Wednesday, July 31, 2013

ISM 915MHz band availability in different countries

I found this handy table today that lists ISM band availability around 915MHz in a variety of different countries:

While closer scrutiny is required, it is the best compilation of this data I have found yet, especially for countries outside of ITU Region 2.

Sunday, July 14, 2013

View of Mesh Extender prototype with 50Wh external battery

Just a couple of quick snaps from in the lab today showing the Mesh Extender prototype with an external battery that should be sufficient to run it for 24 hours.

A lot of the battery case is empty air, so it would be possible to make something even more compact.

We might also look at making the shell of the mesh extender so that it can physically attach to a battery pack for simple one-piece operation when required, but still allow replacement with an external battery pack.

The battery pack is about the size of a Galaxy S2 phone, but 20mm thick.  As mentioned, this could be shrunk with a better battery pack than this cheap one we bought off deal extreme.

We have since refined the prototype further, and found a smaller battery.

I really could have used some mesh communications gear last weekend

(Please also take a look at our crowd-funding campaign at

Last weekend my family and I stayed down at Second Valley, an hour and a half drive from home in Adelaide.  It is a popular beach holiday spot, with a variety of beach shacks, permanent homes and farm land.

We arrived Friday night in the rain, only to discover that the power was out.  There are a lot of trees around the area, and with the woolly weather it was quite possible that a one had fallen across the power lines somewhere, and so given the late hour we figured we would just get ourselves and the kids into bed in the dark, and worry about it in the morning.

What we didn't know until after, was that the problem was the power junction on the property filling up with rain water, and so no crew was going to be called out automatically:

This box should have been full of electricity, not water.

We soon discovered that our assumption that the local infrastructure would be available had caused us to be quite poorly prepared.  We only had our mobile phones and a laptop computer for light sources, and hadn't bothered to fully charge any of them.

This was a bit of a problem as our little people still needed story time, and some light in the room to get to sleep.

But electricity wasn't all we were without, because water makes its way from this tank:

The rain tank

 Through this electric pump:
The water pump

And into this rather vital household appliance:

The toilet
Fortunately we could use a bucket to fill the cistern from the water tank. Oh yes, and the septic system is electric as well, so even using buckets to operate the cistern we were only buying time before the manure hit the, well, lawn.

The gas stove also had an electric starter, but fortunately we had a gas lighter, and could at least boil water.

To add to the challenge, there is no land-line phone in the house, and the mobile phone coverage is of the kind that requires you to stand on the toilet seat and make a strange face, or stand outside facing 17 degrees NNW and hold the phone exactly three Royal Cubits above ground level to get signal.

This made it a difficult proposition to call the electricity network to find out when the power would come back.  So we weathered the night, but in the morning there was still no power.  But there was still rain.

After eventually making the call standing in the rain, we were informed that a work crew had been dispatched and would come as soon as they were able.

They did eventually turn up, but only after they had tried to call us, and as far as we can work out, didn't come immediately because they couldn't get in touch with us to make sure we would be home.

All in all, the lack of effective communications meant that we were without power for probably 12 hours longer than necessary.

So here I am running a crowd-funding campaign for mobile mesh telephony, and so I should be amongst the most prepared in the event of things going wrong communications wise.

But of course I hadn't bought any gear along with me, not a mesh extender, or even a mesh-cellular gateway made of duct-tape, so that I could at least have waited inside for the call back from the power company.

Maybe next time I go away, I will actually carry the tools I make.  

This is also why it is important that we make the Serval Mesh useful, even when things aren't going wrong, so that people are carrying it and know how to use it when the need arises.

Friday, July 12, 2013

Smaller, more elegant Mesh Extender prototype

(Please check out our crowd-funding campaign at to help us turn the Mesh Extender into a product that can be used to help people communicate in difficult circumstances.)

There is nothing like being busy to increase the incentive to just get something done in spite of the things that make it impossible.

So today I decided that I would actually get around to making a Mesh Extender prototype out of a TP-Link MR3020 wireless router.

These have an Atheros 9k chipset, so they have the good Wi-Fi chip that we need, which is usually the tricky bit. They have a USB port for the mass storage that Rhizome needs to run.  So far so good.

The previous TP-Link (WR703N) routers we used didn't have an easily accessible serial port -- they just had a couple of tiny points on the PCB you could solder to.

In contrast, the MR3020 has real hole-through pins that can take a standard 0.1" single-inline header, and some of the earlier builds actually have the connector installed.

With the help of a work-experience student we put OpenWRT on several of the MR3020s and got them about half-way to working as Mesh Extenders, able to mesh over Wi-Fi.

That just left the serial connection to the UHF packet radio, which is what I attacked today.

The serial port normally has a login and is used to log kernel messages.  After figuring out how to turn off as much of that as possible short of recompiling the kernel, I then set about connecting up one of the packet radio modules and getting some holes drilled so that we could put the RFD900 packet radio in the little case.  Well, mostly in the case, because the heat-sink sticks out a bit.

The following pictures show how this makes a neat little unit:

We have some Verbatim nano USB memory sticks on order so that we can replace the big USB drive on the side with a little tiny bump, which will further stream line the shape.

That leaves only the battery to consider, and we have a few options for those.

Importantly, we have shed the USB hub and USB to serial adapter, which between them were consuming perhaps 20% of the total power budget, and made the resulting prototypes frustratingly large.

Indeed, compare the images below of the older prototypes to the images above, where the only missing component is a battery connected to the USB port of the MR3020.

Once we cut a hole for the heat-sink and get the lid fitting back on, and install the nano USB sticks, we should have a nice pocket-sized prototyping platform from which to refine.

