It is an unofficial feature in the 0.07 release that you can use WebSMS and SMSDroid to send and receive mesh-based SMS (MeshMS). Recent versions of those applications will automatically detect Serval MeshMS if it is installed.

Version 0.08 (due in February 2012) will also allow mesh file transfer via Serval Rhizome.
Version 0.08 will also let you send and receive MeshMS even when there is not a connection possible via the mesh between the endpoints.  This works using a store-and-forward mechanism.

See the following posts for some information:

Intro to Serval Rhizome and demo video

A more advanced demo video

The Serval tech roadmap discussion

This version offers compatibility with many more models of Android handsets, fixes to a large number of bugs, and some general usability improvements.  Full functionality still requires root permissions on your phone, but it does its’ best to work on phones without root access.

Personally, what I am most happy about is that we have substantially improved our software development and release process, which has already resulted in a much better quality in the current release, and I am confident will result in increasing quality in future releases.

In the longer-term this is the first of several planned releases that will hopefully bring us to a 1.00 public release around September 2012.  The planned intervening releases and their headline improvements, subject to resourcing, slippages and all the other usual hazards of software development are:

0.08 – Serval Rhizome/MeshMS store-and-forward file distribution core technology.  Automatic update of Serval software over the mesh.

0.09 -  Revamped and consolidated MeshMS API and functionality.  Integrated UI for sending and receiving MeshMS messages, including using the Rhizome store-and-forward facility.The underlying technology has already been demonstrated by sending an SMS between Africa and Australia without reliance on infrastructure.

0.10 – Revamped UI for entire Serval Mesh application.  Voicemail and Push-to-talk group communications.

0.11 – Serval Mapping Integrated (using Rhizome to get map tiles without internet access).  Voice calls moving from SIP+RTP via SipDroid and Asterisk to Serval MSIP and MDP, which will improve performance on lossy wireless networks, allowing clear calls over multi-hop links each with 50% packet loss, and greatly reducing the APK size by removing the full Asterisk deployment it currently contains.

0.12 – Consolidate existing work, and address outstanding issues for first public release.  0.12 will become 1.00RC1 when accepted.

Together, these releases will culminate in a software package that will let you do most mobile-phone type things (voice calls, SMS/MMS, voice mail, push-to-talk) when there is no infrastructure, or for free when infrastructure is available, such as transferring and sharing data among friends and community without having to pay up to $2/MB for SMS traffic or $2/MB for cellular data services.

This is good news for people with very low purchasing power, such as large portions of the population in developing countries as well as for those living in more developed countries such as Australia, where infrastructure is not universally present, especially for those living in rural and remote areas.

But for now, it is back to working on Serval Mesh release 0.08 …

On the hardware side, we have ordered 50 Huawei IDEOS X1 phones for AUD$71.10 each — pretty cheap for a fully featured Android smart phone, complete with digital compass, GPS, accelerometers, a capacitive screen and enough CPU and RAM to meet our needs. These phones are network locked, but that doesn’t worry us as we won’t ever be putting a SIM card in them, because they will be used solely as mesh communications devices.

It turns out that buying 50 phones at a time is an interesting exercise.  People want to know that you are not planning to resell them or export them.  Some shops will even try to limit you to buying five or less at a time.  Those tactics are legal in some places, but not here in sunny South Australia, where we have an excellent Fair Trading Act that correctly classifies such terms of trade as discriminatory, or in some cases as tantamount to bait-and-switch.

In fact, until quite recently the South Australian consumer protection and gambling laws were so strong that cereal-packet competitions could not require a South Australian resident to purchase a product to participate in a trade promotion lottery, because it would be considered gambling, i.e., pay for chance to win.  So when I was growing up it was quite common to see terms and conditions looking quite normal, and then have sections specially for South Australians, that would often say something along the lines of:

“10. To enter, cut the bar code from a packet of XYZ and include it in your entry as proof-of-purchase.  Residents of South Australia are permitted to submit a hand drawn facsimile of the proof-of-purchase.

…

17. Only one entry per household per day.  Residents of South Australia may enter more than once per day.”

There were certainly occasions in my childhood when we would go to the effort of getting the coloured pencils out to draw a barcode by hand to get the chance to win a prize.  I think we sometimes did it just because we could.  Sadly, with the introduction of poker machines (Australia has fully 21% of the worlds poker or slot machines — a statistic that is disturbing given that Australia has only about 1% of the developed-world population) and the general liberalisation of gambling in South Australia the situation has changed, and I believe we have lost our little parochialism.

Anyway, I digress.

