FEC in the MAC doesn't work. FEC in the PHY might be OK.
The common-mode reasoning goes:
802.11 provides a PER of N% for XXX byte packets, so adding FEC
reduces the PER to an acceptable level with no retransmissions.
N(%) is arbitrary. For any data rate, the error rate increases as
range increases. There is no breakpoint where a given data rate is
out of range. So we define "out of range" as the point where PER for
XXX byte packets passes N%.
Adding FEC does not alter than N% number, its probably an OK technique
for increasing range at a given data rate.
If your application can't live with N% PER, you're not going to be any
better off with FEC.
jim
Daniel Dobkin writes:
> Dear Akash:
>
> There is always a tradeoff between speed and coding. The problem is
> that the highest speeds already have rather limited link budgets; if
> you reduce or eliminate the FEC the required S:N goes up. This is
> probably fine if you are doing a line-of-sight link in the same room
> for a video delivery, but not so good farther away. A nice reference
> you can check out is
>
> A Comparison of HIPERLAN/2 and IEEE 802.11a
> Angela Doufexi1, Simon Armour1, Peter Karlsson1,2, Andrew Nix1, David
> Bull1
> 1Centre for Communications Research, University of Bristol, UK
> 2Telia Research AB, Malmoe, Sweden
> E-mail: [EMAIL PROTECTED], [EMAIL PROTECTED]
>
> Unfortunately I don't know where this was published but you can get it
> from the U Bristol folks directly over the web. They did a followup
> paper coupling these results to propagation modeling as well.
>
> --Dan Dobkin
>
>
> On Friday, July 11, 2003, at 04:33 PM, Akash Malhotra wrote:
>
> > Dear Mr Dobkin,
> >
> > Thanks for the reply. I am student whose interest lies in wireless
> > networks especially 802.11 technology. I just got curious to find out
> > be cause I read somewhere that few comapnies are trying to reduce this
> > overhear(FEC) to increase speed to 72 Mbps.
> >
> > Thanks
> >
> > akash
> >
> > On Fri, 11 Jul 2003, Daniel Dobkin wrote:
> >
> >> Response to post of 7/11:
> >>
> >> <<Is it compulsary to use FEC with OFDM technology
> >> such as 802.11 a/g. (?)>>
> >>
> >> It is certainly not compulsory to employ FEC with any OFDM technology;
> >> OFDM is a modulation technique quite independent of encoding.
> >> However,
> >> it certainly is compulsory to employ this method in 802.11a (and g),
> >> as
> >> encoding is part of the standard.
> >>
> >> An encoded data stream is incomprehensible unless you decode it, so
> >> one
> >> could not use FEC on only one end of a link. It is certainly possible
> >> to use FEC in only one direction on a link, presupposing that both
> >> stations are coordinated (that is, EAST transmits coded messages to
> >> WEST, WEST receives and decodes; WEST transmits unencoded messages to
> >> EAST, which receives them and passes them without decoding). It's
> >> unlikely such an arrangement would be worthwhile.
> >>
> >> Why do you want to turn off encoding?
> >>
> >> Daniel M. Dobkin
> >> Enigmatics
> >> 1-408-314-2769
> >> [EMAIL PROTECTED]
> >>
> >
> >
> >
> Daniel M. Dobkin
> Enigmatics
> 1-408-314-2769
> [EMAIL PROTECTED]
> Dear Akash:
>
> There is always a tradeoff between speed and coding. The problem is
> that the highest speeds already have rather limited link budgets; if
> you reduce or eliminate the FEC the required S:N goes up. This is
> probably fine if you are doing a line-of-sight link in the same room
> for a video delivery, but not so good farther away. A nice reference
> you can check out is
>
> A Comparison of HIPERLAN/2 and IEEE 802.11a
> Angela Doufexi1, Simon Armour1, Peter Karlsson1,2, Andrew Nix1, David
> Bull1
> 1Centre for Communications Research, University of Bristol, UK
> 2Telia Research AB, Malmoe, Sweden
> E-mail: [EMAIL PROTECTED], [EMAIL PROTECTED]
>
> Unfortunately I don't know where this was published but you can get it
> from the U Bristol folks directly over the web. They did a followup
> paper coupling these results to propagation modeling as well.
>
> --Dan Dobkin
>
>
> On Friday, July 11, 2003, at 04:33 PM, Akash Malhotra wrote:
>
> Dear Mr Dobkin,
>
> Thanks for the reply. I am student whose interest lies in wireless
> networks especially 802.11 technology. I just got curious to find out
> be cause I read somewhere that few comapnies are trying to reduce this
> overhear(FEC) to increase speed to 72 Mbps.
>
> Thanks
>
> akash
>
> On Fri, 11 Jul 2003, Daniel Dobkin wrote:
>
> Response to post of 7/11:
>
> <<Is it compulsary to use FEC with OFDM technology
> such as 802.11 a/g. (?)>>
>
> It is certainly not compulsory to employ FEC with any OFDM technology;
> OFDM is a modulation technique quite independent of encoding.
> However,
> it certainly is compulsory to employ this method in 802.11a (and g),
> as
> encoding is part of the standard.
>
> An encoded data stream is incomprehensible unless you decode it, so
> one
> could not use FEC on only one end of a link. It is certainly possible
> to use FEC in only one direction on a link, presupposing that both
> stations are coordinated (that is, EAST transmits coded messages to
> WEST, WEST receives and decodes; WEST transmits unencoded messages to
> EAST, which receives them and passes them without decoding). It's
> unlikely such an arrangement would be worthwhile.
>
> Why do you want to turn off encoding?
>
> Daniel M. Dobkin
> Enigmatics
> 1-408-314-2769
> [EMAIL PROTECTED]
>
>
>
>
> Daniel M. Dobkin
> Enigmatics
> 1-408-314-2769
> [EMAIL PROTECTED]
--
"Speed, it seems to me, provides the one genuinely modern pleasure."
-- Aldous Huxley (1894 - 1963)
--
general wireless list, a bawug thing <http://www.bawug.org/>
[un]subscribe: http://lists.bawug.org/mailman/listinfo/wireless
