DIY MOBILE AD-HOC WIRELESS MESH NETWORKS (MANETs)
Rutgers team proposes Net alternative
by Rick Merritt / 4/28/2011
San Jose, Calif. – A team of researchers at Rutgers University have launched the latest of a group of wireless network initiatives aiming to create a more open alternative to the Internet. MondoNet aims to enable a mesh network that lets a hybrid collection of new and existing Wi-Fi, WiMax and other wireless devices connect to each other without going through a central carrier. A draft proposal for MondoNet describes its premise as well as how it will gather the best of existing technologies for mobile ad-hoc wireless mesh networks (MANETs). The project’s goal to create a system that provides both greater freedom and privacy for individual users than today’s Web. Aram Sinnreich, organizer of MondoNet and an associate professor at Rutgers, outlined the proposal in a recent video. Today’s Web is subject to censorship and manipulation due to close links between a handful of carriers and their governments, he said, citing examples in China, Egypt and the U.S. “All the information [on the Internet] has to go through the eye of the needle of a few companies beholden to their governments,” Sinnreich said. “What we need is a new network,” he said. MondoNet aims to be as fast and feature-rich as the Net while being more immune to censorship and spying. Researchers hope to get funding to create a prototype of their concept in the Rutgers area. Many legal and technical challenges remain open, the researchers said. For example, they propose use of tcpcrypt for security, although they admit it is not immune to malicious attacks. It’s also not clear how MANETs will get permission to let users act as broadcasters or what form of licensing MondoNet will use for its software.
The effort aims to adopt techniques from a number of other pioneering efforts in MANETs including:
- BATMAN: A Better Approach to Mobile ad-hoc Networks launched earlier this year.
- Babel: A distance vector routing protocol
- Daihinia: A tool to turn Wi-Fi devices into a mesh network.
- Freedom Box: A simplified Linux server for distributed networks
- GNUnet: A software framework for secure peer-to-peer networking
email :sinn [at] rutgers [dot] edu
WEAVING a NEW NET: a MESH-BASED SOLUTION
Weaving a New ‘Net: A Mesh-Based Solution
for Democratizing Networked Communications
by Aram Sinnreich, Nathan Graham, & Aaron Trammell / Rutgers University
Recent developments, from the mass release of sensitive diplomatic cables by Wikileaks to the social media–fueled revolutions and protests currently gripping the Middle East and North Africa, have underscored the increasingly vital role of information and communication technologies (ICTs) in geopolitical affairs. Further, a wealth of recent research demonstrates the growing importance of digital networks in fostering cultural innovation and a vibrant public sphere, and the increasing centrality of these technologies to the daily lives of billions of individuals across the globe. Given the centrality of ICTs to these emerging changes in our social, cultural, and political landscapes, and the oft-invoked observation that “code is law,” it is essential that we develop and maintain a communications infrastructure that will enable individuals and communities (especially those in danger of political repression) to participate and contribute fully and actively to the public sphere, and to communicate confidently in private. Unfortunately, today’s infrastructure is not fully adequate to achieve this end. As U.S. Secretary of State Hillary Rodham Clinton recently observed, “the internet continues to be constricted in myriad ways worldwide.” While this is certainly the case in repressive political regimes from China to Iran, we face significant obstacles to “internet freedom” in America, as well. Although the internet is highly decentralized in its communication and social patterns, its technical and regulatory foundations are extremely hierarchical, due to centralized control by organizations like ICANN and oligopolistic ownership of network access. As a result of this centralization, digital communications are vulnerable to a degree of surveillance and censorship that would be unthinkable in traditional social arenas, threatening free speech and cyberliberties. Many laws and regulations exploit, rather than ameliorate this threat. Seemingly disparate factors like tiered access, intellectual property laws and national security measures, taken in combination, threaten to produce a communications environment in which cultural innovation is stifled, normative behaviors are criminalized, and political dissidence is dangerous or impossible. We believe that a new architecture is required in order to protect the continuance of civil liberties in networked society. In this article, we propose 10 “social specifications” describing the requirements of such an architecture, and outline a project called MondoNet designed to meet these specifications using ad hoc, wireless mesh networking technologies. We also address the legal and technical challenges facing the MondoNet project, and anticipate future developments in this field.
Weaving a New ‘Net: A Mesh-Based Solution
for Democratizing Networked Communications
On February 15, 2011, U.S. Secretary of State Hillary Rodham Clinton gave a speech entitled “Internet Rights and Wrongs: Choices & Challenges in a Networked World,” in which she reaffirmed America’s commitment to “internet freedom” as an increasingly vital element of our foreign policy (Clinton, 2011). In her words, internet freedom is “about ensuring that the internet remains a space where activities of all kinds can take place, from grand, ground-breaking, historic campaigns to the small, ordinary acts that people engage in every day.” Or, to put it simply, the internet is essential to the exercise of free speech and civil liberties in networked society. Recent political developments around the world appear to support this argument. Although the internet has been a platform for political speech and social action virtually since its inception (Rheingold, 1993), digital communications platforms have become an increasingly central component of resistance movements and other organized social action over the past five years, and consequently an increasingly popular target for repression, censorship, and surveillance. As Secretary Clinton herself observed, social and mobile media were important tools for both organizing and publicizing the massive antiregime protests in Iran in 2009 and Egypt in 2011, leading to government-imposed internet shutdowns in both cases, and contributing to the eventual ouster of Egyptian President Hosni Mubarak. The complete list of relevant examples is far longer; in countries ranging from China to Tunisia to Myanmar, political resistance and repression have moved from streets and cafes to mobile phones and laptops, and governments have devoted an ever greater number of resources to controlling and policing the flow of digital communications within and without their borders. In addition to its role in political struggle and change, the internet has also become central to the social, economic, and creative lives of billions of people around the world. A wealth of recent research (e.g., Deuze, 2006; Benkler, 2006; Coté & Pybus, 2007; Sinnreich, 2010; Baym, 2010) illustrates the growing importance of information and communication technologies (ICTs) in fostering cultural innovation, emerging markets, and a vibrant public sphere. Unfortunately, the challenges posed to online political speech and cultural innovation don’t end at America’s borders. Despite Secretary Clinton’s assertion that “on the spectrum of internet freedom, we place ourselves on the side of openness,” American citizens face numerous threats to free speech and civil liberties online, from both governmental and commercial institutions.