After the difficulties I encountered initially in sourcing 50 mobile phones, I eventually went to our local post office to see if they would order me 50, since they had the model I wanted in their catalog.  Not only were they willing to order them (I promised to pay up-front, since it was a large purchase), they also kindly sliced 10% off the price.  So I now have a box of 44 phones in my lounge room, with the remaining 6 on back order:

At this price, the phones don’t come with SD cards, which we do need for the Serval Rhizome software to do anything useful, as store-and-forward requires somewhere to store!  So I hunted around, and after some luke-warm experiences with suppliers either being unwilling to give me a decent price for what is clearly a wholesale quantity, or failing to stock enough, I tried OfficeWorks.  They had enough in stock, but were charging $16 for 4GB cards, which I had quotes from other local stores for $7.  So I loaded my basket with 50 memory cards, and proceeded to try their price-matching policy:

However, the large price difference and quantity did trigger concern from their store manager.  He was very friendly and amiable, but did point out that to get the price matched, I had to find a local supplier with the quantity I wanted actually in stock.  But that wasn’t easy to arrange, as most of the cheaper suppliers only had a few on hand.  There is of course a silliness in their rule, as I could have bought them in small batches such that each batch was small enough that the cheaper shop held sufficient stock.  But in the end I did manage to find a cheap supplier on eBay who had plenty in stock at $5.80 a piece for SANDisk 4GB class 4 microSD memory cards.  Now we just have to see if they will arrive.

So that covered the phones and memory cards.  The next challenge was to come up with a way to easily charge 50 phones.  Not only did it need to work in my home, but it also had to have the potential to work in a real disaster zone, and one which might not be in Australia.  Having enough power boards to plug 50 little USB charger bricks was not an attractive option, especially since I need to fit all of this into my 23kg luggage limit when I fly, and 50 phones alone are not that light.

What I have decided to do is to chop up a pile of USB leads and wire them up to some automotive 5v accessory power supplies that are commonly used to power portable DVD players and the like.  Apart from being the cheapest 5v high-current power supplies I could find (AUD$40 for 5A), they have the benefit of running on 12v, or in the case of some models, on 12v-24v, so that they can even work in a truck.  This means that in-vehicle mass-charging is possible when the power grid is off.  It also means that they can all be charged using a single cheap 13.8v power supply, which are also quite cheap; you could even use an old computer power-supply.  You can see my prototype here:

So now I just need to get back to writing the software that needs to go on these phones …

• Automatic Wi-Fi chipset detection process that tries to guess how to put you phone into
• adhoc mode. This should now work with most rooted handsets.
• New screen for displaying the progress of tasks we have to do on install.

Bugs Fixed :
The following bugs have been fixed since Release 0.06-

• 0000002: Handset auto-detection needs fixing
• 0000029: Wifi clients may drop broadcast packets
• 0000098: Preparing information sent back to project for display
• 0000127: java.lang.NullPointerException : SetupActivity.setAvailableWifiModes()
• 0000131: When operating in AP-Client modes, AP cannot be called
• 0000141: Experimental scripts may be run and crash the software (or phone) on install
• 0000145: Call to node A fails second time after call to node B on three node network
• 0000147: Adhoc edify interpreter needs rebuilding so that we can use “generic” wifi
• driver loading
• Stopped DNA from corrupting its data file and crashing which would leave the phone
• unreachable (General root cause of several other bugs).

Bugs can be viewed in more detail via our bug tracker – registration is required to access.

On a mesh, it is possible for us to deliver messages to other phones that are reachable on the mesh at the same time.  This works great, and we have had this capability in the Serval BatPhone software for a while now.

However, it does have some limitations, in particular, if there is no link all the way from the sender to the receiver on the mesh at the instant the message is sent.  This is rather unfortunate, as we usually think of SMS as being the most reliable fall-back on a mobile communications network.

So we set about replicating this resilience by creating a store-and-forward SMS-like service on top of the Serval Rhizome mesh file distribution framework.  We call this MeshMS, for Mesh Message Service.

This service uses direct mesh links if they are available by making use of the existing SMS capability in the Serval BatPhone software.  However, if there is no direct link, then it uses a store-and-forward scheme, that asks any passing phones to copy the message and distribute it to other phones on the mesh, until it (hopefully) eventually reaches its intended destination.  The cartoon below shows how this works.

The man wishes to send the message “Prepare the Zeppelin” to his hench-person, but there is no direct mesh link at that time due to the challenging topology.  The message gets picked up (stored) by the compatible phone in the pram, without any action on the part of the woman.  The woman then keeps walking along, and eventually the phone in the pram is able to automatically deliver (forward) the message to the ultimate destination, whereupon the minion knows to ready his master’s zeppelin, to do the weekly grocery shopping, one trusts. Of course, there could be many intermediate steps, instead of just the one shown here.