Thursday, July 11, 2013

Android phone prices hit new low

The march of technology keeps driving the price of low-end Android phones ever lower.

Economics is a strange beast at the best of times, and all the more so when it comes to the price of mobile phone handsets as various subsidies can reduce the price substantially.

The result is that the cheapest countries to buy mobile phones are those where the carriers are willing to offer the deepest subsidies, and it isn't always obvious which countries those are.

My anecdotal experience suggests that the two cheapest countries, with prices generally similar, are Kenya, which makes sense since GDP-per-capita is relatively low, and Australia, where, well, to be honest I don't really know.

We have previously bought Android phones like the Huawei IDEOS U8180 for AUD$50 - AUD$70.

Then the other day I noticed that the local supermarket is selling an Android phone for AUD$40, and I thought that was cheap:

But then the local Post Office catalogue arrived offering the Huawei Y100 (U8185) for AUD$29 including a 2GB microSD card and $2 starter SIM card with Vodafone:

This $29 wonder includes GPS, Bluetooth, an 800MHz processor and a screen so small it almost hurts. Basically they are a 50% faster version of the U8180's we are so familiar with.  The extra speed makes a noticeable difference.

We had the latest version of the Serval Mesh running on it within a few minutes:

Sure, both of these phones are also benefit from discounting through the retail chain, in much the same way that all the Android phones around the $50 mark were a year ago.  History suggests that Android phones at this kind of price-point will become regularly available over the coming year.

Tuesday, July 9, 2013

Interfacing between the Serval Mesh and Cellular Networks (Part 2)

In a previous post we described our initial work towards creating a two-way mesh to cellular gateway using mobile phones as the only major component.

Thanks to Michael's great work, we have this now working.

Calling out is automated with the Serval Mesh software announcing itself as a GSM gateway when you attempt to call an off-mesh number, as you can see in the screen shot below.

He also managed to get incoming calls working.

He even managed to get DTMF parsing working, so that when you call into the mesh you get prompted for the number on the mesh you wish to be connected to, and you can then just dial the number on your phone, and get put through.

This last feat is quite amazing given that the DTMF codes have to survive being played through tiny ear buds, re-recorded through a tiny microphone before actually being decoded.

We know that the frequency response of the audio path is very non-linear, but Michael managed to come up with a commendably simple signal processing solution that could work in spite of this.

You can see how easy it is to setup in the following video:

On the same day we filmed that video, I also called one of our friends in New Zealand Red Cross using this arrangement.

That is, from our lab in Australia, I dialed his normal New Zealand number from on the mesh, selected "GSM Gateway" when offered, and then waited to be connected.

There was a lot of delay early in the call, bit it then caught up to within a couple of seconds for the rest of the call.

That delay was longer than normal for a mesh call, possibly because one of the mesh phones was acting as the access point for the phone I was using, and there may have been some CPU contention (we were using 500MHz IDEOS U8180s that are very slow), combined with Wi-Fi packet loss in the University building here due to the presence of lots of other Wi-Fi devices.

Anyway, Matthew and I had a usable and effective conversation for about five minutes discussing recent news.

Matthew's assessment of the audio quality was "not great, but definitely usable, and certainly much better than nothing."

We agree, and really, this is the point: there are 100 better ways to connect a mesh to the outside world, but if all you have on hand during a disaster is a couple of phones and some duct-tape, there was previously no solution -- but now there is.

We plan to integrate this capability into the mainline Serval Mesh app in the future.

To help us do this, and generally advance the Serval Mesh, please consider supporting our crowd-funding campaign at, and spreading the news about that campaign far and wide.

Monday, July 8, 2013

Serval in the MIT Technology Review

See the current issue of the MIT Technology Review for a write up about Serval and Commotion and mesh telephony in general:

They provide a nice summary of our long running problems with Android failing to support ad-hoc Wi-Fi, and that this prevents various humanitarian uses of Android phones.  We encourage Google to review their policy.

Shameless plug: The crowd-funding campaign they mention as launching soon is

Sunday, July 7, 2013

Crowd-funding the Serval Mesh Extender Announcement & Video

Those who have been following the blog here will have heard about our prototype Mesh Extenders that we have been developing.

In short, the Mesh Extender lets us break the "Wi-Fi barrier" of Mesh Telephony, typically extending range 10x to 100x depending on the environment.  This means that a single mesh hop can span a block or two in typical suburban or urban areas, and into the kilometres in open terrain.

This has tremendous potential to increase the utility of mesh telephony.

Also, because the Mesh Extender acts as a Wi-Fi access point for phones, it completely removes the need for "rooting" or "jail breaking" a phone for it to participate on the mesh.

This also means that lots of nearby phones can use a single Mesh Extender.

In other words, it makes the mesh better and easier to join.

We now want to refine these prototypes into something that can end up being a saleable product so that we can get to the point where the general public can buy them and use them.

We thought about venture capital and other routes, but we don't want to compromise the openness of the technology we are creating.

This means that private and public philanthropic funding have been the mainstay until now, and has been vital in getting the Serval Mesh and Mesh Extenders to their current stage where we can make secure mesh phone calls, send secure mesh text messages and share files over the mesh.

While this sort of funding has been great, it is not really geared to moving to the product stage.

This is why we decided to launch a crowd-funding campaign so that the people who want the technology we are creating, and will ultimately benefit, can participate in funding its development, and help make sure that the Serval Mesh is available as soon as possible, and remains completely free and open software and hardware.

Our target is US$300,000 which will allow us to design a product based around the prototype Mesh Extenders, and attack the outstanding software work required to make the Mesh Extender an easy-to-use and automatically updating appliance.

I invite you to take a look at our campaign at, and encourage you to spread the word far and wide, and if you would like, to contribute.