Hotspot in a Box

Ive finally managed to put it all together! I believe I have managed to build the ultimate embedded hotspot-in-a-box solution.
What I have done is, put together Chillispot, FreeRadius and phpMyPrepaid on the same box. While this itself is not unique, what separates this from other such initiatives is that all of this comes on a 512MB Compact Flash Disk.
The basic platform is a Soekris Engineering net4801 appliance. This has a decent 128MB of RAM and three Ethernet Interfaces. More details on the Soekris box are here.
The challenge was to look for a Linux distribution that recognised all the hardware on the Soekris, and was small enough to fit on a CF Card. My search lead me to Voyage Linux which is a debian derived distribution with a 66 MB footprint in its basic avatar.
Since I didnt have a CF Card reader, I used my Nikon Coolpix n770 camera as the reader. Attached it to my Ubuntu laptop and the card showed up as a usb disk. Using the voyage installer was a trivial exercise. Once the basic OS was installed, I then plugged the CF Card into the Soekris and watched it boot beautifully over the serial console!!
Once this was done, I connected the Soekris to the Internet and ran apt-update to update the package list.
I then went about installing MySQL, PHP, Lighttpd (lighty) instead of Apache, FreeRadius, Chillispot and Phpmyprepaid.
After two days of fiddling around, I figured out why Chillispot wasnt working properly, rectified it (eth1 has to be 0.0.0.0).
I now have a full fledged hotspot controller which does not need to connect to any centralized Radius Server for authentication.
The Soekris box is currently undergoing burn-in.

The only problem I am facing is that of MySQL tables crashing! I live in an area where power outages are the norm. Thank you DLF for wonderful buildings and crappy infrastructure.
High on my to-do list is change over from MySQL to firebirdSQL (www.firebirdsql.org). Wonder how long will it take to modify phpmyprepaid to work on firebirdsql...

For all your uber-geeks out there, I can send you a copy of the CF card image if you feel a little adventurous. #free -m reports
124 M Total 59M Used 64M Free
The total size of software on CF card is around 370 MB.

Wi-Fi in Trains

Wi-Fi Networking News reported here that Icomera, a Swedish company is rolling out Wi-Fi in UK Train Operator GNER's trains. I called up the Icomera people and spoke with a nice gentleman named Jan Lindberg who said he was in charge of International Business. I asked him whether Icomera was interested in coming to India to roll out Wi-Fi in Indian trains. His answer completely blew me off! He said very apologetically that Icomera was a very small company (about 20 people) and did not have enough bandwidth to go out of Europe. So they don't have any business anywhere except Europe in the "foreseeable future" as he put it.
I really wish some Indian giant takes cognizance of this and goes out and buys these guys out so that we can benefit from their technology. We have after all bought out Corus and Arcelor havent we?

Campus WiFi Tenders: Realistically Speaking..

The other day, a prominent government lab put up a notice inviting tenders for Wi-Fi connectivity at their residential colony. We promptly downloaded the tender specs from their website. The tender appeared to be well written with a lot of care taken to ensure that only the best solution gets selected. On closer look, I was appalled to see some of the so-called specifications! It indicated an absolute lack of expertise in Wi-Fi network capabilities, and technologies. The most damaging spec read as follows:-
" Only one site (also called base station) to be used to install all access points/repeaters and antennae and provide access throughout the campus."
The campus consists of 50 odd buildings, each of 2-4 stories height and medium vegetation. The campus itself is about 1 sq km in size. I decided that the place was worth a visit and some discussions with the concerned people was definitely in order.
The laboratory (like most Indian government institutions) had an antiquated IT department. They even had a fancy name for it: "Computer Facility Center". We couldnt meet with the Head and instead two subordinate "Scientists" met with us.
I basically asked to be shown around the colony where coverage was desired. One of the gentlemen accompanied us to the Residential colony which was across the road from the lab. On seeing the largest building there (about 12 floors), I went up to the roof of this building to get a Bird's Eye view of the entire campus. What I saw wasnt very encouraging. The building was at one corner of the campus and there was thick tree cover all over. I quickly realized that there was no way that anyone could provide coverage to the entire area with just one "base station". We then went back to the IT department for discussions. I explained to the two people that they must release an addendum to the tender notification doing away with the requirement for a single base station. The explanation given by them for this totally blew me away! They said that they didnt want to be bothered with the maintenance of the Wi-Fi network and hence if a single base station were used, only one point would be required for maintaining the equipment.
This even after another requirement in the tender that the prospective bidder had to provide support services for 3 years!
I tried explaining to them the fallacy of the logic. Moreover, I tried to tell him that there was no technology that could provide access to laptops and PCs in the colony with just one base station mounted on one building at one corner of the area.
They just refused to listen to reason! They said that it was too late to change the tender and it would only involve additional paperwork and too much trouble and other such blah blah. They had not even heard of Mesh networking. They kept insisting that there were some experts who had assured them that they were in possession of such equipment which could fulfill this requirement.
After all this, I am left with the thought that woe be to the poor guy who does win the bid. How would he go about providing access?

