Home | Help | Search | Login | Register

mulonga.net forum  |  IT Centers  |  Binga Wireless Project  |  Topic: WELCOME TO JOIN THE DISCUSSION ON BINGA WIRELESS « previous next »
Pages: 1 [2] Print
Author Topic: WELCOME TO JOIN THE DISCUSSION ON BINGA WIRELESS  (Read 9416 times)
Peter Kuthan / AZFA
Global Moderator
Hero Member
*****
Posts: 819


« Reply #15 on: December 13 2007 »

i agree. it is feasible but not easy at all. a lot of preconditions have
to be met. and EIRP limit of 100mw at 2.4GHz can *not* be met. maybe
EIRP of 1000mw @ 5.8GHz can be enough but for 85km. for 2.4GHz i would at least suggest a 24dbi gain antenna and ~10-15dbm radio output (exact value has to be figured out on site). so EIRP would be far beyond 20dbm.

as aaron said another important precondition is the line of sight
and a obstacle free fresnel zone [1]. a 85km link would result in a
fresnel zone of ~50m diameter in the middle. se how to calculate it here: [2]
this means the antennas would have to be at least 25m over the ground (or on a hill,...) to have good physical conditions.

the ~40 km links can be done with 2.4GHz more or less within the 100mw EIRP with good equipement.

in generall, to have a good QoS and redundancy in case of failures we
should link B-D, D-F and E-F if feasable (LoS, FresnelZone)

in combination with olsr mesh routing [3] this would result in more robustness and performance and the extra costs for a couple of antennas
and routers are marginal.

best
/roland

---
[1]
http://en.wikipedia.org/wiki/Fresnel_zone

[2]
http://www.bcwireless.net/tools/rf/fresnel.main.cgi

[3]
http://en.wikipedia.org/wiki/OLSR
Logged
Peter Kuthan / AZFA
Global Moderator
Hero Member
*****
Posts: 819


« Reply #16 on: December 14 2007 »

Greetings Simango

I think someone has already made comments in relation
to the Geo-physical distances that you would like to
cover in Binga and that the power budget calculations
could be influenced by free space loss for those
distances of 85km between masts. I would be concern
that reliability will be compromised unless there is
clear LoS between the points, and knowing Binga I
guess not.

My experinece so far has indicated that using
off-the-shelf commercial grade (with hardened outdoor
electronics!) WiFi kit, you will get equipment that
has throughput curves effective at between 15 and 20
km. That means in your case, repeaters along the way
and therefore a question of power (in the bush) and
security.

I am sure you have done modelling in terms of what
connectivity solution you would prefer and your Total
Cost of Ownership points you towards WiFi. However
with all this equipment that needs to be installed,
protected, powered  - a less equipmnet intense
solution should prevail. If you get a satelitte link
to a central point and then mix and match with WiFi
for sites at less than 20km, that might help.

The project that my organisdation is executing and you
came and saw at the school in Highfields Harare is
working at distances of not more than 8 km betwen
access point and bridge and all I can say is that the
network has been tempremental at times and distnaces
issues and alignmnet have been problems.

I will give you regulatory recommendations for ISM.
Countries like South Africa and its "FCC" equivalent
ICASA, have simply turned a blind eye and self
regulation is working in places. More later

Muroro Dziruni
Connect Africa - Zimbabwe
Logged
Peter Kuthan / AZFA
Global Moderator
Hero Member
*****
Posts: 819


« Reply #17 on: December 14 2007 »

I do agree that 85km+ would be quite ambitious. I have talked to a few
RF engineers who do believe you can come up with a workable link
budget with some fairly sized antennas etc. Having been on site i would
think for this particular link, your legacy type long distance licensed microwave would be best . Wifi type could radiate from either site there of with i think the furthest points from either site being at most ~46km(R Simango can confirm) . There are a couple of who claim fellows with working ~70km links around.

