Add congestion control parameters to config

[cnderrauber]

The loss-based congestion control get poor
performance under high bandwidth, high rtt
and packet loss case since the congestion
window becomes 1 mtu and increase slowly
after retransmit timeout. And fast recovery
retransmit cause exit slowly in consecutive
packet loss. This change add paramters to
the config then the user can set them to get
higher throughput in such cases.

Description

Reference issue

Fixes #...

[codecov]

Codecov Report

Attention: Patch coverage is 82.60870% with 4 lines in your changes missing coverage. Please review.

Project coverage is 81.43%. Comparing base (943ac50) to head (1d98af6).
Report is 1 commits behind head on master.

Files with missing lines	Patch %	Lines
association.go	82.60%	2 Missing and 2 partials ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #354      +/-   ##
==========================================
+ Coverage   81.24%   81.43%   +0.19%     
==========================================
  Files          51       51              
  Lines        3348     4337     +989     
==========================================
+ Hits         2720     3532     +812     
- Misses        483      660     +177     
  Partials      145      145              

Flag	Coverage Δ
go	81.43% <82.60%> (+0.19%)	⬆️
wasm	67.21% <26.08%> (-0.03%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

---

🚨 Try these New Features:

• Flaky Tests Detection - Detect and resolve failed and flaky tests

[boks1971]

lgtm, is it possible to add a UT for this?

[cnderrauber]

The benchmark shows improvements under high rtt/loss condition with custom congestion parameters, 50mpbs@200ms, 6.48mbps@100ms/1%loss, 3.76mbps@200ms/1%loss, compared to original 37mbps@200ms, 1.36mbps@100ms/1%loss, 0.72mbps@200ms/1%loss.

[davidzhao]

what does control look like? it would be great to see the before & after benchmarks.

[Sean-Der]

@cnderrauber fantastic!

I don't fully understand the change. Would it be possible to make the diff only be related to congestion control? Like the bundleDataChunksIntoPackets

[Sean-Der]

What does usrsctp do here? @tuexen Do you have any opinion/any insights? What should we be doing better

[cnderrauber]

what does control look like? it would be great to see the before & after benchmarks.

Here is 2 linux server with tc to simulate 100ms rtt, with default parameter the fast recovery costs seconds to exit and the congestion window is throlttled, and a large fast retransmit window make it exit quickly and the congestion window increase quickly as well as the bitrates.

[cnderrauber]

@cnderrauber fantastic!

I don't fully understand the change. Would it be possible to make the diff only be related to congestion control? Like the bundleDataChunksIntoPackets

All changes and paramters are related to the congestion control, what does only be related to congestion control mean?