Building shared wifi infrastructure from home wifi in Phnom Penh

[Rithy RAY]

Dear Group,

Anyone who interested to build this with us?

Regards/Rithy

[T Gillett]

Hi Rithy

Can you describe some details of  what you would like to set up?

Regards
Terry

--
You received this message because you are subscribed to the Google Groups "Village Telco Development Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to village-telco-...@googlegroups.com.
To post to this group, send email to village-...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/village-telco-dev/CALC5eokVJr2fgLZmWXpBuxhUVtM4Hu-BtQ0MYBVgdd9minknKg%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.

[Rithy RAY]

Thanks for your interest Terry.

There are some dots that we need to see:

-wifi are being installed at home by two main consumer grade ISPs

-those wifi router is not reliable and often create frustration when it hang

-several homes that got 300m or less distance from each other tend to use difference ISP but when their neighbor's ISP got down they cannot share the one next door that still working

-most wifi are idle at working hours and start using by owner when they are back from office

-ISP manage the contents so those user from other ISP cannot consume those content (movie and TV show and newspaper) and startup cannot get to these ISP easily to introduce their content

=> Why don't we solve all of these problems and charge USD 1 per home, suppose we got 10,000 home in Phnom Penh city of Cambodia? Imagine the amount of revenue we can generate from this and create jobs for people here?

Regards/Rithy

[T Gillett]

Hi Rithy

Providing Internet access to 10,000 homes is a large undertaking that would involve some serious commercial aspects, and is no doubt what your local ISP industry is trying to do.

It is unfortunate that the equipment that they provide is not proving to be reliable.
What type of distribution is being used eg ADSL, fibre, coax cable, 3G/4G cellular, or WISP?

Can other equipment be substituted?

It is true that many individual home ISP connections are under-utilised due to patterns of use, and it would make sense to be able to share a connection between several homes in many cases. The success of this depends largely on the type of usage by each home.

If all the homes sharing an Internet connection want to stream video at the same time, it's not going to work, simply because there won't be enough bandwidth or data capacity available.

But if the usage in each home is simple ad-hoc web browsing, email and the like, then it can work quite well.

You also have to be careful about the legal implications. There may be state regulations that prohibit using wifi between homes, and also the ISP agreement may prohibit such usage.

But assuming that there are no state legal issues, and you can find an ISP that will provide a good service at a reasonable price, then using wifi to connect small clusters of homes (~10) to a shared Internet service can work quite well.

This is a community based system where the group of homes can negotiate a sensible working and cost sharing arrangement, possibly assisted by an organisation providing installation and technical support.

A typical simple arrangement is to have a master wifi node at the point where the Internet connection is provisioned (the gateway), physically positioned so that it can be 'seen' clearly by nodes located in each home.

For networking, you can have a simple hub-and-spoke arrangement with wifi point to point connections from each home to the main node (a mini WISP), or you can use a wifi mesh to provide more complex networking eg with more than one gateway for capacity/redundancy.

In either case, it generally makes sense to have a node in each home, to which client devices connect, and the node acts as a router to connect to the wider network.

I think the keys to success of these schemes are to have a good understanding of the community needs, strong community groups, and an organisation to provide technical support. And on the technical side, the most important thing is to make it absolutely as simple as possible so that the system is robust and reliable.

Taking this approach allows you to start small, with minimal outlays, and to grow over time.
So you might end up with 1,000 simple, robust community systems each supporting 10 homes, rather than a monolithic system trying to support 10,000 homes with a huge administrative overhead.

Regards

Terry

To view this discussion on the web visit https://groups.google.com/d/msgid/village-telco-dev/54cf2f7e-08af-4a45-a518-624abab15218%40googlegroups.com.

[Rithy RAY]

Hi Terry,

Let's me try to address your comments:

Providing Internet access to 10,000 homes is a large undertaking that would involve some serious commercial aspects, and is no doubt what your local ISP industry is trying to do.
It is unfortunate that the equipment that they provide is not proving to be reliable.
What type of distribution is being used eg ADSL, fibre, coax cable, 3G/4G cellular, or WISP?

=> The two major ISP that aiming for home users are using fibre-to-home quality and very easy to get cut because it use utility poles.

Can other equipment be substituted?

=> I used to think about this too because home users tend to demand a good wifi such as eero.com but that cost USD 199 in the US (if we add cost to move that to Cambodia its will be very high since this vendor don't ship to Cambodia). Plus no one going to do electronic recycling for those old and not unstable wifi devices.
These two major ISPs use these wifi models:
http://opennet.com.kh/news/new-devices-on-april/
http://digi.com.kh/digihome/en

It is true that many individual home ISP connections are under-utilised due to patterns of use, and it would make sense to be able to share a connection between several homes in many cases. The success of this depends largely on the type of usage by each home.
If all the homes sharing an Internet connection want to stream video at the same time, it's not going to work, simply because there won't be enough bandwidth or data capacity available.
But if the usage in each home is simple ad-hoc web browsing, email and the like, then it can work quite well.
You also have to be careful about the legal implications. There may be state regulations that prohibit using wifi between homes, and also the ISP agreement may prohibit such usage.
But assuming that there are no state legal issues, and you can find an ISP that will provide a good service at a reasonable price, then using wifi to connect small clusters of homes (~10) to a shared Internet service can work quite well.

=> People will still use whatever ISP provided and comply with everything in their contract with ISPs but the idea is to create “FireChat” likes of these wifi. So when User A's wifi not stable or its WAN link is down, the user can select near by wifi of user B (provided both of them join our membership scheme).
Video contents is increasingly popular, the bandwidth is not an issue since each wifi device can be 300m or less from each other, if we can provide video on demand or other service on top of a “virtual network” created on top of all these home wifi.