I do believe that RSA's ICASA has a more favourable approach to ISM use in
as far as giving mandate to local authorites/municipals to allow/disallow the use
of the band in the specific geographical locations. I am looking forward to a meeting
spectrum engineers from ICASA to get whatever framework (if any) they used on
the use thereof. I would be glad to share the insight afterwards.

http://www.internet.org.za/ICASA_2.4_GHz_warning_notice.pdf

TC.
Logged
Peter Kuthan / AZFA
Global Moderator
Hero Member
*****
Posts: 819


« Reply #18 on: December 16 2007 »

reply from aaron:

On Dec 14, 2007, at 1:26 PM, Muroro Dziruni wrote:

> Greetings Simango
>
> I think someone has already made comments in relation
> to the Geo-physical distances that you would like to
> cover in Binga and that the power budget calculations
> could be influenced by free space loss for those
> distances of 85km between masts. I would be concern
> that reliability will be compromised unless there is
> clear LoS between the points, and knowing Binga I
> guess not.
>
> My experinece so far has indicated that using
> off-the-shelf commercial grade (with hardened outdoor
> electronics!) WiFi kit, you will get equipment that
> has throughput curves effective at between 15 and 20
> km. That means in your case, repeaters along the way
> and therefore a question of power (in the bush) and
> security.

With 5.8 GHz osbridges we achieve 31km without problems or tuning anything.
Line of sight is a strict requirement in any case.

>
> I am sure you have done modelling in terms of what
> connectivity solution you would prefer and your Total
> Cost of Ownership points you towards WiFi. However
> with all this equipment that needs to be installed,
> protected, powered  - a less equipmnet intense
> solution should prevail. If you get a satelitte link
> to a central point and then mix and match with WiFi
> for sites at less than 20km, that might help.

i might drop in the term: "mesh" and "multi hop" .
Smiley
It works.

Is SAT uplink allowed in Zim? I thought that VSAT was strictly forbidden, no?

>
> The project that my organisdation is executing and you
> came and saw at the school in Highfields Harare is
> working at distances of not more than 8 km betwen
> access point and bridge and all I can say is that the
> network has been tempremental at times and distnaces
> issues and alignmnet have been problems.
>
> I will give you regulatory recommendations for ISM.
> Countries like South Africa and its "FCC" equivalent
> ICASA, have simpoly turned a blind eye and self
> regulation is working in places. More later
>
good Smiley
aaron

> Muroro Dziruni
Logged
Peter Kuthan / AZFA
Global Moderator
Hero Member
*****
Posts: 819


« Reply #19 on: December 29 2007 »

Re: [Tongaonline] mailinglist ... easier

Dear Aaron, hi all

good idea to look for ways how to make our communication and further discussion on Tonga.Online extension and especially "Binga wireless" more easy and clearly laid out.

I discussed this topic with Klaus and Stefan yesterday and we found that such a platform for communication should be integrated into the mulonga website as the discussion forum we would like to promote. This would also better provide for maximum transparency and participation.

What do you think? How to make the forum better accessible?

best wishes - for 2008!

Peter

Aaron Kaplan wrote:
> Hi!
>
> I subscribed everybody to a new mailings list. This makes it easier  and we don't have to CC everybody.
> In case somebody wants to be administrator of this list, let me know.
>
> I created the list from the CC list we had in the previous mails. We  found out that Daniel (funkfeuer) was not in the CC list although he  is very interested.
>
> So!
>
> Feel free to use this list (just reply to the list and everybody gets  it) . If you don't like the list, then you can also unsubscribe (or  change your settings ) on the link below.
>
> cheers,
> aaron.
Logged
Pages: 1 [2] Print 
mulonga.net forum  |  IT Centers  |  Binga Wireless Project  |  Topic: WELCOME TO JOIN THE DISCUSSION ON BINGA WIRELESS « previous next »
Jump to:  

Powered by SMF 1.1.8 | SMF © 2006-2008, Simple Machines LLC