Infrastructure, Access, and Speech
We cannot understand the operation of the internet without first understanding the commercial interests of the private companies that provide its infrastructure, and control access to that infrastructure (deNardis, 2010). There is almost a complete lack of competition between these companies; at present, 97 percent of American consumers are forced to chose between at most two broadband providers (Turner, 2009). As Lawrence E. Strickling (2010), administrator of the National Telecommunications and Information Administration (NTIA), recently argued, “Broadband service providers have an incentive to use their control . . . to advantage their value-added services or to disadvantage competitive alternatives. In the absence of robust broadband competition, those providers may be able profitably to act on those incentives to the detriment of consumers and competition.” Consumers face a similar lack of choice in the wireless data market, an arena in which federal regulators possess even less power to exercise oversight.1 This lack of competition and effective regulation gives broadband and wireless providers a great deal of unchecked market power, which they have used, and have an incentive to continue using, in ways that undermine the ability of their customers to freely exchange information. In practice, we have already seen several instances of service providers exploiting this power to block communications for ideological, rather than purely profit-driven, motives. AT&T, for instance, has been criticized for censoring speech critical of President Bush during a live webcast (Marra, 2007). Similarly, Verizon Wireless has blocked text messages from NARAL, a pro-choice political group (Liptak, 2007). The consolidation of the Internet access business raises political concerns beyond these anticompetitive implications. It also contributes to an environment in which free speech is constrained by the federal government itself. One notable example is the NSA electronic surveillance program, a massive federal initiative to eavesdrop on the private communications of American citizens in the wake of the September 11, 2001 terrorist attacks. This program, which violated federal laws (ACLU, 2008), was only possible because the NSA was able to monitor the majority of communications by compelling a relatively small number of oligopolists to participate, presumably using federal regulatory power as leverage.
1 At the time of writing, AT&T has just announced its plans to acquire T-Mobile, potentially bringing the number of major American wireless data service providers from 4 to 3.
Of course, most governmental threats to free speech online come from laws, treaties, and policies that have been introduced and/or ratified by Congress. Although this is not the place for an exhaustive survey, a short list of troubling examples includes the revised Foreign Intelligence Surveillance Act (FISA), the Stored Communications Act (SCA), the Anti-Counterfeiting Trade Agreement (ACTA), the Combating Online Trade Agreements and Copyrights Act (COICA), and the as-yet-unnamed “backdoor bill,” a law requested by the White House that would give the Department of Justice unilateral power to compel ISPs to censor entire domains from the American public. Understood collectively, these examples indicate that the emerging legislative consensus accords “e-speech” less protection than traditional channels and forums (Sinnreich & Zager, 2008). In addition to these concerns, Zittrain (2009) and Moglen (2010) have pointed to the ways in which the emerging “cloud” architecture also undermines democratic and participatory communications. The consolidation of capital and information within a set of centralized corporate servers leads to the complete disempowerment of the user, to a point where ownership of all networked data skews away from local computers toward a set of centralized, corporate-owned servers. In Moglen’s words, the “dis-empowered client [is] at the edge and the server in the middle. [Information was stored] far from the human beings who controlled, or thought they controlled, the operation of the computers that increasingly dominated their lives. This was a recipe for disaster.”
Resistance and Reinvention
The constraints on free speech and civil liberties we have mentioned have met with various forms of resistance over the years. From the beginning, as Turner (2006) relates, the internet’s military and hegemonic origins have been recast as an opportunity for democratic, or even utopian, sociopolitical action. From John Perry Barlow’s seminal 1996 manifesto, “A Declaration of the Independence of Cyberspace,” to today’s position papers and legal interventions by groups like Free Press, Electronic Privacy Information Center, and Electronic Frontier Foundation (a group Barlow cofounded), there has been a consistent effort to define and preserve online free speech and civil liberties, and to develop an ethical and legal framework surrounding these issues. Similarly, we may understand the emergence of networked participatory culture (Banks & Humphreys, 2008), convergence culture (Jenkins, 2006), and configurable culture (Sinnreich, 2010), and the mass adoption of alternative communication protocols like peer-to-peer file sharing, as a largely nonideological form of resistance against the monopolization and privatization of communication. Although an individual mash-up or remix may not be positioned as a challenge to copyright law (or even produced with effective understanding of such laws), for instance, the collective interest in producing and sharing these emerging cultural forms by the billions indicates an emerging set of norms at odds with the increasingly draconian conditions under which cultural expression may legally occur. However, despite the prevalence and effectiveness of these forms of resistance, which are positioned in opposition to cultural regulation through commercial and legal means, the threats to civil liberties and free speech we have identified can ultimately be attributed to a network architecture that lends itself to exploitative and hegemonic ends. As Lessig (1999) has written so concisely, “code is law.” And, despite our ambitions of “internet freedom,” and the wealth of democratized cultural forms and breadth of political opinions currently flowering online, we believe these freedoms will continue to be undermined by a network architecture that fundamentally privileges centralized control over collective deliberation. If power corrupts, as Lord Acton’s oft-quoted phrase suggests, then absolute power over global communications will inevitably corrupt the public sphere and undermine the democratic process. Thus, we propose that the best way to vouchsafe civil liberties in the networked age is through an architectural intervention. The internet’s infrastructure must be fundamentally reimagined if it is to serve as an effective platform for democracy. Though the benefits of hierarchical DNS regimes and long-distance terrestrial backbone infrastructure are clear from an engineering standpoint, they also may be at odds with the same political values they were ostensibly built to serve (Mueller, 2002). Not only does the current architecture place the United States in an exceptional and politically unsustainable role as global regulator, it also allows for the interests of consolidated capital to be furthered above all else. In the interest of promoting civil liberties in a democratic society, our network architecture must encourage free, unregulated speech (Balkin, 2004, p. 49). To guide ourselves and others in understanding what a reimagined networked architecture would require if free speech and civil liberties are to be prioritized above all other considerations, we have developed a set of 10 “social specifications.” Our hope is that these may be understood as fundamental principles informing the development and deployment of next-generation networking technologies. Below, we will describe our own solution to these challenges, in the form of an ad hoc, wireless mesh network called MondoNet.
