish1, 2017-08-01 09:42:20]

The CQ bit was checked by gnish@webrtc.org

[gnish1, 2017-08-01 09:42:21]

The patchset sent to the CQ was uploaded after l-g-t-m from philipel@webrtc.org
Link to the patchset: https://codereview.webrtc.org/2982233002/#ps80001 (title:
"Unittest fix")

[commit-bot: I haz the power, 2017-08-01 09:42:30]

CQ is trying da patch.

Follow status at:
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.webrtc.org/...

[commit-bot: I haz the power, 2017-08-01 10:06:18]

CQ is committing da patch.

Bot data: {"patchset_id": 80001, "attempt_start_ts": 1501580540986850,
"parent_rev": "773be36bd61d0c595790a00c8cfb0eab898798fd", "commit_rev":
"d339dbc7d4758e62b02740feda85c82dd6b2eba5"}

[commit-bot: I haz the power, 2017-08-01 10:06:23]

Description was changed from

==========
Added implementations for entering/exiting STARTUP, DRAIN, PROBE_BW, PROBE_RTT
modes, also updated MaxBandwidthFilter class, with the filter implementation
which stores three best estimates for the filter window.

BUG=webrtc:7713
==========

to

==========
Added implementations for entering/exiting STARTUP, DRAIN, PROBE_BW, PROBE_RTT
modes, also updated MaxBandwidthFilter class, with the filter implementation
which stores three best estimates for the filter window.

BUG=webrtc:7713

Review-Url: https://codereview.webrtc.org/2982233002
Cr-Commit-Position: refs/heads/master@{#19196}
Committed:
https://chromium.googlesource.com/external/webrtc/+/d339dbc7d4758e62b02740fed...
==========

[commit-bot: I haz the power, 2017-08-01 10:06:24]

Committed patchset #5 (id:80001) as
https://chromium.googlesource.com/external/webrtc/+/d339dbc7d4758e62b02740fed...