All of the above just goes to say that the people/end users who have the grave responsibility of ensuring that the organisation gets the best solution based on sound technological inputs choose to behave immaturely. I really wish people would take the trouble of reading up on technologies instead of designing it for the system integrator! What they dont seem to understand is that every Wi-Fi vendor has his own architecture/philosophy of designing a wireless network. By specifying a particular approach, they have effectively ensured that only a couple of vendors can qualify. The least that could have been done was to have appoint a consulting organisation which would have helped the Laboratory make a technologically sound RFP which can be realistically delivered by the industry.

I'm off to hunt for a solution that can do all that they're asking for at the lowest cost!!

Freakquency: Meraki Wireless Repeater Makes Extending WiFi Easy!

Freakquency: Meraki Wireless Repeater Makes Extending WiFi Easy!

Meghalaya goes for Wi-Fi hotspots

The beautiful state of Meghalaya in India's North-east has announced plans to set up Free Wi-Fi hotspots across Shillong, the capital city.
Mr Robert Garnett Lyngdoh, the IT minister of Meghalaya says that Meghalaya will be the first state in India to offer free Wi-Fi to tourists and locals for non-commercial use.
Shillong is a tourist paradise and is more westernised than most other Indian cities. This step is one of the many initiatives in India towards promoting the use of Wi-Fi to bridge the Digital Divide.

Plan afoot to make entire Delhi Wi-Fi

The Delhi edition of the "Times of India" reported yesterday that taking a cue from the other cities like Bangalore (now Bengaluru) and Pune which are already rolling out Pilot Wi-Fi networks,
The city government's information technology department is exploring options to ensure Delhiites have wireless net access on laptops

A Delhi Government official is also quoted in the article as follows

A senior IT department official said, "We are looking at a model that integrates all the existing technologies like GPRS, Wi-Fi and broadband. We are also looking at WiMax, which is the latest in the field. The final decision will be taken on the basis of the economics involved. But right now, our focus in on providing high-speed internet connectivity to all in an unwired way. This will come at a cost. But once the government enters the picture, the cost of internet access for an individual will come down drastically."

This is a very welcome step forward for a city starved of high speed wireless broadband access!! However, there seems to be a tendency amongst the so-called telecomunications experts in the Government to look down upon Wi-Fi and at the same time be in awe of WiMax! I attended a meeting recently in which a very senior telecommunications officer in the Government innocently asked of the rest of us:

" Is the Wi-Fi technology standardized like WiMax? Does it have a forum like the WiMax forum?"

Like the old song "I didnt know whether to stand there or run!!"
There are also many factual irregularities in the article, which can be attributed to the reporter's lack of research or knowledge or both!
For example, she creates a major blooper in this paragraph which describes Wi-Fi:

It is a short-range system covering many hundreds of metres. It uses a licensed bandwidth to provide access to a network — typically used by the end-user to access their own network, which may or may not be connected to the net.

Whatever the above paragraph means, it conveys a very wrong message to the community at large. This is how I posted a reply on the website (though it has a very bleak chance of getting posted):

"Dear Abantika,
You are wrong about Wi-Fi using licensed bandwidth. It doesnt. In fact due to the pioneering efforts of people in India such as Dr Arun Mehta, both the 2.4GHz and the 5.8 Ghz spectrum on which the Wi-Fi networks can operate (802.11b/g and 802.11a resp) have been delicensed for both indoor and outdoor use by the Government. This is to enable the common man to setup and use the wi-fi spectrum. Articles like yours only serve the interests of the incumbent telcos who are pushing the WiMax/3G agenda while being fully aware that all of what these not-yet-proven technologies can do is already in production use in many countries with Wi-Fi."