10 Social Specifications for a Democratized Network
The network should not be operated, maintained, or in any way reliant upon a single or minimally differentiated set of entities or technologies. No individual, entity, or group should be central to the network to the extent that its absence would measurably impact the network’s functionality or scope. Network participation should not require access to fixed, physical infrastructure of any sort.
2. Universally Accessible
The requisite technology and expertise required to participate in the network should be available at minimal cost and effort to every human being on the planet. Furthermore, all users should be able to extend the network’s content and functionality to suit their own needs, or those of others. No aspect of the network’s functioning should be reliant upon proprietary technologies, information, or capital.
The network should be resistant to both regulatory and technical attempts to limit the nature of the information shared, restrict usage by given individuals or communities, or render the network, or any portion of it, inoperable or inaccessible.
The network should enable users to choose exactly what information they share with whom, and to participate anonymously if they so desire. Users should only have access to information if they are the designated recipients, or if it has been published openly.
The network should be organized in a way that minimizes the risk of malicious attacks or engineering failure. Information exchanged on the network should meet or exceed the delivery rate and reliability of information exchanged via the internet.
The network should be organized with the expectation that its scale could reach or even exceed that of today’s internet. Special care should be taken to address to the challenge of maintaining efficiency without the presence of a centralized backbone.
The network’s density and redundancy should be great enough that it will operate persistently on a broad scale, and be available in full to any user within range of another user.
8. Fast (Enough)
The network should always achieve whatever speed is required for a “bottom-line” level of social and cultural participation. At present, we assert that the network’s data transfer rate should, at a minimum, be enough for voice-over-IP (VoIP) communications, and low-bitrate streaming video.
While the network will have the capacity to exchange information with internet users and nodes, it should also be able to operate independently. A large-scale failure or closure of internet infrastructure and content should have minimal effect on the network’s operations.
The network should be built with future development in mind. The platform should be flexible enough to support technologies, protocols, and modes of usage that have not yet been developed.
Our Solution: MondoNet
There are undoubtedly several potential technological routes to address the social specifications outlined above, and as network technology continues to evolve, we are certain that additional solutions will arise. Given today’s technological and social landscape, we believe the most promising approach is the development of a mobile, ad hoc wireless mesh network (sometimes abbreviated MANET; Rheingold, 2002). In a MANET, users connect directly to one another via WiFi or a similar wireless networking protocol, and each device becomes client, server, and router at once, sharing bandwidth and information with other devices, and enabling users to relay third-party information on behalf of their indirectly connected peers. Such a network requires no centralized infrastructure or access service provider; to join the mesh, one simply logs on within range of another peer, and to exit the network, one simply logs off. Ideally, the number and density of peers should be great enough that the network persists despite the continuing entrance and exit of individual nodes.
MANET technology is not a silver bullet to address the challenges and specifications we outlined above, and none of the existing initiatives yet fit the bill (a subject we will address in greater depth below). In order to meet our social specifications, the network, which we call MondoNet, would need to be conscientiously designed and adapted with these challenges in mind. Given the limitations of today’s networking technologies, no MANET can be universally accessible or completely decentralized at launch. Because of the geographical proximity required for participation, MondoNet would need to develop first within local communities. Over time, those communities would themselves become regionally interconnected, and, ultimately, global (or at least intracontinental) networks could be established. (Other network-based technologies such as the telephone have grown in similar fashion [Wu, 2010]) The first phase of MondoNet’s rollout would require access to internet points-of-presence to bring users in contact with one another, and to provide them with access to content and services that are currently located exclusively on internet servers. This reliance on the internet as a prosthetic network undermines MondoNet’s independence and security by bringing network data back into the range of ISP and wireless carrier scrutiny and control. However, as MondoNet grows in size and coverage, we envision a greater number of content and service providers hosting their data on MondoNet peers, rendering traditional internet access increasingly unnecessary. Over the long term, MondoNet should operate as a completely detached network, independent of the internet. Security, censorship, and surveillance are additional challenges. In a normal ad hoc network, malicious peers can obtain sensitive information simply by joining the network and capturing the data they route on behalf of third parties. We envision MondoNet as a natively encrypted platform, in which security, rather than openness, is the default status of all data. By leveraging existing, open-source encryption protocols, all intrapersonal communications would be accessible only by intended participants. For communications and network publications intended to be globally accessible, we can encrypt information with “everybody” as a recipient and integrate the public key for decoding this data into the platform, so that the user experience will be identical to today’s experience of viewing unencrypted data via the internet. Another censorship threat comes in the form of locatability; in repressive regimes, the operation of a MondoNet peer may be seen as a punishable offense. Once the signal is traced to a given device, it may be deactivated, and the operator may be liable. Although there is an inherent risk in participating in any prohibited network, we believe there is greater safety in numbers. We hope to ameliorate the risk of participation, and boost the stability and efficiency of MondoNet, by introducing “repeater” peers into the network, which would function independently of human operators. Consisting of little more than a small antenna, a power supply, and a tiny flash memory chip, repeater peers could be produced cheaply in large quantity and distributed throughout MondoNet’s geographical coverage areas, hidden within public and private spaces (e.g., attached to street signs and automobiles with magnets, buried in trash cans, stashed behind inventory on store shelves). This would drastically decrease the ability of censors to identify human operators, and increase the density, speed, and permanence of network coverage. An additional challenge we must address is access. All formats, standards, and documentation associated with MondoNet must be freely licensed or in the public domain, to ensure that (1) the technology cannot be monopolized or rendered inoperable by any given party; (2) development of the platform is accessible to all users, and will remain so for perpetuity; and (3) the cost to gain access to MondoNet software and services is as close to zero as possible.
