Hi all,

We are located in Hungary and the biggest cable (DSL) internet provider released its brand new packages. Just giving an example, they can provide 2.5 Mbit/s for about 20 USD. (This price was the 1.25 Mbit/s package's price previously, just in case.) Being a small wifi (5GHz) provider, we have a package of 1.5 Mbit/s for 32 USD. The problem is that how can we compete with this price and bandwidth? If we are decreasing prices, we will have a profit loss. If we are making a bandwidth increase, we can't provide it, especially in case of point-multipoint links. What would be your strategy or even ideas?

Ps: Is there any outstanding development in case of wifi technology or so. It is really an urgent need in order to be able to be up to cable providers.

Hope we do not have to wait for long for this novelty such as wimax. On the other way, 802.11n standard is _STILL_ not available in Mikrotik OS. Moreover this works only on 2.4 GHz yet that is fairly overladen and also haven't seen any outdoor solutions of it at all.

Any help, idea, suggestion, solution is greatly appreciated.

gyoztes

Switch to Alvarion VL or Motorola Canopy for wireless and use MT for routing only. Otherwise target the customers outside of DSL's reach.

Lol, jwcn

How does that allow him to provide better speeds at the same price as the DSL guys?

My suggestion would be to target the rural markets where DSL is not available, and will not be available.

They have proven products that can deliver speeds DSL is not capable of.

Point is that things should be used what they are meant for.

Wifi is not for broadband area coverage. If you need to provide wireless broadband think of WiMax.

Goal is not to provide wireless Internet, but to provide Internet with all means that are appropriate regarding costs. So, if ADSL is cheaper than WiFi, then ADSL will prevail. WiFi would be used where ADSL is not an option.

If you are making business on broadband Internet, then think of switching to ADSL, or other more appropriate technology.

what's about turbo mode?

IMHO the key to supplying any market is to understand what your competitive advantages are or Unique selling points. You need to understand your environment and why your customer chooses you. If it is only price then you should leave the market because there is nearly always a competitor who is prepared to undercut the price to try and gain a market share.
This applies to anything you are trying to sell matters not wether it is wifi, cars or food. Look at quality (build a reliable network don' connect someone on a weak link just to grab the money), customer service (keep in touch with newsletters), kept promises (adsl is often up to 20Mb but rarely delivers it), coverage areas, speed of delivery, local people to talk to not remote centralised office. Just somethings to think about.

The replies you have had have a similar sentiment. everything changes with time and the cable companies are worried about 3Gand its performance improvements. No market stands still, it is very much about being able to grab he moment.

this is possible,
only one way to compete price war, if you use fast internet caching servers for clint's then you provide more then double speed with same price!

IMHO the key to supplying any market is to understand what your competitive advantages are or Unique selling points. You need to understand your environment and why your customer chooses you. If it is only price then you should leave the market because there is nearly always a competitor who is prepared to undercut the price to try and gain a market share.
This applies to anything you are trying to sell matters not wether it is wifi, cars or food. Look at quality (build a reliable network don' connect someone on a weak link just to grab the money), customer service (keep in touch with newsletters), kept promises (adsl is often up to 20Mb but rarely delivers it), coverage areas, speed of delivery, local people to talk to not remote centralised office. Just somethings to think about.

The replies you have had have a similar sentiment. everything changes with time and the cable companies are worried about 3Gand its performance improvements. No market stands still, it is very much about being able to grab he moment.

+1 very good post

I did some investigation into WIMAX and from what I saw WIMAX still has a way to go. I have read about several WIMAX deployments of which some are successful and some aren't. Case and point http://www.infoworld.com/article/08/03/ ... nce_1.html. These success stories doesn't really impress me much so far, here are my opinions:

Some manufacturer boast of FDD ptmp equipment:
With the little networking knowledge that I have cmsa/cd is not a feature of a full duplex systems, this is why we use switches not hubs. With wireless clients fighting for available airtime I think FDD ptpm a bit far.

Longer ranges and increase throughput:
Laws of physic states that the lower the frequencies the longer the distance but with lower data rates. Within the right circumstances a 2.4ghz system may reach farther than a 3.65ghz but with lower throughput. So to achieve the throughput as boasted by WIMAX may possible within a PTP system and not ptmp which is what we are interested in. WIMAX is amazing for now until everybody start using it. Cell providers migrate from 900mhz to higher frequencies like 1700 and 1900 (sorry my call just drop, just it a dead zone "Can you hear my now"), they do this as 900mhz frequencies was to noisey and prone to interference like a car being started or microwave oven. Increase throughput seems to be a result of MIMO technology similar to 802.11n. Crap I can currently get this with me current setup: Access point as 2 radios one on vertical polarity and the other on horizontal polarity antenna, client side same with bonding. Voila, I have twice increase throughput. The concept of an indoor self installed cpe is great, no truckroll, but what could the possible distance be from the POP.