This is a community based system where the group of homes can negotiate a sensible working and cost sharing arrangement, possibly assisted by an organisation providing installation and technical support.
A typical simple arrangement is to have a master wifi node at the point where the Internet connection is provisioned (the gateway), physically positioned so that it can be 'seen' clearly by nodes located in each home.
For networking, you can have a simple hub-and-spoke arrangement with wifi point to point connections from each home to the main node (a mini WISP), or you can use a wifi mesh to provide more complex networking eg with more than one gateway for capacity/redundancy.
In either case, it generally makes sense to have a node in each home, to which client devices connect, and the node acts as a router to connect to the wider network.

=> We might need to shape the model. The installation and support team can be hire hence allow us to create more jobs which is great.

I think the keys to success of these schemes are to have a good understanding of the community needs, strong community groups, and an organisation to provide technical support.

=> I can manage to make this part happen.

And on the technical side, the most important thing is to make it absolutely as simple as possible so that the system is robust and reliable.
Taking this approach allows you to start small, with minimal outlays, and to grow over time.
So you might end up with 1,000 simple, robust community systems each supporting 10 homes, rather than a monolithic system trying to support 10,000 homes with a huge administrative overhead.

=> That's why I come to this mailing list. Will you collaborate with us? Another potential is to share this network with university student and those who are in needs but don't have the budget to get their own home wifi. When we can make this work without legal and other barriers (when we do it right) the potential is huge. This is first city (Phnom Penh), there are few other major cities that can benefit from this if we can do this right. Probably we can expand to other country in ASEAN because the last time I checked there is no such implementation in Singapore or China. 

--
You received this message because you are subscribed to a topic in the Google Groups "Village Telco Development Community" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/village-telco-dev/gDMxNx1wyRQ/unsubscribe.
To unsubscribe from this group and all its topics, send an email to village-telco-...@googlegroups.com.

To post to this group, send email to village-...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/village-telco-dev/CAD58EkzaiNfNB9P80OrL%2BQMEpgVaptbSEGt0ghogiFq9b_cCRg%40mail.gmail.com.

[Vickram Crishna]

Elegantly put!

Thanks

Vickram

To unsubscribe from this group and stop receiving emails from it, send an email to village-telco-dev+unsub...@googlegroups.com.
To post to this group, send email to village-telco-dev@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/village-telco-dev/54cf2f7e-08af-4a45-a518-624abab15218%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to a topic in the Google Groups "Village Telco Development Community" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/village-telco-dev/gDMxNx1wyRQ/unsubscribe.

To unsubscribe from this group and all its topics, send an email to village-telco-dev+unsub...@googlegroups.com.

To post to this group, send email to village-...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/village-telco-dev/CAD58EkzaiNfNB9P80OrL%2BQMEpgVaptbSEGt0ghogiFq9b_cCRg%40mail.gmail.com.

For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "Village Telco Development Community" group.

To unsubscribe from this group and stop receiving emails from it, send an email to village-telco-dev+unsub...@googlegroups.com.
To post to this group, send email to village-telco-dev@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/village-telco-dev/CALC5eomAKKp-GRh%2BQxpFG9cbGERtBMDUxq4V1qgFOu7tRL4bQA%40mail.gmail.com.

[T Gillett]

Hi Rithy

Apologies, I still don't understand what you are proposing to set up.

A few thoughts on the points you raised:

1. Reliable ISP Service

The first issue seems to be that some ISP services are unreliable due to cables being run via utility poles. 

Obviously you can select the most reliable ISP, but you can also create redundancy by setting up a community network with multiple gateways using different ISP services so that failure of a single ISP service will not leave the network without Internet access.

2. Reliable equipment

The second issue seems to be around the equipment provided by the ISPs, and in particular the performance of the wifi routers.

The Eero routers that you mentioned are a high-end dual-band router that uses mesh networking to allow the use of multiple access points around a home to ensure good wifi coverage. It appears from their website that the system is based on OpenWrt software.

There is other equipment that is of good quality, economical and more readily available in your area that you can use for the wifi router function eg TP-Link and similar.

You can use this equipment either with the original firmware, or with OpenWrt based firmware if you want to do something not supported by the original firmware, such as mesh networking.

3. Bandwidth

Extensive use of video content will rapidly saturate a community network and upstream links. Typical ADSL connections of 5Mbps will only support one or two video streams from a video service such as Netflix. So if video is a fundamental requirement for each home, then you probably won't be able to do much sharing of ISP services. 

The speed of the local wifi network and the speed of the upstream ISP service are only part of the equation. What really matters is the data throughput that can be achieved from the local access point all the way to the video service provider. This is often a much lower speed than what is achievable in the local network and on the ISP link.

 4. VT SECN

Just for background, the VT SECN firmware is designed to provide simple small scale wifi mesh networks typically comprised of around 10 nodes with moderate data use requirements. 

The firmware is available for the MP02 device, as well as various TP-Link and Ubiquity devices.

The software is Open Source, based on OpenWrt and Batman-adv, and is community supported by volunteers. 

It is not a commercial product.

The system is not intended to scale out to hundreds of nodes with high volume data requirements. 

You can of course assemble larger networks comprised of small independent mesh cells, with each cell connecting to one or more upstream networks for Internet access. 

Data throughput in a mesh is dependent on many factors, but the realistic best case for simple single-radio 802.11g nodes is 10 - 20 Mbps, so when you share that around 10 nodes you end up with 1 - 2 Mbps on average per node. That won't support a lot of video users connected to each node.

Regards

Terry

To view this discussion on the web visit https://groups.google.com/d/msgid/village-telco-dev/CALC5eomAKKp-GRh%2BQxpFG9cbGERtBMDUxq4V1qgFOu7tRL4bQA%40mail.gmail.com.