While open-source hardware will also help lower cost of access, we see an even greater opportunity in repurposing consumers’ existing mobile devices for inclusion in MondoNet. Throughout the developed world, it is common for consumers to upgrade their mobile phones and entertainment devices every 18–24 months, discarding previous-generation models or relegating them to a back closet. By downloading an easy-to-install firmware upgrade to devices such as smartphones, portable media players, and tablets, users should be able to access MondoNet with hardware they already own but have no current use for. The immediate incentive will be access to voice and video communications, peer-to-peer file sharing (P2P), and other valuable network services without the cost or liability associated with accessing the same services over the internet. What this means for MondoNet as a whole is potential installation on hundreds of millions of devices already in the hands or homes of users. There are further challenges to the successful deployment of MondoNet. The internet benefits immensely from a centralized governance and organizational structure, in terms of network efficiency, security, and operations. IP number assignment, DNS, and protocol compatibility are just a few of the issues that will be difficult to implement without a central authority. However, we do not view these challenges as insurmountable, and other interested parties are already working to address them. For instance, Pirate Bay developer Peter Sunde recently announced a P2P DNS initiative, which would theoretically address some of these challenges. Below, we will explore this and other platforms, technologies and initiatives that we believe MondoNet can emulate, partner with, adopt, or otherwise learn from.
A MANET that would conform to the social specifications outlined earlier faces numerous well-documented engineering challenges, but we are optimistic about its potential for success. In MondoNet, we are proposing a peer-based architecture for data transmission that breaks from the server/client model that has come to dominate popular conceptualizations of the internet (Moglen, 2010; Schollmeier, 2002). With current advances in battery life, mobile routing protocols, WiFi (802.11), WiMAX (802.16), encryption techniques, and human-centered design, MANETs could emerge as a viable alternative to current hierarchical systems. In addition, several high-profile projects, such as Freedom Box, Open Mesh, One Laptop Per Child (OLPC), the Serval Project, and a Better Approach to Mobile Ad Hoc Networking (B.A.T.M.A.N.), have proposals that could fortify the initiatives set forth by MondoNet by providing specialized routing protocols, persistent stationary nodes, and security measures, in addition to piquing user awareness of and interest in mesh networking. For security, overlay encryption built upon existing TCP/IP architecture is one popular and viable solution, although encryption and key verification introduce a heavy traffic load and bandwidth restraints to nodes that often have limited battery and computational power (Mamatha & Sharma, 2010, p. 276). Recent testing has demonstrated the success of a distributed security scheme, which utilizes multiple strategies at the network and link layers to efficiently reduce network security vulnerabilities (Gada et al., 2004; Khokhar, Ngadi & Mandala, 2008; Dhanalakshmi & Rajaram, 2008). A powerful layer-two encryption with a distributed scheme would put all users on the same broadcast domain while reducing traffic load. However, in order for anonymous users to connect and push data across heterogeneous networks, layer-two encryption may not be the best option. Instead, a solution such as tcpcrypt, which uses opportunistic encryption by adding TCP header options to encrypt all traffic, would provide greater security, backwards compatibility with legacy TCP stacks, and minimized strain from negotiations on the server (36 times more connections than SSL). Using this approach to encryption provides a fallback to standard encryption if the endpoint does not support the method, and tcpcrypt is capable of staggered deployment. Another important benefit is that tcpcrypt has no requirements for preshared keys (psk) or certificates. Tcpcrypt makes “end-to-end encryption of TCP traffic the default, not the exception” (Bittau, 2010). We face an additional major security challenge, as well: the current operating frequencies are known, limited, and therefore easily jammed. Although MANETs are capable of creating diverse communication paths across a network, allowing for versatile rerouting around localized interference and overcoming a common problem with radio-based technologies, broad-interference attempts could still cripple the network. MondoNet must address potential interference from the environment, and jamming attempts from totalitarian governments and other malicious actors because, as Frankel et al. (2007) establish, the current standard from the Institute of Electrical and Electronics Engineers (IEEE; 802.11) offers no defense against jamming or flooding (p. 39).