Largest WiFi Network in the World will be in West Bengal, India

Move over Philadelphia and San Francisco; the worlds largest Wi-Fi network is being built right here in India.
It has been reported in Biztech India (among other media) that "smartBridges-SREI To Set Up Wi-Fi Network In WB, India". The article goes on to state that:

smartBridges and SREI have announced an alliance to provide low-cost technology enabled Common Service Centres (CSCs) in West Bengal, India. smartBridges will provide the wireless infrastructure to set up 4937 CSCs in 14 districts of West Bengal while SREI will manage these centres and develop a self-sustaining business model.

This is bigger than muni-wireless; this is state-wide wireless! What is more interesting is that the article further goes on to explain how the network will be used:

These CSCs will serve as a revenue tool for village entrepreneurs as well as several self help groups. While providing additional employment, services like VoIP, eEducation, eMedicine and other technology based solutions will also be implemented to improve the quality of life of the rural population.

This is a classic example of how Wireless technologies can and are being used to bridge the so-called "Digital Divide". It has always been emphasized that Wi-Fi is not just about providing Internet connectivity which is a by itself a natural by-product. With Wi-Fi so much more can be achieved. It is common knowledge that the incumbent telcos including the state owned ones have miserably failed in their obligations to provide connectivity in the rural areas. Instead, they are comfortable giving a percentage of their revenues to a fund known as the USO (Universal Services Obligation) fund. This is become a joke because the fund has now grown to a few billion dollars which is what could have been used to provide the connectivity in the first place. Hence these initiatives are part of the Indian Governments attempts at spending the billions that have accumulated over a period. The West Bengal project will slowly be replicated across this vast country. Though we will not have a massive Wi-Fi cloud over any state or even a district (The connectivity model in these cases is mostly point-to-multipoint) which is fine, since we dont expect every rural Indian to be toting a laptop or other computing device. What is more important is that with back-haul reaching the CSC at every village, the village level entrepreneur can set up his own profit-earning hotspot with business models like the "Tomizone" model.

There are other parties which are experimenting with wireless technologies like 700 Mhz Wireless DOCSIS (equipment by Axcera/VCOM), WiMax and so on.
India is the right place to be if one is in the wireless business!

Setting up a Wi-Fi Hotspot: Profitable yet Affordable. Part-II

Continuing from where we left off...
I will attempt to discuss how a Wi-Fi hotspot can be built using cheap hardware ( Linksys WRT54GL - Rs 4100, Standard Barebones PC - Rs 15,000) and some free opensource tools. Though it is a non-trivial task, the fruits of labor are well earned!
To start off, lets identify the components of our hotspot:

Sl No	Component	Our Solution	Website
1	Wireless Router	Linksys WRT54GL	http://www.linksys.com
2	Captive Gateway	Coova	http://www.coova.org
3	AAA Server	Freeradius	http://www.freeradius.org
4	Subscriber Management Software	phpMyPrepaid	phpmyprepaid

The following figure (from Chillispot) succinctly illustrates our setup.

To keep this post short and also to avoid making it sound like rocket science, I will focus upon a solution that offers a readymade package which is a combination of Sl No 1 & 2 above. As some people say, its like your Linksys router on Steroids!

Step 1: Preparing the Hotspot Controller

The first step is to convert a commonly available wireless router into a Hotspot Controller. To do this, we have to perform a little surgery on the WRT54GL Linksys router. Remember that this will void the warranty on the router. Though it will not go up in smoke, you might end up just "bricking" the router.

Download the latest version of the Coova firmware from here. Connect your laptop/PC to the WRT54GL with the blue patch cord that came with the router. Remember that you need to connect to one of the LAN ports and not the "WAN" or Internet port. The Linksys will assign an IP address to your PC in the series 192.168.1.x. This means that your PC will in all probability get an IP address of 192.168.1.2. Now fire up your favorite browser and type in http://192.168.1.1/ in the address bar.
You should be presented with a authentication window. Type in "admin" as the username and "admin" as the password. You should be logged in into the WRT54Gl management console. Now navigate to the Administration-->Firmware Upgrade tab. Click on the Choose FIle button. You will be presented with a File Chooser dialog box. Now point it to the firmware file that you downloaded eg. openwrt-wrt54g-squashfs.bin on your PC. When you now click on Upgrade the progress bar will indicate the upgrade in progress. Remember not to disturb this actvity as it may lead to "Bricking" your router!
Once the upgrade is over (it might take a couple of minutes), the router will reboot. Once the reboot is over (the DMZ led on the router will go off), you will be presented with the Coova web interface. It would be a good idea to connect your internet connection (DSL etc) to the "Internet" port on the router at this stage. This is to enable automatic download and installation of additional components by the new firmware. Now Coova is a project built upon the now-legendary OpenWRT project. It takes out the pain of configuring OpenWRT with Chillispot etc and presents everything in a nice Interface.
Enter the new password for the Coova Router (notice how we re-christen it?). I would suggest that you follow this link to setup the Coova router. It has some neat screenshots to ensure you recognize what you are doing.
Once you have reached this point in the above installation wiki, that says "Once the settings are applied, the router will be a HotSpot", you are all set to go!
Coova offers three ways of setting up a hotspot, Internal, Chillispot and WiFiDog. We will explore the first two.

