Virtual Mobile Ad-hoc Networks Jonathan Lahav, Ariel Stulman, Member, IEEE Department of Computer Science Jerusalem College of Technology Jerusalem, Israel e-mail: j.lahav@gmail.com, stulman@jct.ac.il. Abstract—Ubiquitous use of MANETs (Mobile Ad-hoc Net- works) is greatly hindered by the technological cap on broadcast distances. In this work we propose a method of extending the cover area of a MANETs by taking advantage of pre-existing mesh networks among routers. This is done in a transparent manner, so participating ad-hoc nodes are oblivious to this mechanism. They see both ad-hoc networks as one, effectively creating a large virtual network. We implemented a proof-of- concept that uses a MANET client application over the Android operating system for providing voice calls between smartphones. Index Terms—MANET, mobile networking I. I NTRODUCTION With the proliferation of mobile platforms (i.e., smart- phones, tablets, etc.), it has become feasibly possible to create mobile ad-hoc networks between such machines. Like in many other networks, it is technically possible to stream through these networks many types of data (voice, video, and the like). These networks provide users with great advantages including, but not limited to, free calls - as there is no need for a carrier, security - since the big brother controlling central nodes isn’t present, and robustness - as routes are found on-the-fly given whatever network nodes exist. These networks, however, are not widely adopted for daily use. This can be attributed to two basic problems: 1) The range of WIFI which is limited to a few dozens of meters, requiring adjacent nodes to be within those parameters. And, 2) Limited battery life and memory capabilities of mobile devices today induce users to turn off this functionality. Allowing data flow through one’s phone is a form of al- truism that rapidly diminishes due to these technological caps. There can be numerous solutions to these problems, including an increase of WIFI range, stronger batteries and larger memory capabilities. These advances are only a matter of time, but for now they are quite crippling. In this research, we focused on using existing infrastructures to bridge the gap between ad-hoc nodes. We route MANET data, both data packets and node discovery, through home routers, effectively creating a virtual ad-hoc network (see Figure 1). People are much more inclined to allow data to flow through their routers, since they are not constrained by power, memory or bandwidth (see, for example, city-wide mesh networks like Freifunk [1], and P2P networks like Tor [2]). In a similar manner, we use routers to route ad-hoc traffic on our behalf. This paper is a proof-of-concept to this method. Figure 1. Test architecture II. PROOF OF CONCEPT In order to create the proof of concept, we first need a MANET (Mobile Ad-hoc Network) client. This is an ap- plication that provides us with the two essential services required to setup such a network; packet routing and node discovery. For network routing, mesh networks usually use a variation of OLSR [3] or B.A.T.M.A.N. [4], so using a MANET application that utilizes these protocols will require least adaptation on routers. For this reason we chose Serval [5], an Android [6] based MANET application, as our basis (see Figure 2), and routers using OpenWrt [7] as an operating system and OLSR as the routing protocol. 2013 IEEE 10th International Conference on Mobile Ad-Hoc and Sensor Systems 978-0-7695-5104-3/13 $26.00 © 2013 IEEE DOI 10.1109/MASS.2013.95 433