Fortunately, wireless bandwidth is becoming more difficult to disrupt. In 2008, the Federal Communications Commission (FCC) unanimously agreed to open a large portion of the unused wireless spectrum (frequencies previously reserved for analog television) for unlicensed use to white-spaces devices, or WSDs (Wu, Wang, Liu & Clancy, 2008, p. 9). In 2010, the FCC voted to distribute unlicensed spectrum for the first time in 25 years, setting two channels aside for wireless microphone use previously used by analog television (Kim, 2010). This newly available spectrum, with its longer wavelength and better penetration, will allow wireless broadband access as part of the FCC’s National Broadband Plan. The Office of Engineering and Technology (OET) has selected nine administrators to manage and maintain the white spaces database (FCC, 2011). WSDs are capable of detecting the local frequencies in use such as television stations and avoiding interference, which are outlined in proposals from IEEE 802.11, 802.22, and the White Spaces Coalition (IEEE, 2009; Stevenson, 2009; Bangeman, 2007). An opportunistic multihop ad hoc network would create intermediate nodes through distributed storage, which would alleviate some of the problems associated with a network where “user disconnection is a feature rather than an exception” (Conti, 2007). These intermediate nodes would store data when no nodes are prepared to receive it and then forward the data to other peers within transmission range. Conti notes that an opportunistic ad hoc multihop network is “well-suited for a world of pervasive devices equipped with various wireless networking technologies” such as WiFi, WiMAX, Bluetooth, ZigBee, and plug-in servers that are “frequently out of range from a global network but are in the range of other networked devices” (p. viii). It is now more realistic than ever to lay the groundwork for MondoNet. By employing some of the improvements outlined above, the B.A.T.M.A.N. project claims to have achieved a wireless mesh consisting of 4000. Other MANET routing protocol efforts include Babel, a distance vector routing protocol based on AODV, which utilizes a unique variation of the ETX link cost estimate instead of the hop metric used for most multihop ad hoc deliveries; WING, which provides added support for radio interfaces, uses Weighted Transmission Time (WCETT) routing metric, and allows automatic channel assignment; and Roofnet’s SrcRR, which also takes advantage of the ETX metric. Babel, WING, and Roofnet are useful because all are capable of eliminating transient routing loops commonly associated with MANETs, which helps reduce unnecessary route duplication.
Several individuals and organizations have proposed hardware to enable an effective MANET. These proposals include Freedom Boxes, which are “cheap, small, low-powered plug servers” (Dwyer, 2011) that run on Linux-based software; LANdroids, pocket-sized wireless network nodes that were created to travel autonomously with troops and disaster relief teams (Menke, 2010); OLPC XO, which is a Linux-based subcomputer built around an 802.11s WiFi mesh networking protocol; Roofnet, which, in addition to providing open source routing protocols, has set up 70 nodes in Portland, OR; and Solar Mesh, a solar-powered wireless mesh network developed by McMaster University, providing hotspot coverage and WiFi endstations. In addition to hardware and routing solutions, several other relevant projects and software frameworks are under development, such as GNUnet, a software framework allowing friend-to-friend (F2F) sharing; Daihinia, which transforms a common ad hoc WiFi network into an efficient multihop solution for communities; and SMesh, a hierarchical mesh network built upon Spines in which peers rely on the infrastructure to forward packets instead of relying on other peers. Finally, Open Mesh Project and Open Source Mesh are both recently launched projects that aim to establish effective mesh networks. Unlike MondoNet, these initiatives do not require network mobility, and have not presented a clearly delineated set of social specifications informing their technological development. However, we see these projects as valuable potential partners sharing a common set of goals and interests with MondoNet.
An important set of remaining questions concerns the legal and regulatory environment for MondoNet. First, to ensure that the platform remains accessible to all potential users and available to current and future developers, we must undertake efforts to make sure that it is built on firm legal foundations. This means taking precautions not to violate any existing patents, as well as developing and sharing our own intellectual property under appropriate terms. Although we are already committed to an open license, there are dozens of free software licenses listed by the Open Source Initiative (n.d.), each offering a slightly different definition of openness and a slightly different solution for achieving it (Lamothe, 2006; Waugh & Metcalfe, 2008). Currently, our online documentation is licensed under a Creative Commons Attribution-Sharealike 3.0 license,2 but we have yet to establish which software licensing platform will be most effective in supporting our ambitions for MondoNet. Second, although we envision MondoNet as a tool made expressly for maintaining free communications in the face of institutional and governmental censorship, we will operate under the shelter of law to the greatest extent possible. This entails a number of concerns, the details of which may differ from region to region and from year to year. One example is the question of spectrum licensing; namely, what permission do individual citizens have to broadcast information at given energy levels within given frequency ranges? One of the benefits of using WiFi (specifically, IEEE 802.11) is its broad international recognition as a platform for consumer communications technologies. Of course, there are more powerful frequency ranges that may be available as well (e.g., lower frequency “white spaces” recently unlicensed in the U.S. by the FCC (FCC, 2011; IEEE, 2009; Wu, Wang, Liu, & Clancy, 2008). As we will discuss further below, a related question concerns the extent to which we can integrate these multiple networking standards into a single mesh framework.
