Rietveld Code Review Tool
Help | Bug tracker | Discussion group | Source code | Sign in
(168)

Issue 5314055: csma module improvements (propagation delay and carrier sense)

Can't Edit
Can't Publish+Mail
Start Review
Created:
12 years, 5 months ago by Vedran Miletić
Modified:
9 years, 9 months ago
Reviewers:
Mehdi Moussouni
Visibility:
Public.

Description

The implementation is still very rough on the edges and has lots of commented out code (please ignore it), and I will clean it up for inclusion. What seems to work: * realistic propagation delays * carrier sense What isn't tested: * propagation delay and loss models (hopefully didn't break anything, but unlikely) * large number of devices and applications (it shouldn't break) What isn't implemented: * jam signal (probably easy, just need to think about it a bit) * collision detection (sort of, the code is partially there, will fix it before inclusion) * tests (will do once I know it works) Idea: the channel has it's delay (e.g. 2ms) and BooleanValue PositioningEnabled; false => old style csma, true => device positions on the wire are taken into account when computing delay. Device Position is a number betwen 0.0 and 1.0, where those represent ends of the wire. (I started with length but in the end this approach made more sense).

Patch Set 1 #

Total comments: 3

Patch Set 2 : CSMA module improvements rebased on top of ns-3-dev as of 2014-06-10 #

Total comments: 1
Unified diffs Side-by-side diffs Delta from patch set Stats (+129 lines, -50 lines) Patch
M src/csma/helper/csma-helper.h View 1 8 chunks +14 lines, -7 lines 1 comment Download
M src/csma/helper/csma-helper.cc View 1 4 chunks +18 lines, -16 lines 0 comments Download
M src/csma/model/csma-channel.h View 1 5 chunks +11 lines, -4 lines 0 comments Download
M src/csma/model/csma-channel.cc View 1 10 chunks +27 lines, -16 lines 0 comments Download
M src/csma/model/csma-net-device.h View 1 5 chunks +32 lines, -2 lines 0 comments Download
M src/csma/model/csma-net-device.cc View 1 7 chunks +27 lines, -5 lines 0 comments Download

Messages

Total messages: 3
Lalith Suresh
Had a quick look. Looks good. I'm curious to see how you plan to implement ...
12 years, 5 months ago (2011-10-29 14:07:59 UTC) #1
Mehdi Moussouni
https://codereview.appspot.com/5314055/diff/1/src/csma/model/csma-channel.cc File src/csma/model/csma-channel.cc (right): https://codereview.appspot.com/5314055/diff/1/src/csma/model/csma-channel.cc#newcode323 src/csma/model/csma-channel.cc:323: tEventProp = Seconds (std::abs (m_deviceList[srcId]->GetPosition () - (*it)->GetPosition ()) ...
9 years, 9 months ago (2014-06-26 18:31:06 UTC) #2
Mehdi Moussouni
9 years, 9 months ago (2014-06-27 14:35:21 UTC) #3
Realistic propag. delay system looks good for me but it's only for a wire bus
(1-Dimension) : CSMA is also used for wireless communications (3-Dim).

https://codereview.appspot.com/5314055/diff/3002/src/csma/helper/csma-helper.h
File src/csma/helper/csma-helper.h (right):

https://codereview.appspot.com/5314055/diff/3002/src/csma/helper/csma-helper....
src/csma/helper/csma-helper.h:102: * \param distance Distance in metres from one
end of the wire.
Realistic propag. delay system looks good for me but it's only for a wire bus
(1-Dimension) and CSMA is also used for wireless communications (3-Dim).
Sign in to reply to this message.

Powered by Google App Engine
RSS Feeds Recent Issues | This issue
This is Rietveld f62528b