Step 2: Testing the Hotspot Controller

Lets now perform some simple tests to ensure our upgrade went off well and the router is working as expected.

Coova as standard router

Your laptop should be able to connect to the Internet while being connected to the Coova router wirelessly.

Coova with internal hotspot

Coova offers an embedded internal hotspot which allows you to create a table of internal users with a limited edition 'landing page'
Create a couple of users and enable the Internal hotspot. Now when you try to connect to the Internet through Coova, it should redirect you to the 'Landing page'. One you enter the username and password correctly, It should permit you to surf the Internet. For many applications like libraries etc, this is more than sufficient since you are not permitting people to access the Internet without registering for it (or you knowing about it). The obvious pitfalls of such a mechanism is that there is no session timeouts, users will have to be logged out manually and so on.

Coova as chillispot hotspot controller

This also has two options. Automatic settings as well as manual. With automatic settings, you will be redirected to a simple ToS page hosted on Coova's servers.
Manual settings are what we are interested in. This is one place that I found that the Coova wiki does not document at all!
We will explore Coova manual settings in the next part because we have to satisfy some pre-requisites before this can be achieved. This involves setting up of the components mentioned in Sl No 3 & 4 in the table above. I will outline how these can be installed and tested on a standard Linux PC (Though I am given to understand that this can be acheived with a Windows machine, I am a Linux guy and cant tell you much about that part!). Stay tuned for Part 3 of this series...

Wi-Fi in Gurgaon: A Tall Tale?

We read in the papers recently that "DLF to get Wi-Fi connectivity: Cyber City to get it first, later residents to enjoy benefit".
The article goes on to quote Mr AS Minocha , Chairman of DLF Commercial who says that they are in talks with two American companies for deploying wi-fi throughout DLF Cyber-city.
Well, better late than never!
I still recall meeting with the DLF top brass some 8 months back and proposing for a full city (DLF has developed a mini city within Gurgaon, a suburb of New Delhi) wi-fi network. The value proposition there was focused on providing IP Video Surveillance (because Gurgaon is notorious for lack of law & order). We said that Wi-Fi for Internet access would be a natural by-product. There would be so much surplus bandwidth that they could invite multiple service providers to ride their services on the network. We were literally brushed aside by them. The reason given: "DLF does not believe in investing in infrastructure; However, we will permit you to install your network and also charge a fixed revenue per month for this right of way".
What a load of baloney! We had to naturally drop the idea altogether and wanted to sit back and wait for some intelligence to creep into the organization.
It appears that after their mega IPO this month, they really have had some fresh ideas. This one about Wi-Fi is more than welcome! I am moving house to DLF Phase-1 next month. Lets see how the service (If it is in place by then) performs!
The article also quotes Mr Manocha as saying "at a later stage, routers would be installed at residential colonies and people living in proximity of that access point would get automatic connectivity. This means that within the next few months people living in the DLF city will no longer need wire connectivity to access Internet services. Any person with a wi-fi enabled device such as a PC, cell phone or PDA can connect to the Internet when in proximity of an access point".
Phew! only if it were so easy, people in places like Mountain View, Tempe wouldnt be complaining!

Mesh Networking in India: Do the Big Guns have a strategy?

When I read about the plans of the large mesh technology companies, I often wonder about their plans for markets such as India.
I have had the opportunity to be closely associated with Belair Networks, Strix Systems and RoamAD in India.
I have deployed Belair's first mesh network in Mumbai, India, been involved in a Strix Demo at the Presidential palace and very recently involved in the Proof of Concept that RoamAD has deployed in the a large city in the west of the country.

I will attempt to analyze what their strategies could be, based upon my personal reading of their activities in the country for the last 2 years.

Belair Networks. I have admired Belair for the simplicity of design, and the tremendous technology that goes into their products. Now I have often wondered why such a company would totally ignore the Indian market. Their sole attempt at coming close to the subcontinent has been limited to setting up a sales/support office manned by two people in Singapore. All their attempts at trying to appoint distributors in the country have come to nought. At times, I wonder whether they tried hard enough. I think the primary reason for this is the lack of a strategic vision for this market and Singapore is far more mature as a market than India. Someone who has sold in Singapore will never do that in India. What they dont realize is that the Indian market is a volumes-driven market. Even MNCs like Pepsi and Coke have had to re-write their marketing/sales strategies for this market.
In my opinion, Belair has not addressed the following issues: -

a) Creating brand awareness
b) Creating a special pricing strategy
c) Creating a supply chain
d) Creating a support infrastructure.

I have often been told by Belair executives that the Belair product line is not comparable with either Cisco or Motorola. What they dont realize is that these two companies have existed in India for the last 10-15 years and have a solid brand reacall. People would rather buy a single or dual radio Cisco/Motorola than go for a product from a Startup that is 20 times more expensive!
I have personally seen that even though Belair offers a 3 year warranty on their products, it is next to impossible using this warranty. I tried in vain to send back a BA200 back to their facility in Canada. The BA200 has still not left the country after 8 months beacuse the rules in India state that a Belair engineer in India has to certify that the product cannot be repaired within the country! Also the shipping constitutes as "re-export" which involves too much paperwork to even attempt. The end result was that the BA200 that we hold has gone to the C&F agent twice and come back with some observation or the other. It is now lying at our offices unusable when the only thing that might have gone wrong with it is the power supply module.
The biggest problem with the Belair products is that there are no field/India serviceable parts. One cannot even change a blown fuse (simply because there is no external interface to do this). One of the ways that they could bring the price down would be to source most of the parts in India or at least assemble the nodes here. I happen to know for a fact that the software that runs inside the Belair Nodes is made by a company out of Bangalore!
I happened to attend a webinar yesterday organised by Belair. One of the questions that I asked and which went unanswered was predictably "What is your strategy for India?"
Though they were quick to answer my next question "How does Belair compare with RoamAD" with "We dont consider RoamAD as a serious threat because they are a 2-radio platfom and we are a 4-radio and upwards platform". I then pointed out that the RoamAD nodes we were using in India were all in fact 4 Dual-band Radio nodes each !!!

Strix Systems

Strix has an interesting approach to the Mesh network. They dont bundle antennas at all! The advantage is that the integrator is free to decide what antennae will best suit the purpose but setting up a Strix Network is a pain. I also heard on the Grapevine that Strix is phasing out the Indoor Mesh product line. Strix's India strategy has been very aggresive. Though not directly present (not even through a representative office), they found a distributor/system integrator called AMI India who invested heavily in Strix equipment. Now AMI has had some amazing wins in Mesh thanks to some strong marketing. There is now a second company called Wireless Tech Pacific (out of Singapore initially) who are distributors for (hold your breath) Tropos, Strix and Motorola! Its anybody's guess which products get pushed more.

RoamAD This is one company that is certainly worth watching very closely. Though they started off as a mesh hardware company, they then changed course midway and are now a pure-play mesh software company. The USP is that the integrator is now free to assemble his own nodes using COTS (Commercially Off-the-shelf) or commodity hardware. This is a sure-win situation for a country like India. Now, one does not have to ship any nodes back and forth across the world for repairs.
I have deployed a RoamAD mesh in India and have to say Im impressed by the technology behind it. I have also been impressed by the fact that RoamAD's CEO and CTO have both made visits to India (at least twice each) to ensure that the pilot networks go off well. Now this is a very positive sign as far as strategy goes. RoamAD has also quickly snapped up Raychem RPG (a very well respected name in telecommunications) as their main distributor/integrator. Network tests on the RoamAD system are very encouraging as the software is built from the ground up keeping Voice and Video applications in mind.
The icing on the cake? RoamAD software along with hardware from ADI/Acure costs about 1/10th of competing vendors. You can't get a better combination than this for a price-sensitive market like India.

I havent remarked upon the other companies like Motorola, Tropos, Skypilot, and Cisco because they have entered the market through the traditional route of appointing channel partners who are more into the business of selling
indvidual boxes rather than providing solutions. It remains to be seen how they fare in the months ahead.

3G is dead! Long Live Wi-Fi

We read in the news yesterday that the Minister of Information Technology & Communications Mr A Raja has advised BSNL to drop 3G from their forthcoming expansion plans and go ahead with existing 2G (or whatever its called). The news also mentions that he has questioned why Motorola was disqualified from the 45 million lines tender.
This is like music to my ears! For a long time the incumbent telco lobby has been stymying all attempts for Wi-Fi to grow in India.
It was common for people in very high places to tout 3G as the answer to all our connectivity woes. It is hard enough trying to get a word edge-wise (cell wise?) in many metros these days with abysmal connectivity on GSM/CDMA cellular services.
Now that 3G is apparently put into cold storage, we can all get down to the business of doing some city-wide wireless.
Some real headway is being made in the city of Pune where Microsense has won the contract for unwiring around 40 sq km.
The Bangalore project seems to be in limbo for reasons unkown.

Setting up a Wi-Fi Hotspot: Profitable and yet Affordable

Wi-Fi Hotspots are great business... or so people thought. It sounds so attractive, you connect a wireless access point/router to your "home" DSL line and hey presto! people would come in droves to your establishment to surf, be online whatever and you made a lot of money.
Huh?

Whats the revenue model here? Do you charge for access? Do you give it away for free? If its the latter, would you be arrested if someone abused your service? Are there people/services available who would help you set up such a service?
While it is true that offering Wi-Fi by itself is not the way to making your millions, you can make a start by offering Wi-Fi in a controlled environment. This means that you control: -

Who connects to the hotspot (Authentication)

What they can access through the hotspot (Authorization)

How long they use the service (Accounting)

In addition to this, there also needs to be a mechanism to ensure that adequate bandwidth is available for all users (no mp3 and DVD downloads choking others).

All of the above is available as a managed service offered by companies like Pronto Networks, Airpath, Aptilo, Nomadix and so many others.

The advantages of going in for these services is obvious; You sign up with the company and become a "WISP" (Wireless Internet Service Provider).

They give you a nifty little device called the "Hotspot Controller". This device sits between the Internet connection (DSL, T1) and the rest of your wireless network. At times, this lil box may be the rest of the network because your hotspot might not need any more access points than the one bundled with the Hotspot Controller.

Now the Hotspot Controller will typically connect to the OSS (Operations Support System) of the provider. It can then manage, provision your network. The OSS has a web interface that the WISP (you) login to and create packages, prepaid coupons and so on.

It also lets you customize the "landing page" on your hotspot so you can use that space for advertisement revenue!

When the prospective customer (read road warrior/kid across the road with laptop) tries to get on to the Internet through your network, she is redirected to the "Landing Page" which asks her to "Pay up or else..". In other words, she needs to "sign-in" or "sign-up" for access.

There usually are cool features like letting her sign up for the account by creating an account and paying via credit card and so on.

All this is plain and simple and can be done in a jiffy. But, is this affordable?

Well not really. You (as the WISP) have to pay the managed services provider a flat upfront fee, and then a portion of the revenue per subscriber.

Now what happens when there is no customer at all? Well, you pay the provider a minimum guarantee to ensure he gets back his investment in your business.

There are services that follow the middle-path like not charging you a fee for the hotspot controller but charging a portion of the revenue etc. These include Zonerider, FON.

Now, coming back to the main issue we started out to discuss. Is there a way to get all the goodies that the OSS provider provides and also not lose money? More importantly is there a way to make money?

In part 2 of this series, we will look at how this can be done using some very cool tools available in the Open source world.

Contd....

Wi-Fi in Hotels: A sustainable model?

In india, the incumbent telcos have more or less killed the Wi-Fi in 5-star Hotel markets by putting in the network for free and then sharing revenue with the Hotel. This apparently is a win-win situation for the Hotel since there is no up-front investment and there is also a steady revenue.
The flip side is that in their greed to earn revenue, the hotels end up charging astronomical sums for access. The typical rates are about INR 900 ($ 20) per day!
The operator on his part tries to cut costs by using sub-standard equipment and inferior design leading to lack of proper connectivity in the hotel.
This has led to a lot of people (domestic business travellers) using their GPRS cards for Internet access.