The great thing about this approach is that it doesn’t require a complete path to the destination at the time of sending, but can propagate progressively across the mesh as it is able, and can make use of nodes that move between otherwise isolated mesh networks, creating an asynchronous link between them where it would not otherwise be possible.

While the delay in such a service is huge, the bandwidth is also great, as potentially gigabytes of data can be transferred between nodes.  One side use of this protocol that we intend to exploit is to provide an efficient means of distributing updates to the Serval software suite, so that field updates can occur without dependence on infrastructure, and with much greater aggregate bandwidth than any single-cast cellular or wireless approach.  In fact, this will even allow the update of software, maps and other resources during a disaster.

We have even demonstrated it to deliver an SMS message between South Africa and Australia, using nothing but compatible phones to carry the message more than 10,000km:

This feature has profound utility in allowing the exchange of messages, files, and all manner of data in what are otherwise very difficult settings.

It also provides the capabilities required to enable disadvantaged communities, perhaps in remote locations, war zones or informal settlements, to create their own infrastructure, carrier and cost-free SMS networks using compatible handsets, so that they can enjoy the benefits of digital communications that many of us take for granted.

The procedure for Mac/Linux has now been reduced to (assuming you want the wifi-autodetect branch, which for now you probably do):

1. git clone git://github.com/servalproject/batphone.git -b wifi-autodetect
2. cd batphone
3. ./BUILD.txt

After that, you should be able to compile the software from in Eclipse without difficulty, as all the messy stuff has been taken care of (git sub-modules, JNI/NDK compilation, NaCl preparation, among others).

This easy build process will be pushed up to the master branch fairly soon, but for now, it is only available with the wifi-autodetect branch.

Note that building on Windows remains unsupported, as end of line markers in files get messed up if great care is not taken.  If you want to build on Windows, we recommend you install a Linux virtual machine, e.g., using the free VirtualBox software, and run the build process from in there.

Six months ago we did not have a developer software offering, or even a single Android application that contained all the components for Serval. Today we do.

Six months ago we did not have MeshMS (Mesh SMS service), interactive mapping,  Rhizome file distribution or store-and-forward SMS working.  Today, we do, and have even sent photos and SMS messages more then 10,000km between Australia and South Africa using mobile phones as the only infrastructure (more on this in a blog post and video in the next few days).

On-site in South Africa after testing the Serval MeshMS store-and-forward SMS Service, delivering files and SMS messages up to 11,000km without infrastructure, or a continuously connected mesh.

Six months ago we had a number of legal issues to work through with regard to giving Flinders University an appropriate holding in the Serval Project’s commercial arm, and the Shuttleworth Foundation had not yet come on board.  Today, we have just about finalised the details for these arrangements, with the final execution to occur in the coming weeks.  It will be fantastic to get this settled.

Six months ago we were not greatly recognised internationally for our work on mesh communications.  Today, we are increasingly recognised as the leaders internationally in this space, which is a tremendous honour and also acts to spur us along to deliver on all the potential that mesh networking offers.  Recognition has included support by the Shuttleworth Foundation (South Africa), reaching the finals of the World Embedded Software Competition for University students (South Korea), strong engagement by and with the IEEE 802.11 standards process (USA and international), presenting at the Adelaide Festival of Ideas and TEDx Adelaide (Australia), and also reaching the finals of the Ashoka Foundation World Changer’s Citizen Media competition (international) from a field of more than 400 entrants.  We are also seeing university and other partnership opportunities in Australia, Europe, the Middle East, Africa, the Americas, and Asia begin to take shape.  Now all we need is to be able to clone ourselves a few times over so that we can engage fully with the various opportunities!

Our plans for the next six months are to fill out the feature set that we have generated (voice, SMS, MMS, interactive mapping, file and software distribution), and add in the missing pieces, primarily the security and authenticity components that we have planned from the outset.  This will take us to mid-2012 when we hope to focus on maturing all of this into a general public release, as compared to the early-access developer software we have released to date.

We are also looking seriously about getting advanced mesh support built into one or more models of mobile phone to offer energy efficiency and/or range improvements over the basic service.

Simultaneously, we are beginning to explore appropriate trials of the technology that if they go ahead will gain us valuable feedback, and refine the software to maximise it’s utility for the general community, including in South Africa (assisting in the distribution of educational material without cost or dependence on cellular or school IT infrastructure), North America and Australia.

So all in all, the next six months should keep us fairly busy.

We also welcome Lyn Stephens, who was helping us out part time, but will be working with us full time on the Senior Development team as a student benefiting from support from the( ever appreciated and generous) Shuttleworth Foundation.

Updated profiles are available on our Team Page.