This is an old revision of the document!
“Mobile telephony for those in need.”
The Serval Project is a small team of academics, contracted engineers and students in the Resilient Networks Lab of Flinders University in South Australia, developing revolutionary, free, open-source software for mobile telephones. It was founded by Dr Paul Gardner-Stephen and Romana Challans in 2010 in response to the Haiti Earthquake, and has received funding and equipment from several supporters.
The Serval Project aims to bring infrastructure-free telephony to people in need, such as during crisis and disaster situations when vulnerable infrastructure like phone cell towers and mains electricity are cut off. In practice, this means finding ways to make telephone handsets communicate without mobile cell towers, repeaters, WiFi hot spots, or cables.
The main focus to date has been on developing the free Serval Mesh app for Android to provide voice calls, text messaging and file sharing directly over Ad Hoc WiFi links between phones. The app communicates with distant phones via intermediate phones using mesh routing techniques, and uses strong, elliptic curve encryption to guarantee privacy and identity even though some phones in the mesh network may not be trusted.
The Serval Project also develops the Serval Maps app for Android which uses Serval Mesh file sharing to provide decentralised mapping. This is a useful situational awareness tool for emergency response teams.
The Serval Project is developing the Serval Mesh Helper device to overcome the range limitations of WiFi on smartphones and to extend Serval Mesh services to handsets other than just Android devices which support Ad Hoc WiFi mode. This will allow more kinds of smartphones to participate in the Serval Mesh.
The Serval Project is cooperating with the Commotion Wireless project to integrate the Serval Mesh into the OpenBTS GSM base station, which will bring conventional GSM “feature” handsets into the mesh and will extend the reach and security of Commotion coverage by meshing its OpenBTS base stations together.
WiFi was the obvious first choice for wireless communication in disaster situations because of its widespread regulatory availability – most countries have allocated the WiFi frequency bands for short-range domestic use – and the maturity and broad adoption of the WiFi 802.11 standards.
The “smart phone” was chosen as the initial device because all smart phones support WiFi, whereas most “feature phones” do not. Also, the operating systems of the most popular makes of “feature phones” are all proprietary, which presents financial and legal hurdles.
The Android platform was chosen for initial software development over other smart phone platforms, because only Android permits the Ad Hoc WiFi mode to be enabled (although most commercial handsets must be “rooted” in order to do so). Android's terms and conditions on developers also place fewer obstacles to the kind of freedom-oriented development that the Serval Project is pursuing.
The Serval Project hopes to eventually reprogram the GSM transponders on some makes of smart phones, so they can intercommunicate using bands other than WiFi, eg the 900 MHz GSM band. That frequency band offers a greater range but lower data rate than WiFi, and thus will produce a much larger scale, lower speed mesh. This is a very ambitious goal, and will require collaboration with industry partners.
The Serval Project aims to eventually use Bluetooth for short range meshing, so that the larger mesh can be spared high-volume data transfers between nearby neighbours.