2 Details about this license can be found at http://creativecommons.org/licenses/by-sa/3.0/
Finally, there are questions of legality surrounding the kinds of information that may be shared on MondoNet, and the treaties, laws, and precedents regarding the liability of network operators and technology providers who enable such sharing. Most sovereignties have taken pains to distinguish between permitted and unpermitted speech (e.g., intellectual property infringement, politically inflammatory messages, child pornography), and have developed systems to coerce platform providers to surveil and police their user bases. In the United States, for instance, the Digital Millennium Copyright Act (DMCA) makes it a felony to provide technology that is “designed or produced for the purpose of circumventing a technological measure that effectively controls access to a [copyrighted] work.” Even more to the point, President Obama in 2010 requested Congress to draft a new law that would require all communications technology providers to create a “back door” enabling wiretapping functionality for law enforcement (Savage, 2010; other nations, such as United Arab Emirates and Saudi Arabia, have made similar moves in recent years). Given that the previous administration illegally used its wiretapping powers to surveil American citizens without a court order (Savage & Risen, 2010), we consider such back door functionality to be anathema to the MondoNet project. We therefore anticipate that it is likely that developers, distributors and users of MondoNet will come into conflict with American and international laws, regardless of the content or legal status of their individual communications. There are also several pending questions related to technologies and platforms. Creating a system that conforms to MondoNet’s social specifications poses several hurdles. In order to establish the credibility necessary to encourage users to adopt a new conceptualization of the internet, it is essential that MondoNet adheres to human-centered design principles. Maguire (2001) identified five benefits of an effective, usable system: Increased productivity, reduced errors, reduced training and support, improved acceptance, and enhanced reputation (p. 587). For users under duress during disaster relief efforts, and for populations under oppressive government censorship, system usability, security, and interface simplicity are especially essential. To achieve this end, several key issues still must be addressed, such as determining the optimal routing protocol and encryption method. When designing any MANET, the main goal is to reliably transmit data from one node to another while still delivering a reasonable quality of service given the resource-limited environment. Although Transmission Control Protocol (TCP) is the common standard at the transport layer, it has the notable disadvantage of being slower than User Datagram Protocol (UDP). Future investigations will look at the potential benefits of traffic dispersion using multipath routing, which could improve performance and reduce the amount of energy consumed between nodes (Karygiannis, Antonakakis & Apostolopoulos, 2006; Nácher et al., 2007). Since energy consumption is always an important consideration with mobile devices, answering the design questions limiting battery life such as more routing and security power-aware protocols (Toh, 2001; Liang, & Yuansheng, 2004) are paramount but hardware remedies such as the recent three-fold improvement to the lithium-ion battery using solid state technology (Voith, 2010) are also being investigated.
Security is one of the primary obstacles to adoption in a MANET because the system is peer-based. Although opportunistic encryption through tcpcrypt provides many benefits, opportunistic encryption is inherently susceptible to active attacks. However, Bittau et al. (2010) describe an interesting approach using session IDs to prevent active attacks using tcpcrypt in MANETs (p. 7). Detecting malicious nodes in a MANET poses security problems because, unlike wired networks, anonymous, participatory MANETs are currently incapable of monitoring traffic and therefore lack an Intrusion Detection System (IDS; Karygiannis, Antonakakis & Apostolopoulos, 2006). Before MondoNet can be safely, effectively utilized by populations communicating under governments hostile to open information exchange, rigorous security protocols must be in place. Ensuring anonymity is also a complicated process that could be accomplished through tcpcrypt and by performing network address translation (NAT). Another major area of complication arises when attempting to connect networks. On top of every host running a MANET routing protocol (e.g., OSLR, B.A.T.M.A.N.), there will be a device bridging the two networks together. A very simple example would be the relationship between a wireless device and a conventional wireless router. When a wireless device is connected to the router, the devices communicate via 802.11 (WiFi) but then the router converts the transmission to 802.3 (Ethernet) via network cable. Basically, a smartphone cannot directly communicate with a network cable and needs another device such as a wireless router to allow the WiFi equipped smartphone to send and receive packets from the network. This would similarly work on 802.11s, which is the new routing protocol specifically being developed according to IEEE standards defining how interconnected wireless devices communicate. 802.11s also provides a security protocol, Simultaneous Authorization of Equals (SAE), which, although inflexible, provides protection against passive, active, and dictionary attack (Hartkins, 2008). Finally, determining an open source platform to use on repurposed mobile devices will be important to further development. One serious possibility could be a pared-down version of Fedora Linux, similar to the OS used in the OLPC XO-1. In keeping with MondoNet’s commitment to an open source environment, the repurposed devices will utilize IEEE standard compliant open firmware, and a variant of the Xfce GUI. Although there are many additional pending technical questions, previous efforts by MANET and open source developers and researchers have helped illuminate the path toward the comprehensive solution proposed here.
In this article, we have outlined the theoretical rationale, social specifications, and initial technical considerations for a large-scale, ad hoc wireless mesh network, which we call MondoNet. Although we feel this is a promising start, we hope to develop the network from an abstract idea to a concrete reality in the coming years. In the near term, this means addressing the pending questions we have outlined above, and sharing ideas, information and technology with like-minded individuals and initiatives. Over the longer term, we aim to develop and test a prototype, and to distribute the resulting technology to users and communities that may have a need for it. Ideally, this initiative, like many other open source projects, should develop a life of its own, adapting itself to uses we haven’t even considered at present, and evolving with the changing needs and technical capacities of connected individuals around the globe.
ACLU (2008). Foreign Intelligence Surveillance Act (FISA). American Civil Liberties Union.
Retrieved from http://www.aclu.org/national-security/foreign-intelligence-surveillance-act-fisa
Banks, J., & Humphreys, S. (2008). The labour of user co-creators: emergent social network markets? Convergence: The International Journal of Research into New Media Technologies, 14(4), 401-418. doi:10.1177/1354856508094660
Bangeman, E. (2007). The White Spaces Coalition’s plans for fast wireless broadband. Retrieved
Baym, N. (2010). Personal connections in the digital age. Cambridge UK ;;Malden MA: Polity.
Benkler, Y. (2006). The wealth of networks: how social production transforms markets and freedom. New Haven: Yale University Press.
Bittau, A., Hamburg, M., Handley, M., Mazieres, D., & Boneh, D. (2010). The case for ubiquitous transport-level encryption. In USENIX Security, 10(1).
Clinton, H. R. (2011). Internet rights and wrongs: choices & challenges in a networked world.
Retrieved from http://www.state.gov/secretary/rm/2011/02/156619.htm
Conti, M. (2007). Mobile multi-hop ad hoc networks from theory to reality. New York, NY: Hindawi Publishers.