Increase business oppurtunity:
Really. So far all those who are part of the WIMAX forum are big name service providers who currently have 100 millions of dollar to throw around. This is their opportunity to compete against cable operators to provide IP capable services. I like the fact that they are interested in the technology as we are, but with their presence it is just a matter of time before the small guy gets overwhelm. Base station are ridiculously price, cheapest I have seen so far is over 1500USD, cheapest cpe 270USD(suprisingly) and not available for usage within the US(maybe because of current power limitation).


We all know that WIMAX is here. Will it replace wifi, not according to the WIMAX forum. These two technology will be coexist and compliment each other. So in the mean time I am going to continue with wifi deployments until something else comes along. A carefully plan network deployment can stand the test of time for a long time.

I stand to be corrected.

Why don't you find a cheaper wholesale provider? The big Hungarian telcos are give you very bad prices and want 2-3 year contract. But Internet is a fast changing market so you need to find a partner which better for you.
I think so that the post of buzz57 are very good. You have to find areas where is no dsl, or only lower speed.

Küld el a magic@magicnet.hu-ra, hogy hol, milyen sebességet és mennyiért kaptok most és küldünk egy normális ajánlatot.

gyoztes You can actually one up them on the bandwidth, and equal them in quality if you build the network correctly and carefully. We have been offering 3mbps/256kbps to our residential customers for years with Mikrotik APs and CPEs. jwcn knows first hand the way a well designed (and in this case pure Mikrotik) network can perform and what it can do to competition, as we walked in and took over his entire market area over a year ago inside of a few months time and he no longer has any towers left now, which may explain part of his current animosity towards Mikrotik (http://forum.mikrotik.com/viewtopic.php ... a&start=50).

I'm not going to spend a bunch of time repeating old posts I've made explaining how we build our network (feel free to look them up), but the key technical factors are these (and buzz57 is absolutly correct on the business and customer service side of things):

NStream+Polling to every CPE (sometimes CSMA helps, sometimes it doesn't, test each deployment before deciding)

Psudobridge on the CPE

on the AP set Client TX-Limit for radio links (normally set to 1mbps for all our links) to prevent CPE flooding the AP

PPPoE w/RADIUS authentication and RADIUS defined bandwidth caps for PPPoE sessions (reccomdend using bursting policys)

Have a PC router or large Routerboard (with a decent amt of RAM and CPU) at any tower with more then 25 customers connected to terminate the PPPoE

vlans are your very good friend (there's a visio drawing somewhere on the forum I made with an example)

route everything with OSPF, and have at least 2 routes to every tower you can

either build rings or mesh into your network topology to allow for failure of any tower to prevent outtages of other towers

small subnets sizes

use just the minimum number of OSPF routers on any particular segment required, seperate segments with VLAN's if necessary (if one OSPF router starts acting up, it causes all adjacent routers to do a re-calc, sometimes causing momentary packet loss, so keep the number to a minimum, and absolutly keep redundant links on seperate network segments)

I don't see your problem! We are in exactly same position and here we compete with 8mbps services at $10 (£20)/ month. Actually I feel wireless is in an excellent position to compete! It costs us next to nothing to reach a customer when the DSL providers pay a wholesale fee to the telephone company we do not. They also have much more expensive kit and big overheads. The secret is to have a good bandwidth supply. You can cheat and use the competitions bandwidth to supply your network or like us if your lucky you can reach a data center by wireless and buy bandwidth direct with no transit fees. You see that usually you have to buy bandwidth from your competition if your using leased lines and it's no surprise they try price you out the market. Search for hosting centers in locality or try to team up with a university etc or anywhere which has a good supply. We can offer a pretty good 8mbps service (not guaranteed) shared by 20 or so users per AP? I'm sure we could go a bit more too? It's all in the configuration to keep a good quality of service. I also don't think WiMax is ready yet. It's too expensive to make a business model and with the cost of 802.11a now you might as well just have more base stations! You can be more fexible than big firms with no contracts, pay as you go etc etc.

gyoztes You can actually one up them on the bandwidth, and equal them in quality if you build the network correctly and carefully. We have been offering 3mbps/256kbps to our residential customers for years with Mikrotik APs and CPEs. jwcn knows first hand the way a well designed (and in this case pure Mikrotik) network can perform and what it can do to competition, as we walked in and took over his entire market area over a year ago inside of a few months time and he no longer has any towers left now, which may explain part of his current animosity towards Mikrotik (http://forum.mikrotik.com/viewtopic.php ... a&start=50).

