Re: 802.11 10Mhz Channel

> hi,
>
> The patch overall looks good. A couple of comments:
>
> 1) it feels like you could move the WifiMacHelper::Configure* methods to
> the WifiMac base class (that way, you could access
> WifiMac::GetDefaultMaxPropagationDelay directly)
> 2) maybe you could add a Standard attribute to WifiMac which is similar
> to the YansWifiPhy::Standard attribute and does the same thing as
> YansWifiPhy::SetStandard
> 3) maybe also add a Configure80211a and Configure80211b to WifiMac. Erm,
> I just realized that we currently have no set of 802.11b MAC parameters
> for the 802.11b PHY contributed by gary. If you can't bother with
> finding the right values in the spec, just add an empty Configure80211b
> method and we will fill them in later.
>
> Mathieu
>
>
>
> On Thu, 2009-07-09 at 19:23 +0200, Ramon Bauza wrote:
> > Hello,
> >
> > I would like to submit for code review the attached patch implementing
> half
> > and quarter-clocked Wifi operation using 5 and 10 Mhz channels spacing.
> We
> > would be glad to merge this new functionality in the main branch if ns-3
> > developers think it may be interesting for other people. The following
> files
> > have been modified (ns-3-dev):
> >
> > src/devices/wifi/interference-helper.cc
> > src/devices/wifi/interference-helper.h
> > src/devices/wifi/wifi-phy.cc
> > src/devices/wifi/wifi-phy.h
> > src/devices/wifi/wifi-phy-standard.h
> > src/devices/wifi/yans-error-rate-model.cc
> > src/devices/wifi/yans-wifi-phy.cc
> > src/devices/wifi/yans-wifi-phy.h
> > src/helper/nqos-wifi-mac-helper.cc
> > src/helper/nqos-wifi-mac-helper.h
> > src/helper/qos-wifi-mac-helper.cc
> > src/helper/qos-wifi-mac-helper.h
> > src/helper/wifi-helper.cc
> > src/helper/wifi-helper.h
> >
> > Apart from the modifications pointed out in the previous email posted
> some
> > weeks ago, we have also modified some of the wifi helpers. In order not
> to
> > make the user set weird values for the MAC parameters (slot,sifs, etc)
> from
> > the script, we have modified the WifiMacHelper incorporating two new
> > functions (Configure80211_10MhzParameters and
> Configure80211_5MhzParameters)
> > that allows the user to automatically select the default MAC parameters
> for
> > 5Mhz and 10Mhz channel spacing (we have also moved the function SetType
> and
> > the ObjectFactory m_mac from NqosWifiMacHelper and QosWifiMacHelper to
> > WifiMacHelper).
> >
> > Looking forward to your comments.
> >
> > Best regards,
> > Ramon Bauza.
> >
> > 2009/5/29 Ramon Bauza <monbauza@...>
> >
> > > Hello everyone,
> > >
> > > I am Ramon Bauza, a researcher at Miguel Hernandez University (Spain).
> > > Currently, I am involved in the EU FP7 project iTETRIS which aims at
> > > creating an integrated traffic and wireless simulation platform that
> will
> > > allow evaluating traffic management strategies and communication
> protocols
> > > in vehicular environments. The iTETRIS project employs SUMO and ns-3 as
> > > traffic simulation and wireless simulation platforms respectively.
> > >
> > > Our research group at UMH, it is charge of implementing the 802.11p
> (WAVE)
> > > standard in ns-3. 802.11p is based on 802.11a, being the channel
> bandwidth
> > > and thus also the data rates the main differences between both (802.11p
> uses
> > > a 10Mhz whereas 802.11a uses 20Mhz). After implementing 802.11p in
> ns-3, we
> > > will also look into and implement propagation models for vehicular
> > > environments that consider LOS/NLOS visibility conditions,
> > > vehicle-to-vehicle and vehicle-to-infraestructure communications, and
> > > shadowing and multipath fading effects.
> > >
> > > In order to model a 10Mhz channel in ns-3, some additions/modifications
> > > should be done (some weeks ago I posted a message in the user mailing
> list
> > > regarding this
> > >
> http://groups.google.com/group/ns-3-users/browse_thread/thread/f27b1ddcf9f1d532#
> ).
> > > Apart from 802.11p, I think other 802.11 standards employ 10MHz
> channels and
> > > thus probably the modeling of 10Mhz channels would also be useful for
> the
> > > ns-3 community working on 802.11-related topics. In fact, the IEEE Std
> > > 802.11TM-2007 standard considers 10Mhz channel spacing using OFDM. If
> you
> > > consider that the modeling of 10Mhz channels may be beneficial for
> other
> > > users we would be glad to merge the 10Mhz modeling in the main branch.
> > >
> > > Below, I compile the needed functions and modifications we plan to
> > > introduce in ns-3 at the PHY layer for modeling a 10Mhz channel:
> > >
> > > + wifi-phy-standard.h
> > >
> > >   Add WIFI_PHY_STANDARD_80211_10Mhz in the enum WifiPhyStandard
> > >
> > > + wifi-phy.h and wifi-phy.cc
> > >
> > >   Declare and define the following new functions:
> > >
> > >     static WifiMode Get3mb10Mhz (void);
> > >     static WifiMode Get4_5mb10Mhz (void);
> > >     static WifiMode Get6mb10Mhz (void);
> > >     static WifiMode Get9mb10Mhz (void);
> > >     static WifiMode Get12mb10Mhz (void);
> > >     static WifiMode Get18mb10Mhz (void);
> > >     static WifiMode Get24mb10Mhz (void);
> > >     static WifiMode Get27mb10Mhz (void);
> > >
> > > + yans-wifi-phy.h and yans-wifi-phy.cc
> > >
> > >   Declare and define the void Configure80211_10Mhz (void) function.
> > >   Modify the void SetStandard (enum WifiPhyStandard standard) function
> so
> > > that the user can also select a 10Mhz channel.
> > >
> > > + interference-helper.h and interference-helper.cc
> > >
> > >   Declare and define void Configure80211_10MhzParameters (void);
> > >
> > > One function that should be modified is
> > > InterferenceHelper::CalculateTxDuration (uint32_t size, WifiMode
> > > payloadMode, WifiPreamble preamble) const since it assumes that the
> symbol
> > > duration is always 4us (20 Mhz channel). This function should employ a
> > > symbol duration according to the channel used (10Mhz or 20Mhz). One
> possible
> > > solution to solve this would be to have a new variable in the class
> > > InterferenceHelper which allows differentiating the channel used. This
> > > variable could be of type WifiPhyStandard. In the function Time
> > > InterferenceHelper::CalculateTxDuration, this new variable would be
> examined
> > > and the packet txon duration will be calculated based on the bandwidth
> > > channel.
> > >
> > > Please, I would like to have your comments on this in order to proceed
> with
> > > the implementation.
> > >
> > > Thank you in advance.
> > > Best regards,
> > > Ramon.
> > >
> > >
> --------------------------------------------------------------------------
> > > Ramon Bauza
> > > Ubiquitous Wireless Communications Research Laboratory
> > > Uwicore, http://www.uwicore.umh.es
> > > Signal Theory and Communications Division
> > > University Miguel Hernández (Spain)
> > > Tel: +34 96522 2031
> > > Fax: +34 96665 8903
> > >
> > >
> ----------------------------------------------------------------------------
> > >
>
>