Coté, M., & Pybus, J. (2007). Learning to immaterial labour 2.0: MySpace and social networks. Ephemera Theory and Politics in Organizations, 7(1), 880–106.
Dhanalakshmi, S., & Rajaram, M. (2008). A reliable and secure framework for detection and isolation of malicious nodes in MANET, IJCSNS International Journal of Computer
Science and Network Security, 8(10).
DeNardis, L. (2010). The emerging field of internet governance. Yale Information Society Project Working Paper Series. Retrieved from http://ssrn.com/abstract=1678343
Deuze, M. (2006). Participation, remediation, bricolage: considering principal components of a digital culture. The Information Society, 22(2), 63-75. doi:10.1080/01972240600567170
Dwyer, J. (2011). Decentralizing the internet so big brother can’t find you. Retrieved from http://www.nytimes.com/2011/02/16/nyregion/16about.html
FCC. (2010). National broadband plan. Retrieved from http://download.broadband.gov/plan/national-broadband-plan-chapter-3-current-state-of-the-broadband-ecosystem.pdf
FCC. (2011). TV band (white spaces) administrator’s Guide. Retrieved from http://www.fcc.gov/oet/whitespace/tvwpdda.html
Frankel, S., Eydt, B., Owens, L., & Scarfone, K. (2007). Establishing wireless robust security networks: a guide to IEEE 802.11i. National Institute of Standards and Technology.
Gada, D., Gogri, R., Rathod, P., Dedhia, Z., Mody, N., Sanyal, S. & Abraham, A. (2004). A distributed security scheme for ad hoc networks. ACM Publications, 11(1), 5-15.
Hartkins, D. (2008). Simultaneous authentication of equals: a secure, password-based key exchange for mesh networks. Proceedings of the 2008 Second International Conference on Sensor Technologies and Applications.
IEEE. (2009). 802 LAN/MAN standards committee 802.22 WG on WRANs (wireless regional area networks). Retrieved from http://www.ieee802.org/22/
Jenkins, H. (2008). Convergence culture : where old and new media collide. New York: New York University Press.
Karygiannis, A., Antonakakis, E., Apostolopoulos, A. (2006). Detecting critical nodes for MANET intrusion detection systems. Second International Workshop on Security,
Privacy and Trust in Pervasive and Ubiquitous Computing (SecPerU’06).
Khokhar, R.H., Ngadi, M.A., & Mandala, S. (2008). A review of current routing attacks in mobile ad hoc networks. International Journal of Computer Science and Security, 2(3),18-29.
Lamothe, A. (2006). Degrees of openness. Retrieved from http://linuxdevcenter.com/pub/a/linux/2006/11/09/degrees-of-openness.html
Lessig, L. (2006). Code (2nd ed.). New York: Basic Books.
Liptak, A. (2007). Verizon blocks messages of abortion rights group. New York Times. Retrieved from https://www.nytimes.com/2007/09/27/us/27verizon.html
Maguire, M. (2001). Methods to support human centered design. Internation Journal of Computer-Human Studies, 55, 587-634.
Marra, W. (2007). Pearl Jam’s anti-Bush lyrics jammed by AT&T. ABC News. Retrieved from http://abcnews.go.com/Entertainment/story?id=3467093&page=1
Menke, S. M. (2010). Retrieved from http://fedtechmagazine.com/article.asp?item_id=375
Moglen, E. (2010). Freedom In the cloud: software freedom, privacy, and security for web 2.0 and cloud computing. Internet Society New York Branch. Retrieved from http://www.softwarefreedom.org/events/2010/isoc-ny/FreedomInTheCloudtranscript.html
Mueller, M. (2002). Dancing the quango: ICANN and the privatization of international governance. In Conference on New Technologies and International Governance (School of Advanced International Relations, Johns Hopkins University, Washington, DC.
Nácher, M., Calafate, C. T., Cano, J., & Manzoni, P. (2007). Comparing tcp and udp performance in MANETS using multipath enhanced versions of dsr and dymo. Proceedings of the 4th ACM workshop on Performance evaluation of wireless ad hoc, sensor,and ubiquitous networks.
Rheingold, H. (2002). Smart mobs: the next social revolution. Cambridge, MA: Perseus.
Savage, C. & Risen, J. (2010, Mar. 31). Federal judge finds N.S.A. wiretaps were illegal. Retrieved from http://www.nytimes.com/2010/04/01/us/01nsa.html
Savage, C. (2010, Sep. 27). U.S. Tries to make It easier to wiretap the internet. The New York Times. Retrieved from http://www.nytimes.com/2010/09/27/us/27wiretap.html
Schollmeier, R. (2002). A definition of peer-to-peer networking for the classification of peer-to-peer architectures and applications. Proceedings of the First International Conference on Peer-to-Peer Computing, IEEE.
Sinnreich, A. (2010). Mashed up : music, technology, and the rise of configurable culture. Amherst: University of Massachusetts Press.
Sinnreich, A., & Zager, M. (2008). E-speech: the (uncertain) future of free expression. truthdig. Retrieved from http://www.truthdig.com/report/item/20081024_e_speech_the_uncertain_future_of_free_expression/
Stevenson, C., Zhongding Lei, G.C., Hu, W., Shellhammer, S. & Caldwell, W. (2009). IEEE 802.22: the first cognitive radio wireless regional area networks (WRANs) Standard.
IEEE Communications Magazine, 47(1), 130–138.