I'm not going to spend a bunch of time repeating old posts I've made explaining how we build our network (feel free to look them up), but the key technical factors are these (and buzz57 is absolutly correct on the business and customer service side of things):

NStream+Polling to every CPE (sometimes CSMA helps, sometimes it doesn't, test each deployment before deciding)

Psudobridge on the CPE

on the AP set Client TX-Limit for radio links (normally set to 1mbps for all our links) to prevent CPE flooding the AP

PPPoE w/RADIUS authentication and RADIUS defined bandwidth caps for PPPoE sessions (reccomdend using bursting policys)

Have a PC router or large Routerboard (with a decent amt of RAM and CPU) at any tower with more then 25 customers connected to terminate the PPPoE

vlans are your very good friend (there's a visio drawing somewhere on the forum I made with an example)

route everything with OSPF, and have at least 2 routes to every tower you can

either build rings or mesh into your network topology to allow for failure of any tower to prevent outtages of other towers

small subnets sizes

use just the minimum number of OSPF routers on any particular segment required, seperate segments with VLAN's if necessary (if one OSPF router starts acting up, it causes all adjacent routers to do a re-calc, sometimes causing momentary packet loss, so keep the number to a minimum, and absolutly keep redundant links on seperate network segments)

How resilient is your setup to lightning?

How resilient is your setup to lightning?

if a tower gets struck, you have to expect at least some equipment will fail (it is over a million volts and hundreds of thousands of amps of electricial discharge), and anyone directly connected to the damaged equipment will be offline, however a partial, or complete equipment failure at a tower will only affect customers attached to that equipment or tower, customers on other towers, will be automatically re-routed in 40 seconds or less depending on the link configuration, via OSPF to use an alternative path on the network.

Last winter a serious ice storm came though and took out or degraded 6 out of our 10 major backhauls to non functioning signal levels, yet despite some performance bottlenecks, 100% of our towers were still online and operational (the same can't be said for the customers connected to those towers, they had bad icing too and many of the weaker signal customers dropped off with 1/2 inch or more of ice coating their CPE).

Lightning? What, me worry?
I live in the "Lightning Capital of the World", Florida.
My most exposed equipment is on top of a 22 story building with no other structures close by.
Here is where it has been for the last season:


The red circle indicates the lightning rod connected to the building lightning suppression system.
The rod has been hit so many times it has blown the screws out of the railing, and the tip is discolored and bent from the heat of the strikes.
The previous wifi company lost not only their equipment, but caused $80,000 worth of damage to the fire suppression system and elevators during the last storm they were there.

I have yet to even replace a radio in the equipment you see less that 6 feet away. Maybe it is just luck?

Maybe it is just luck?

heh, I wish you all the luck in the world... that many hits only 6 ft away, one of these times you're going to see a secondary, or worse primary hit on your equipment...

it's considered common knowledge that lightning never strikes the same place twice, in reality it will often strike near the same areas of electrical least resistance, and less frequently but just as devastating, anything else in those areas that can conduct a charge...

Your equipment falls in the "anything else in those areas that can conduct a charge." category

A primary strike is highly unlikely. No path to ground means no target.
You should watch the Discovery Channel episode on the high tension line repair with the helicopter. It was very en- "lightning".
Then explain to me why his radios still worked after the repair. "OK, we are ready to move on to our next repair" the pilot calls in to the office...

Actually, I was very concerned about the possibility of flashover, so I intentionally positioned this next to what appeared to be the busiest lightning rod of the group. There were two visual reports of direct hits on that rod during a storm this season by building security on the night shift.

nice info, so i do have one question, i have been thinking about how you properly ground the ethernet with poe(wether you do it inside the case or outside or in the building or both) as well as the antenna's and the pc board inside and what guage wire to use, anyone have and setups they have used?

ForePoint,

For the record, I have never had or operated or had any involvement with a WISP anywhere remotely close to your location - you have me mixed up with someone else.

We have 1,000+/- customers and 30+ towers on the east coast, never even gave thought to the mid-west.

ForePoint,

For the record, I have never had or operated or had any involvement with a WISP anywhere remotely close to your location - you have me mixed up with someone else.

We have 1,000+/- customers and 30+ towers on the east coast, never even gave thought to the mid-west.

then someone here is claiming to use your username on the forum


edit: i wouldn't at all be surprised from him to see him lie and take credit for your posts as his own... he once pointed at a 100ft tower we were 50 ft away from and told me he had an omni and a 25dBi 5GHz panel on the tower, I was looking with a pair of 8x binoculars, it wasn't there.... he also tried to sell me a tower contract with 25 customers on it, for a tower that hadn't been built, and the "customers" were just people that lived within 2 miles of the tower and could be interrested if the tower was ever built (he knew a tower 2 states away that could be purchased and taken down and transported here to be built)..

assuming you're telling the truth, I am sorry for confusing you with someone else, however what I said about your posts still stands.

try going to jwcn.biz Actually that redirects to the wrong page on the domain but you'll get the point.

vlans are your very good friend (there's a visio drawing somewhere on the forum I made with an example)

Forepoint .... i searched ur posts ... could not find ur drawing ...

is there anyway u could let me chk it out .... ??

txs