Add a new overuse estimator for the delay based BWE behind experiment.

webrtc/modules/congestion_controller/trendline_estimator.cc

Index: webrtc/modules/congestion_controller/trendline_estimator.cc
diff --git a/webrtc/modules/congestion_controller/trendline_estimator.cc b/webrtc/modules/congestion_controller/trendline_estimator.cc
new file mode 100644
index 0000000000000000000000000000000000000000..a086f6aaa1bae4055caddaa77fa3d48962f7bcdf
--- /dev/null
+++ b/webrtc/modules/congestion_controller/trendline_estimator.cc
@@ -0,0 +1,87 @@
+/*
+ *  Copyright (c) 2016 The WebRTC project authors. All Rights Reserved.
+ *
+ *  Use of this source code is governed by a BSD-style license
+ *  that can be found in the LICENSE file in the root of the source
+ *  tree. An additional intellectual property rights grant can be found
+ *  in the file PATENTS.  All contributing project authors may
+ *  be found in the AUTHORS file in the root of the source tree.
+ */
+
+#include "webrtc/modules/congestion_controller/trendline_estimator.h"
+
+#include <algorithm>
+
+#include "webrtc/base/checks.h"
+#include "webrtc/modules/remote_bitrate_estimator/test/bwe_test_logging.h"
+
+namespace webrtc {
+
+namespace {
+double LinearFitSlope(const std::list<std::pair<double, double>> points) {
+  RTC_DCHECK(points.size() >= 2);

[stefan-webrtc, 2016/11/16 15:11:20]

Perhaps store points.size() in a local variable instead of calling multiple
times. Probably a no-op, but I think it's nice when using lists. :)

[terelius, 2016/11/16 15:39:12]

Since C++11, size() is required to be constant time.

+  // Compute the "center of mass".
+  double sum_x = 0;
+  double sum_y = 0;
+  for (const auto& point : points) {
+    sum_x += point.first;
+    sum_y += point.second;
+  }
+  double x_avg = sum_x / points.size();
+  double y_avg = sum_y / points.size();
+  // Compute the slope k = \sum (x_i-x_avg)(y_i-y_avg) / \sum (x_i-x_avg)^2
+  double numerator = 0;
+  double denominator = 0;
+  for (const auto& point : points) {
+    numerator += (point.first - x_avg) * (point.second - y_avg);
+    denominator += (point.first - x_avg) * (point.first - x_avg);
+  }
+  return numerator / denominator;
+}
+}  // namespace
+
+enum { kDeltaCounterMax = 1000 };
+
+TrendlineEstimator::TrendlineEstimator(size_t window_size,
+                                       double smoothing_coef,
+                                       double threshold_gain)
+    : window_size_(window_size),
+      smoothing_coef_(smoothing_coef),
+      threshold_gain_(threshold_gain),
+      num_of_deltas_(0),
+      accumulated_delay_(0),
+      smoothed_delay_(0),
+      delay_hist_(),
+      trendline_(0) {}
+
+TrendlineEstimator::~TrendlineEstimator() {}
+
+void TrendlineEstimator::Update(double recv_delta_ms,
+                                double send_delta_ms,
+                                double now_ms) {
+  const double delta_ms = recv_delta_ms - send_delta_ms;
+  ++num_of_deltas_;
+  if (num_of_deltas_ > kDeltaCounterMax) {
+    num_of_deltas_ = kDeltaCounterMax;
+  }
+
+  // Exponential backoff filter.
+  accumulated_delay_ += delta_ms;
+  BWE_TEST_LOGGING_PLOT(1, "accumulated_delay_ms", now_ms, accumulated_delay_);
+  smoothed_delay_ = smoothing_coef_ * smoothed_delay_ +
+                    (1 - smoothing_coef_) * accumulated_delay_;
+  BWE_TEST_LOGGING_PLOT(1, "smoothed_delay_ms", now_ms, smoothed_delay_);
+
+  // Simple linear regression.
+  delay_hist_.push_back(std::make_pair(now_ms, smoothed_delay_));
+  if (delay_hist_.size() > window_size_) {
+    delay_hist_.pop_front();
+  }
+  if (delay_hist_.size() == window_size_) {
+    trendline_ = LinearFitSlope(delay_hist_);
+  }
+
+  BWE_TEST_LOGGING_PLOT(1, "trendline_slope", now_ms, trendline_);
+}
+
+}  // namespace webrtc