www.a00.de > tcpgroup > 1991 > msg00068

TCP-group 1991

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

MAC sublayer protocol

fred k1io writes in response to Steve Johnson:

> user in effect a backbone node.  The problem we're trying to solve is
> ordinary shlubs like me, who don't have dedicated radios, and just run
> packet once in a while, not 7x24.
> In fact, a repeater is MORE spectrum-efficient than a single-freq node.
> Single-freq gives us Aloha, with 17-25% efficiency before congestion
> collapse.  Repeaters (with half-duplex leaf nodes) give us CSMA, with
> 35% or so efficiency and more gracefully self-limiting traffic. SInce
> repeaters take half as long (once to receive and once to send) to carry
> traffic as digis, they are roughly FOUR TIMES as effective, using TWO
> TIMES the bandwidth.  If we really wanted to clean up the 2m mess,
> we'd ban digis and insist that all non-direct L2 traffic go through
> repeaters.  But the yakkers are content to leave us in our cesspool
> around 145.0.

Not to argue with the above (and still biting my tongue WRT a myriad of
previous comments) I think there is a hidden issue here.
  If the repeater is located such that it *only* radiates to<->from the
same set of U users which would otherwise be trying to CSMA it alone I
agree.  BUT, very often the repeater is located at a high level site.
This means that it not only gobbles the spectrum from a much larger area
but also is likely to increase the number of users.  The result can be
that 10*U(or whatever) users now vie for the same channel and we are
pushed further up the congestive collapse curve, even though things are
working better than if those same 10U users tried to all go it alone
(assuming they could all hear and interfere with each other which they
probably couldn't).
  Wide area omni (multiple access) is not better than multiple small
area MA is not better than all-pt-pt.  Frequency reuse improves as
coverage area gets smaller and aggregate channel capacity improves.
  Better directedness of transmitted beams and control of cluster size
(physical and number-of-users) allow better resource reuse but do
require increased coordination/cooperation.  Keeping a local server
(like a repeater) low enough to just serve a bounded number of users
without interfering with others takes planning.  Admittedly, the
non-7x24, omni-omni non-dedicated hardware case, while important, is not
long in this respect. Still, if one goes to the trouble of installing
a repeater (takes planning) I think a look at the overall effect is
in order at the same time.

Glenn n6gn

Document URL : http://www.a00.de/tcpgroup/1991/msg00068.php
Ralf D. Kloth, Ludwigsburg, DE (QRQ.software). < hostmaster at a00.de > [don't send spam]
Created 2004-12-21. Last modified 2004-12-21. Your visit 2021-10-25 00:43.20. Page created in 8.0604 sec.
[Go to the top of this page]   [... to the index page]