Strickling, L. E. (2010). Letter re: national broadband plan. GN Doc. No. 09-5. Retrieved from http://ntia.doc.gov/filings/2009/FCCLetter_Docket09-51_20100104.pdf
Turner, F. (2008). From counterculture to cyberculture : Stewart Brand, the Whole Earth Network, and the rise of digital utopianism. Chicago: University of Chicago Press.
Turner, S. D. (2009). Dismantling digital deregulation: toward a national broadband strategy. FreePress.net. Retrieved from http://www.freepress.net/files/Dismantling_Digital_Deregulation.pdf
Waugh, P. & Metcalfe, R. (2008). The foundations of open: evaluating aspects of openness in software projects. Retrieved from http://pipka.org/blog/2008/07/23/the-foundations-of-openness/
Wu, T. (2010). The master switch: the rise and fall of information empires. New York: Knopf.
Wu, Y., Wang, B., Liu, K. J., & Clancy, T. C. (2008). Repeated open spectrum sharing game with cheat-proof strategies. IEEE Transactions on Wireless Communications, 20(20), 1 12.
Zittrain, J. (2009). Lost in the cloud. The New York Times. Retrieved from http://www.nytimes.com/2009/07/20/opinion/20zittrain.html
City of Austin’s Wireless Mesh Network System
MORE MESH PROJECTS
by Klint Finley / January 28, 2011
In Cory Doctorow’s young adult novel Little Brother, the protagonist starts a wireless ad-hoc network, called X-Net, in response to a government crack-down on civil liberties. The characters use gaming systems with mesh networking equipment built-in to share files, exchange message and make plans.
The Internet blackout in Egypt, which we’ve been covering, touches on an issue we’ve raised occasionally here: the control of governments (and corporations) over the Internet (and by extension, the cloud). One possible solution, discussed by geeks for years, is the creation of wireless ad-hoc networks like the one in Little Brother to eliminate the need for centralized hardware and network connectivity. It’s the sort of technology that’s valuable not just for insuring both freedom of speech (not to mention freedom of commerce – Egypt’s Internet blackout can’t be good for business), but could be valuable in emergencies such as natural disasters as well.
Here are a few projects working to create such networks. Wireless ad-hoc networking has been limited in the past by a bottleneck problem. Researchers may have solved this issue for devices with enough computational power. The U.S. military is alsoinvesting in research in this area.
The OLPC’s XO has meshnetworking capabilities. And some gaming systems, such as the Nintendo DS, have mesh networking built in. But we want to look at projects that are specifically aimed at replacing or augmenting the public Internet.
Openet is a part of the open_sailing project. Openet’s goal is to create a civilian Internet outside of the control of governments and corporations. It aims to not only create local mesh networks, but to build a global mesh network of mesh networks stitched together by long range packet radio. See our previous coverage here.
Netsukuku is a project of the Italian group FreakNet MediaLab. Netsukuku is designed to be a distributed, anonymous mesh network that relies only on normal wireless network cards. FreakNet is even building its own domain name architecture. Unfortunately, there’s no stable release of the code and the web site was last updated in September 2009.
Not to be confused with the mesh networking hardware vendor of the same name, OPENMESH is a forum created by venture captalist Shervin Pishevar for volunteers interested in building mesh networks for people living in conditions where Internet access may be limited or controlled.
Pishevar came up with the idea during the protests in Iran in 2009. “The last bastion of the dictatorship is the router,” he told us. The events in Egypt inspired him to get started.
It’s a younger project than Openet and Netsukuku, but it may have more mainstream appeal thanks to being backed by Pishevar. It’s not clear how far along Openet is, and Netsukuku’s seems to be completely stalled so a new project isn’t entirely unreasonable. Update: One commenter points out that Netsukuku’s developers have checked in code as recently as two weeks ago, so although the site hasn’t been updated the project isn’t stalled.
by Klint Finley / January 31, 2011
Last week we told you about three projects to create a government-less Internet by taking advantage of wireless mesh networking. Wireless mesh networks are networks that don’t require a centralized authority to create networks. These can provide an alternative way to communicate and share information during a crisis such as a natural disaster or civic unrest.
Many of you followed-up by telling us about several other interesting projects, such as P2P DNS to Tonkia. Most importantly, there are at least four other projects that should have been on our original list.
Daihinia is a commercial project that provides software that essentially turns Windows PCs into wireless repeaters. The company’s software makes it possible to use a desktop or laptop with a normal wireless card to “hop” to a wireless access point while out of range of that access point. There’s no Macintosh version, but it’s being discussed.
Digitata is a sub-project of open_sailing‘s Openet, which we mentioned in the previous installment. Digitata is focused on bringing wireless networks to rural areas of Africa. The group is creating open source hardware and software, including its own own IP layer for mesh networking called IPvPosition (IPvP).
Freifunk (German for “free radio”) is an organization dedicated to providing information and resources for mesh networking projects. Its website has a list of local mesh networks all over the world, from Afghanistan to Nepal to Seattle.
One of its main resources is the Freifunk firmware, a free router firmware optimized for wireless mesh networking. Users can replace the standard firmware on their routers with Frefunk’s firmware, enabling them to build mesh networks with cheap off the shelf hardware.
Freifunk also develops a protocol Better Approach To Mobile Adhoc Networking, or B.A.T.M.A.N., an alternative to the older Optimized Link State Routing Protocol (OLSR).
wlan ljubljana and nodewatcher
wlan ljubljana is a wireless mesh network in Ljubljana, Slovenia. In addition to providing its users with Internet access, it appears to also feature a local network.
wlan ljubljana is working with volunteers in other cities in Slovenia to create more local networks, and has created its own firmware package for routers called nodewatcher. Like Freifunk, nodewatcher is based on the embeddable Linux distribution OpenWrt. nodewatcher is designed to be easy to use for a non-technical user.
Here are a few more resources: