|
|
Insight
Wireless
network in the Himalayas
 |
|
David
R Huges
Colonel (Ret), Old Colorado City
Communications, Colorado Springs, Colorado, USA
dave@oldcolo.com |
|
|
A
personal narrative of setting up and operation of the world's highest
cybercafe and its application in the region....
opening up opportunities for promoting tourism and education.
“If
you can connect up by wireless from Everest, you ought to be able to connect
up anywhere”
Thus quipped a writer from
the New York Times when she researched her story about the efforts Tsering
Sherpa and Dave Hughes were going to make to operate the world’s highest
‘Cyber Café” at the 18,000 foot level Base Camp in May, 2003. This was
to support climbers during the 50th
Anniversary of the first climb of Mount Everest by Sir Edmund Hillary and
Tenzing Norgay some 50 years ago.
While that Everest Cyber Café
operation made world news when all eyes were on the large number of climbers
attempting the summit in May, a potentially far more significant wireless
project was started in October further down at 12,000 feet in the Sherpa
village of Namche Bazar. This article is about that effort to link up
wirelessly, first, all the lodges and businesses in Namche – the last
place where climbers and trekkers rest and acclimate themselves to the
altitude before pressing on to their higher mountain goals. And then to
launch the first very high country ‘distance learning’ venture for very
poor Sherpa children whose education is very limited at best.
Beginnings
The
roots of the Sherpa Wireless project were in the vision of 39 year old Tsering,
a native of Namche, who had gotten part of his early education in Darjeeling,
and through his travels and work in Asia and America where he saw the power
and potential of the Internet. He was determined to bring that power to his
home village and people, even though they lived in remote mountains where
communications is very difficult, unreliable, and expensive. And some of it
had already been destroyed by the Maoists who were blowing up Nepalese
government telephone company towers, destroying the last link to Namche.
He already, by 2002, had taken
over some communications equipment others had started to use, including a
small radio-telephone network based on licensed frequencies, and completed his
own ‘Cyber Café’ in Namche, with three desk top computers, and a
satellite system that carried 128kbps of TCP/IP signal down to Kathmandu, 100
miles away, where it linked to the rest of the world via Worldlink Internet
Service.
It was in the spring of 2002
when an American named Gordon Cook, who was interested in the Himalayas,
visited Namche, heard Tsering’s vision and introduced him to me over the
Internet.
Evolution of an Idea
I had
done both community networking work in places like Montana, where I designed
ways to link 114 one-room school houses to the world, and then extended
wireless for the National Science Foundation to places including Ulaanbatar,
Mongolia, Alaska, the jungle of Puerto Rico and my own Old Colorado City. I
was more than familiar with what it would take for Tsering to realize his
vision.
I also knew, from my
philosophy of ‘fitting technology to cultures’ and not ‘cultures to
technology’ that I needed to learn a lot about the Sherpa, and Nepalese
cultures, their economics, and politics as it would affect Namche if I were to
be able to give the right kind of advice on how to make the networking useful
to the economy, education, and culture of the Sherpas. At the simplest
technical level, operating computers or radios at 12,000 feet in Nepal is
little different from operating and maintaining them in high Colorado. But
from that point on – the language differences, the levels of education and
experience with computers or other technical devices - in a village which had
never used telephones, or had electricity, much less ‘personal’ computers,
would make it a steep learning curve.
I set about learning by email,
which Tsering could use, writing English, although it was a chore for him, and
by study of books and on the Internet about the Sherpa culture and Nepalese
Kingdom, and from information from Gordon Cook from his previous trips there.
First, Base Camp
In
early 2003, a novel idea was hatched by Tsering Sherpa and the Namche based
manager of the Sargamatha Pollution Control Committee – SPCC – a partially
Nepalese-government funded NGO which had the responsibility for ‘managing’
the ‘Park’ and travellers. They wanted to set up the world’s highest
‘Cyber Café’ at the 18,000 foot level Base Camp used by all the climbers
going up from the Nepalese side of Everest. It would serve, for a fee, the
hundreds of climbers and their expedition parties, the press, and medical
staffs that would converge on the Base Camp in May 2003. The profits, if any,
would be split between SPCC, Wordlink, and Tsering’s Namche Technical
Services. It would be a benefit to SPCC.
Tsering recruited Dileep
Agrawl, owner of Worldlink, Nepal’s largest Internet Service Provider in
Kathmandu, who had much experience with Satellite Internet links and could
provide the Satellite base unit.

Picture of Namche with wireless
overlay lines
The plan was to set up the
Satellite Base Station at the Base Camp, linked to the Internet via a
geosynchronous satellite
and provide e-mail, web browser and perhaps even voice over the Internet to
the climbers and to their support personnel, camped for a month or more at
Base Camp at the foot of the great Ice Fall at 18,000 feet, from where the
hard technical climb of the mountain starts.
Simple concept. Right? Wrong!
There were plenty of obstacles. The Base Camp sits on the Khumbu Glacier that
moves between three and four feet a day down the flat valley.
Keeping the satellite aligned and linked would be a constant headache.
Secondly, Base Camp was so low in the valley that the Nuptse and Lhotse
mountain masses blocked the best line of sight to the satellite low in the
southern sky. They would have to peek through the passes between them, which
often would be thick with clouds – which could attenuate the signal or even
block it entirely during mountain storms.
So Tsering contacted me in
Colorado. The solution was, to me, straight forward in concept. Place the
Satellite Base unit on the other side of the Khumbu valley on higher, solid
ground that sloped up toward 23,490 foot Pumori mountain, and link the Base
Camp to the Satellite Base by a 4 kilometer 802.11b radio wireless data link
between a radio at the satellite, and the other in the Cyber Café. Each radio
would be fitted with a 13.5 dB gain directional antenna. All powered by plenty
of batteries and solar panels packed up there on the backs of Yaks and Sherpa
porters.
I persuaded Cisco Corporation
to donate three Cisco Aeronet Wi-Fi 350 Access Point radios with the antennas.
They got the radios to Nepal just in time.
The long and short of it was
that solution worked, after taking 18 yaks to carry everything up from Namche.
There were many minor problems, all solved. Doing the simplest things was
hard, slow, work at that altitude. The Cyber Café in the tent at Base Camp
provided, at $1 a minute, the kind of digital link to the rest of the world
that even costly satellite voice telephones brought by many parties could not.
In one case a climber manager said the Base Camp Cyber Café saved him $10,000
in communications costs.
There was tremendous world
wide publicity for several months over the ‘World’s Highest Cyber Café’
operating during the two short climbing season months. Tsering became famous.
And I did not even have to go there. I just give advice and evaluated
feedback, by email.
Wireless Namche
All
the above was preparation - a sort of rehearsal, and lessons learned - for
what Tsering Sherpa really wanted – which was the wireless connecting up of
Sherpas in his 12,000 foot high Namche community. Which is the village on the
trail to Everest, which all Trekkers and Climbers use as the stopping-over
place to acclimate themselves to the altitude, rest, and organize before
pressing on. Its their last place to communicate from.
I had gotten to know Tsering,
and his vision for his people, and understood the economic problems that
Namche faced, with boom and bust tourism, overdependence on one source of
revenue, and only six months of activity. And the fact that the children of
Namche and the surrounding villages were still – in spite of many schools
having been built by the Hillary Trust – poorly educated. Most of the
children faced a life of portering heavy loads for wealthy Westerners up and
down the Solo Khumbu for $5.00 a day. They had little future.
So during the summer of 2003 I
worked out the details with Tsering by email and occasional VOIP calls from
his Cyber Café of what it would take to extend wireless broadband links to
many Lodges where trekkers and climbers stayed from the location of his Namche
Cyber Café and its satellite base station, and 4 desktop computers inside.

Tsering Serpa and David Huges
Planning wireless layout
Especially I studied both
topographic maps of Namche and the surrounding area, and photographs from the
high country, and some aircraft, from above. For the key would be the validity
of my rough, distant, ‘wireless site survey’ for how many of what type
radios, connectors, cables, power supplies, and all the myriad bits and pieces
I had to take in there in order to reach the whole village. It would be a two
day trek with Sherpa porters, to get anything up there from the nearest small
aircraft landing field, and being over 100 miles from Kathmandu would be too
far to get something overlooked and needed, even if it could be procured
there.
My boots on the ground
military experience going back 50 years to the Korean War, from reading maps
to planning operations with attention to detail, and my 15 years experience
specifically deploying unlicensed spread spectrum radios in remote areas
helped immensely.
One look at the aerial view of
Namche shows its unique topography. A high bowl with buildings, including the
lodges, terraced upward from the central “Tibetan Marketplace’ (thus the
name Namche Bazar) with a Buddhist Stupa at one end – a drop off precipice
to the valley far below behind it, and a giant very steep mountain 2km across
from opening of the bowl.
At once that told me that if
an Access Point base radio could be placed right, radios in the lodges could
‘see’ the base radio at a pretty short range – less than a mile. It also
told me the best place would be right on the sheer face of that facing
mountain. A daunting proposition. And that the roofs of the lodges, being made
of tin, would be bad to try and punch through from above. so the 2.4ghz
signals should try to go through the concrete walls on the sides.
I knew that the Cisco radios I
had gotten for the one-time Base Camp project would not be the best radios for
the Namche project or for future expansion in Sherpa villages by Tsering. I
selected a brand I had some experience and knowledge about - Wi-Fi 802.11b
radios from Smartbridges of Singapore. I picked them because:
They were rated outside to -40
degrees F. The Cisco’s only went to 0. They operated with Power over
Ethernet POE cabling, which could work with either battery or grid power
They had excellent receiver
sensitivity
They could be run on 12volts
DC – plain vehicle-class battery. The Cisco’s required 24 to 48v. Multiple
batteries.
They were simpler to configure
than the Cisco’s. Nothing fancy.
They had standard ‘N’
connectors right at the radio – separate pigtails were not required to link
with external antennas.
Finally they were much less
expensive than the Cisco family, from a high of $330 down to $107 for USB port
driven client radio for use in the Lodges. From one third to one half the cost
of Ciscos and similar quality corporate radios. And they could come from a
closer Asian source than from US manufacturers, Singapore.
I pledged to donate $3,000
worth of Smartbridges with associated hardware and fixtures including signal
splitters, battery regulators, cable tools to the project.
There was something else I
felt we needed. If we were going to connect up a Sherpa school to the net, and
try to - through ‘Distant Learning’ techniques - teach Sherpa children
English, Math, Science and other subjects beyond the materials or skills of
the Nepalese teachers who generally ran the schools in the high country, we
needed reliable voice sound. Tsering had made it clear that unless the
children learned English they could not even qualify as guides for Trekkers or
climbers.
So I once again contacted
Cisco and asked them if they would donate four Voice over IP (VOIP) telephone
sets to the cause, even though I had not bought their radios. And would they
configure them for SIP (Session Initation Protocol) – point to point IP
connectibility, so that the classes could have voice over speaker phones. And
once the 24 by 7 wireless and satellite connections were in place, there would
be no cost of ‘crossing over’ into the switched telephone networks in the
US or other places. Essentially ‘free’ voice calling. Once again they came
through. Cullen Jennings, an expert in VOIP voluntarily configured 4 7960
Ciscos for our precise needs, right down to the fixed IP addresses in Namche
they would use. I registered their SIP protocols through IPTEL.COM voip
registry in Germany – a free server service that stores the numbers for voip
phones, so other SIP phone users can reach them from anywhere on the net.
Finally Jim Forster of Cisco,
taken with the project, offered $2,000 personally so I could buy 2 late-model
Windows computers for the classroom in
Kathmandu and get them to Namche for the students to use once their classes
started. With two 70 pound bag weight limits for air travel from the US I
could not buy them, even more cheaply, in the US and carry them to Nepal with
all the other impedimenta I was going with.
After frequent email and phone
calls, one of my employees Marty Yazlowitz and I configured and pretested
every radio and network connections, as well as the Cisco phones ahead of time
in a short range shadow network. There would be little opportunity to start
wrestling with IP address, netmasks, and router configurations once I got to
Namche. No local help desks to call!
The Trek
I also
had to get my 75 year old body in physical shape for the trip. I would have to
fly into Lukla, Nepal from Kathmandu, a very small and scary airport with a 20
degree slope, on the side of a mountain part way up the Solo Khumbu. And then,
with hired Sherpa porters, trek for at least 2 days over 12 rough trail miles
and over 6 hanging bridges up 3,000 feet more to Namche Bazar at its 12,000
feet elevation. No FedEx delivery there!
But living at 6,500 feet in
Colorado, thus not too worried about altitude sickness at 12,000 feet, and
having long ago been a hard walking infantry officer, I faithfully climbed
daily up Barr Trail toward 14,000 foot Pikes Peak for six weeks until I was
fit to tackle the trail to Everest.
On 21 October, 2003 I flew
through Frankfurt, Germany, and Doah, Quatar to Kathmandu with my radios and
very advanced internet phones, plus my own laptop and a very small, Sony
camcorder with which I could take and send pictures from and even video clips.
On the trail I met interesting
people from all over the world. Most were taken aback at my age making this
trip. But to the Sherpas that is everyday life into their 80s. I met Apa
Sherpa who has summited Everest 13 times.
With Tsering who met me, and
Pavan Shakya, a lowland Nepalese employee of Worldlink who desperately wanted
to learn the secrets of wireless from me, we plugged away up the mountain and
got there with all the technical equipment intact.
Training and Installation
My
personal method of working with people who have never installed and operated
wireless devices – and very few have - is to teach them progressively what I
am doing, explaining everything theoretically, then doing the first
installation myself while explaining everything practical while they watch,
then make them install the second one themselves, me watching and critiquing,
then, expect them to do all the rest, answering questions by email, or Voice
over IP.
I went over what the
architecture would be, what is important in terms of Line of Sight, obstacles,
potential interference, possible need for relay radios, and how to do a rough
walk-around site survey before starting installing anything. I want them to
‘visualize’ in their minds eye what the invisible wireless network would
look like. So they would get used to looking for high points, clever
alternative locations, where radio relays might go, obstacles which will
attenuate the signals, and a rough idea of ‘distance’ the radios can
reach, with a variety of external omni and directional antennas. We started
that process the first, rainy day.
It was clear to me I had,
‘visualized’ Namche properly from maps and a National Geographic aerial
photo, and my years of experience to envision a practical layout of radios
within the village, with its steep terraced slopes.

David Huges on a Buddhist Stupa in Namche
installing a radio
But there remained one large
question mark. Where to put the key “Access Point” radio that would have
to ‘see’ the whole village from one point? And towards which all the
client and relay radios would have to point. Given the 28 degree ‘fan’
from the directional antenna built into the primary Access Point Radio, I was
afraid it would have to be placed across the valley on the steep side of
20,644 foot Kongde mountain, at least 2km away from the center of Namche to
‘see’ with enough strength the whole town. My calculations of radio power
and antenna gain told me that would be an acceptable distance for the RF
signal to go, with 15db of ‘margin’ as my benchmark for a reliable radio
link.
But looking directly across
the deep gorge at the mountain I saw it was almost a sheer cliff. Awesome in
fact. With small waterfalls trickling down its face. Tsering joked that when
the water froze crazy ice climbers, would take 4 days to climb to the top. I
asked him how long it would take he and other Sherpas to go around to the top,
and rappel down to anchor a radio, with batteries, and solar panels. He said
it would take at least 12 hours just to get there. Out of the question. Too
many variables. Very very risky.
Then, even though Tsering knew
little except what he had seen at Base Camp months before about Wi-Fi radios,
he said “Why don’t you put it on the Stupa at the bottom of the Tibetan
Marketplace?” I said “But that is a religious site, isn’t it? Won’t
your Sherpa neighbors be offended if this crazy American puts a modern radio
up on its traditional face?” His answer showed how little I still knew about
the Sherpa culture and the nature of their religious icons. “Its not really
a temple. No one would object. Besides, my family owns one of the four posts
that surround it.”
With nothing to lose, and no
better suggestion, we set about to mount the Access Point Radio up about 10
feet above the ground on one of the poles, and aimed it up to the center of
Namche. The 28-degree fan only subtended about half the town. I feared there
would be insufficient signal power to reach the lodges to each side, up to 90
degrees away from the direction the antenna was pointed. Also after Tsering
got a 220v power line run to the inner wall of the Stupa, we strung the power
line from the pole right over a bank of Prayer Wheels. That looked ugly to me.
But cheerfully Tsering said ‘Lets just hang prayer flags on the wire!” So
he did, and nobody was offended, at any of this. Just curious.
To my great relief, when I
dispatched three of the Cyber Café staff members to walk with laptop
computers and small USB powered Wi-Fi radios up and down each side of the
Namche bowl to each lodge to see what our coverage was, everything worked! The
‘side lobe’ power that radiates out at 90 degrees from the directional
antenna that concentrates the power forward was just strong enough to reach
inside the lodges to the side. They were only about 150 feet away, and that
helped.
Then quickly we were able to
put a Client, of Bridge, radio on the roof of the Cyber Café building, using
a wooden post (very few usable pipes were around, and even wood is scarce),
then drop the Power over Ethernet cable into the computer room, hook it to the
Router, which was connected to the IP Satellite dish, and we had end to end
wireless connectivity from any lodge to the Internet. Mission largely
accomplished! After that fundamental ‘backbone’ link was established the
tricky remaining task was to see how we could get the wireless signal, first
up onto the ‘shelf’ of land above Namche where the SPCC Park Headquarters
was, then to the Army barracks close by – both of whom wanted Internet
access, and then with a Relay radio from there to a Sherpa ‘middle’
school. We first were prepared to extend the link to the relatively well
endowed Hillary School which lay north 2km.
The headmaster had showed some
interest in being connected before I got there. But in the end Tsering could
not come to terms with the school. They wanted a totally free lunch. No way.
Tsering was in business. The satellite service costs him $1,200 a month. So
the eager Headmaster of the very poor Thame school, 4km away whose girls would
travel to Namche to do traditional dances to raise money for books, jumped in
to take advantage of the possibility of getting his kids educated from afar
– real distance learning, where the last link to the Internet would be
wireless. He would happily pay the $200 a month needed. This is where the
Cisco 7960 VOIP telephones configured for SIP connectivity, came in. SIP
protocol phones are IP point to IP point phones, which cost nothing above the
basic net connection to use. The basic idea was to extend the wireless link to
the school from Namche to Thame, requiring an intermediate relay radio, put a
small router in the school, there, connect up the two new Windows computers
Jim Forster had donated, and put in one of the Cisco IP Phones with speaker
there in the classroom. Then let all 9 Sherpa children and their Nepalese
teacher hear distant Mingma Sherpa who grew up in Namche, and now lives in
Pittsburgh, USA, use his own Cisco IP phone teach them, first, oral English,
and then written English via email, and later other subjects through the very
long computer connection.
The only hard part from this
point on would be how to set the class schedule, since Pittsburgh is 10 hours
different from Nepal – a 9 AM class in Thame would require that Mingma be up
at 11 PM in Pittsburgh to teach it!
Closure
With
the major wireless deployment tasks completed, and after my calculating what
it would take for a climber on the summit of Everest – which I could see
from Namche, to send and receive email wirelessly from a handheld PDA, or to
put a telescope cam at Namche to watch climbers the last few hundred meters
above the Hillary Step, I returned to the US. I remain in touch by email and
VOIP, including ‘Freeworldialup’ software phones which are the Next Big
thing. And let Tsering, Santosh, and the little band of Internet pioneers
carry on themselves.
|
|