At 07:50 AM 7/18/00, Kent wrote:
>Can you guys tell me what is the difference between CD
>and CA? Detection and Avoidance?

I'm not a guy, but I can explain the difference. In fact, I already did in 
a message yesterday.

Carrier Sense Multiple Access with Collision Detect (CSMA/CD) means listen 
before you send (carrier sense) and listen while you're sending to 
determine if there is a collision (collision detect). When listening before 
sending, if you sense another station already sending, you defer your 
sending. When the frame in transmission is finished, Ethernet says to wait 
the interframe gap, and then send. The problem with this method is that 
multiple stations may sense that the frame in transmission is finished and 
send at the same time. That's not a fatal problem, however, because 
stations listen while sending for collisions. If there is a collision, they 
back off a random amount of time and send again.

Carrier Sense Multiple Access with Collision Avoidance (CSMA/CA) tries to 
avoid collisions by specifying that stations wait a random time before 
sending. After sensing that the medium is no longer in use, a station waits 
the interframe gap plus a random amount of time, and then sends. Most 
CSMA/CA protocols also include a Request to Send (RTS) and Clear to Send 
(CTS) protocol, to further reduce the possibility of collisions.

802.11 wireless LAN has a complex MAC layer that is quite different from 
802.3 Ethernet, despite the fact that people call it "wireless Ethernet." 
The medium access control is different and the frame format is different. 
802.11 uses CSMA/CA. It tries to avoid collisions and does not detect 
collisions if they do happen.

Detecting collisions requires the capability to send and receive at the 
same time to determine if another station's transmission is interfering 
with one's own transmission. This can be costly. Moreover, even if a 
station had collision detection functionality and sensed no collision when 
sending, a collision could have actually occurred at the receiver. This 
situation results from the particular characteristics of a wireless 
network. Suppose that station A is transmitting to station B. Suppose that 
station C is also transmitting to station B. A physical obstruction, such 
as a brick wall, could prevent A and C from hearing each others' 
transmissions, even though their transmissions are indeed interfering at 
the destination, B.

In the 802.11 specification, the physical layer (PHY) monitors the energy 
level on the radio frequency to determine whether or not another station is 
transmitting. The PHY layer provides this carrier sensing information to 
the MAC protocol. If the channel is sensed idle for an amount of time equal 
to or greater than the Distributed Inter Frame Space (DIFS), a station is 
allowed to transmit.

What happens if the sender senses the channel is busy though? In this case, 
the station performs a backoff procedure that is similar to that of 
Ethernet's backoff procedure when a collision occurs. A station that senses 
that the channel is busy defers its access until the channel is later 
sensed idle. Once the channel is sensed idle for an amount of time equal to 
the DIFS, the station then computes an additional random backoff time and 
counts down this time until the timer reaches zero, at which time the 
station transmits its frame.

The IEEE 802.11 frame contains a duration field in which the sending 
station explicitly indicates the length of time that its frame will be 
transmitting on the channel. This value allows other stations to determine 
the minimum amount of time (the so-called network allocation vector, NAV) 
for which they should defer their access.

To decrease the likelihood of collisions even more, the IEEE 802.11 
protocol can also use RTS and CTS control frames to reserve access to the 
channel. When a sender wants to send a frame, it can first send an RTS 
frame to the receiver. A receiver that receives an RTS frame responds with 
a CTS frame, giving the sender explicit permission to send. All other 
stations hearing the RTS or CTS then know about the pending data 
transmission and can avoid interfering with those transmissions.

Priscilla




>Thanks
>
>Kent
>--- "Stanfield Hilman B(Brad)  CONT NNSY"
><[EMAIL PROTECTED]> wrote:
> > 802.3 uses CSMA/CD *BUT* 802.11 (wireless ETHERNET)
> > uses CSMA/CA so, in the
> > truest sense YES, it IS possible for Ethernet media
> > to deploy both. Just not
> > simultaneously.
> >
> >
>************************************************************************
> > Brad Stanfield
> > Network Engineer
> > [EMAIL PROTECTED]
> > Government Micro Resources
> > Network Operations Control Center
> > Bldg 33 NAVSEA NCOE
> > 757-393-9526
> >
> >
> >
> > -----Original Message-----
> > From: Luong, David [mailto:[EMAIL PROTECTED]]
> > Sent: Monday, July 17, 2000 5:19 PM
> > To: '[EMAIL PROTECTED]'; 'Cisco Group Study'
> > Subject: RE: Deferred Packets
> >
> >
> > Further to your response, a fellow co-worker of mind
> > has mentioned that
> > ethernet used CSMA/CD 802.3 instead of CA? Is it
> > possible for ethernet media
> > to deploy both? I have not come across this
> > before....any takers out there?
> >
> > Dave.
> >
> > -----Original Message-----
> > From: [EMAIL PROTECTED]
> > [mailto:[EMAIL PROTECTED]]
> > Sent: Monday, July 17, 2000 12:40 PM
> > To: Luong, David
> > Subject: Re: Deferred Packets
> >
> >
> >
> >
> > Deferred packets are a normal part of an ethernet
> > transmission.
> >
> > If multiple hosts attempt to transmit over the same
> > ethernet media at the
> > same
> > time, a collision occurs. After a specified period
> > of time, both packets
> > involved in the collision are lost and must be
> > retransmitted by the
> > transmitting
> > hosts. Ethernet interfaces employ
> > collision-avoidance methods to detect
> > other
> > traffic on the network and await opportunities for
> > safe transmission,this is
> > also known as ethernet deference.
> >
> > Hope this helps.
> >
> >
> >
> >
> >
> >
> >
> >
> > "Luong, David" <[EMAIL PROTECTED]> on 07/17/2000
> > 01:37:57 PM
> >
> > Please respond to "Luong, David"
> > <[EMAIL PROTECTED]>
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >  To:      "'Cisco Group Study'"
> > <[EMAIL PROTECTED]>
> >
> >
> >  cc:      (bcc: LOMBMML Hill
> > Michale/Lincoln/Metromail)
> >
> >
> >
> >
> >
> >
> >  Subject: Deferred Packets
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> > Hi Folks:
> >
> > On a 10-half ethernet connection between two
> > routers, we are seeing deferred
> > packets happening as well as the normal usual
> > collisions. Are deferred
> > packets just normal as well? Is it expected in
> > ethernet media? Thanks in
> > advance...
> >
> > David Luong
> > CCNP,CCNA,Network+,A+,i-Net+
> > Telecommunications Analyst
> > Insurance Corporation of B.C.
> > Vancouver, B.C CANADA
> >
> >
> > ___________________________________
> > UPDATED Posting Guidelines:
> > http://www.groupstudy.com/list/guide.html
> > FAQ, list archives, and subscription info:
> > http://www.groupstudy.com
> > Report misconduct and Nondisclosure violations to
>[EMAIL PROTECTED]
>
>
>__________________________________________________
>Do You Yahoo!?
>Get Yahoo! Mail � Free email you can access from anywhere!
>http://mail.yahoo.com/
>
>___________________________________
>UPDATED Posting Guidelines: http://www.groupstudy.com/list/guide.html
>FAQ, list archives, and subscription info: http://www.groupstudy.com
>Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]


________________________

Priscilla Oppenheimer
http://www.priscilla.com

___________________________________
UPDATED Posting Guidelines: http://www.groupstudy.com/list/guide.html
FAQ, list archives, and subscription info: http://www.groupstudy.com
Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]

Reply via email to