mirror of
https://github.com/mollyim/webrtc.git
synced 2025-05-13 22:00:47 +01:00
58ed02eae3
This is done by adding a reorder optimizer that estimates the probability of receiving reordered packets. The optimal delay is decided by balancing the cost of increasing the delay against the probability of missing a reordered packet, resulting in a loss. This balance is decided using the `ms_per_loss_percent` parameter. The usage and parameters can be controlled via field trial. Bug: webrtc:10178 Change-Id: Ic484df0412af35610e74b3a6070f2bac7a926a63 Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/231541 Reviewed-by: Ivo Creusen <ivoc@webrtc.org> Commit-Queue: Jakob Ivarsson <jakobi@webrtc.org> Cr-Commit-Position: refs/heads/main@{#34954}
75 lines
2.2 KiB
C++
75 lines
2.2 KiB
C++
/*
|
|
* Copyright (c) 2021 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 "modules/audio_coding/neteq/reorder_optimizer.h"
|
|
|
|
#include <algorithm>
|
|
#include <limits>
|
|
#include <vector>
|
|
|
|
namespace webrtc {
|
|
|
|
namespace {
|
|
|
|
constexpr int kDelayBuckets = 100;
|
|
constexpr int kBucketSizeMs = 20;
|
|
|
|
} // namespace
|
|
|
|
ReorderOptimizer::ReorderOptimizer(int forget_factor,
|
|
int ms_per_loss_percent,
|
|
absl::optional<int> start_forget_weight)
|
|
: histogram_(kDelayBuckets, forget_factor, start_forget_weight),
|
|
ms_per_loss_percent_(ms_per_loss_percent) {}
|
|
|
|
void ReorderOptimizer::Update(int relative_delay_ms,
|
|
bool reordered,
|
|
int base_delay_ms) {
|
|
const int index = reordered ? relative_delay_ms / kBucketSizeMs : 0;
|
|
if (index < histogram_.NumBuckets()) {
|
|
// Maximum delay to register is 2000 ms.
|
|
histogram_.Add(index);
|
|
}
|
|
int bucket_index = MinimizeCostFunction(base_delay_ms);
|
|
optimal_delay_ms_ = (1 + bucket_index) * kBucketSizeMs;
|
|
}
|
|
|
|
void ReorderOptimizer::Reset() {
|
|
histogram_.Reset();
|
|
optimal_delay_ms_.reset();
|
|
}
|
|
|
|
int ReorderOptimizer::MinimizeCostFunction(int base_delay_ms) const {
|
|
const std::vector<int>& buckets = histogram_.buckets();
|
|
|
|
// Values are calculated in Q30.
|
|
int64_t loss_probability = 1 << 30;
|
|
int64_t min_cost = std::numeric_limits<int64_t>::max();
|
|
int min_bucket = 0;
|
|
for (int i = 0; i < static_cast<int>(buckets.size()); ++i) {
|
|
loss_probability -= buckets[i];
|
|
int64_t delay_ms =
|
|
static_cast<int64_t>(std::max(0, i * kBucketSizeMs - base_delay_ms))
|
|
<< 30;
|
|
int64_t cost = delay_ms + 100 * ms_per_loss_percent_ * loss_probability;
|
|
|
|
if (cost < min_cost) {
|
|
min_cost = cost;
|
|
min_bucket = i;
|
|
}
|
|
if (loss_probability == 0) {
|
|
break;
|
|
}
|
|
}
|
|
|
|
return min_bucket;
|
|
}
|
|
|
|
} // namespace webrtc
|