Exploratory Advanced Research Program

FHWA White Paper On Mobile Ad Hoc Networks

Classes of MANETs

MANETs are a broad category of ad hoc networks. This category can be broken down into several subcategories, which encompass other types of ad hoc networks. Over the last five or so years, MANET research has shifted away from the pure general purpose paradigm towards more specific types of uses for MANETs.⁽²⁾ Each type, some of which are further described below, has unique benefits and challenges for which new protocols must be used or created to fully utilize the application.

Vehicular Ad Hoc Networks

A vehicular ad hoc network (VANET) is a specialized mobile ad hoc network consisting of vehicle nodes that communicate wirelessly. The most prevalent means for this type of communication is via the Institute of Electrical and Electronics Engineers (IEEE) 802.11, or Wi-Fi, protocols.⁽⁷⁾ Modern vehicles carry anywhere between 20 to 50 antennas, which facilitate communications within these channels.^[3]

Unlike general purpose MANETs, VANETs are not restricted by typical mobile device battery limits because of their reliance on the built-in vehicle battery, which—in addition to having a much greater capacity than portable electronics batteries—can recharge as the vehicle is moving. VANETs are, however, usually constrained by road and other vehicle-specific characteristics.^[4] Overall, VANETs exhibit a unique network topology when compared to MANETs. VANETs must accommodate more dynamic changes to the number and density of connected nodes, can span greater distances through a large fleet of vehicle nodes, and exhibit high mobility of nodes due to vehicle speeds. An increase in mobility intensity, which is a node that covers a large surface area, will generate more contact opportunities between nodes but also more limited opportunities for establishing and maintaining a connection.⁽⁸⁾ When node density is high and too many radios try to access the spectrum simultaneously, VANETs may experience spectrum scarcity, meaning there is less available space for communication on wireless channels. In traditional infrastructure networks, such as cell phones and Long Term Evolution (LTE)/Long Term Evolution-Advanced (LTE-A),^[5] this problem is minimized through synchronous communications where a tower or central radio directs all other nodes when to schedule their transmissions. VANETs, and MANETs in general, transmit via asynchronous communications, which do not have that central authority.⁽⁹⁾ One possible solution for addressing this issue is Dynamic Spectrum Access (DSA), which allows unlicensed users to access and take advantage of spectrum that are not being used, at the time, by licensed users.⁽²⁾ While there have been many research efforts focused on DSA, there are no direct applications currently possible.^(10,11) Some of the potential applications of VANETs include collision avoidance, road obstacle warning, and safety message disseminations.⁽¹¹⁾ Use-cases can be categorized by the scope of impact as seen in Appendix B.

Flying Ad Hoc Networks

Compared to VANETs and other surface MANETs, flying ad hoc networks (FANETs) are a relatively new and less explored form of MANETs that face even greater spatial challenges. FANETs generally consist of an ad hoc network between small, mobile unmanned aerial vehicle (UAV) nodes. Compared to existing UAVs that are much larger and use satellites to relay data to ground-based users, FANET UAVs have a much lower acquisition and maintenance cost, can cover larger dynamic areas at any given moment, and their ability to relay information or form a network is not dependent on any single node.⁽¹²⁾ Compared to VANETs, average distances between FANET nodes tend to be much larger. Moreover, while propagation issues caused by buildings and terrain can be less problematic for FANETs, establishing and maintaining a link between each node may require different radio links, hardware circuits, sensors, and mobility handling capabilities caused by the large distances.⁽¹²⁾

Because of the 3-D nature of their movement and the speed of nodes flying in varying directions, typical solutions and standards for VANETs are not always applicable to FANETs. For example, while end-to-end routes can be used for VANETs, continuous yet typical topology changes can create significant delays or route failures, especially when transmitting data packets such as multimedia files. Existing VANET routing protocols, such as dynamic source routing (DSR),^[6] ad hoc on-demand distance vector (AODV) routing,^[7] and optimized link state routing (OLSR)^[8] are also not ideal for the packet forwarding in this scenario. Some research suggests that this type of issue can be solved with beaconless opportunistic routing protocols, which minimize packet transmission loss by identifying relay nodes to forward packets, and more easily recovering from route failures.⁽¹³⁾ Similar to VANETs, FANETs also have battery capacity limits related to how much weight a FANET node can carry. FANETs also must consider collision avoidance mechanisms. In one example, the National Aeronautics and Space Administration has explored field tests using DSRC within unmanned aircraft systems to support tactical separation among nodes.^[9]

Some possible applications of FANETs in transportation are uninterrupted live video sequences of traffic and other related safety incidents, general traffic monitoring, and natural disaster event response. FANETs can also provide internet access to infrastructure-poor regions through floating or orbital nodes. Google’s Project Loon and Facebook’s Internet.org provide LTE-equipped floating balloons and solar-powered drones, respectively, with the intent of creating a mobile flying cellular infrastructure. Additionally, other projects, such as SpaceX and OneWeb, hope to provide lower orbit satellites that can create end-to-end communications. While floating applications have demonstrated initial success, orbital applications have not yet been implemented.⁽⁶⁾ Transportation-related communications may have the potential to exploit these types of networks if needed.

MANET Class Variations

As research has advanced, there have been improvements to how each MANET application operates independently and with other networks. These improvements include better routing protocols, efficient end-to-end service, and more.

Opportunistic Networks

Opportunistic Networks (OppNets), derive from MANETs and delay/disruption tolerant network (DTN) topologies, which use technologies to address communication delays or complete disruptions. OppNets can arise as a result of a network’s dynamically changing topology. While OppNets are not restricted to mobile nodes, their use of a “store-carry-and-forward” paradigm lends them particularly well to distributing data to and from portable devices, such as smartphones.⁽²⁾ However, OppNets are not implemented widely, and the current research in this field is relatively immature as compared with other network technologies. Other reasons for this lack of adoption include lack of incentive from mobile operators to share or relinquish control over some of their fixed networks,⁽⁶⁾ and prohibitive battery consumption for nodes that are in constant signal search mode.⁽¹⁴⁾

Heterogeneous Networks

Heterogeneous Networks (HetNets) refer to the deployment of different radio access technologies that create a network using diverse capabilities such as varying bit rates and range. The goal of a HetNet is to acquire optimal use of spectrum bands and energy through heterogeneous nodes.⁽²⁾ While MANETs consist of spatially dynamic nodes, HetNets can consist of non-fixed and fixed nodes. In some application tests, fixed and non-fixed heterogeneous applications, such as vehicle-to-infrastructure (V2I) VANETs, demonstrate lower end-to-end communication delays and higher throughput performance.⁽⁷⁾ This concept lends itself well to transportation applications since HetNets are constantly evolving and would benefit from access to different technology solutions.

Internet of Things

MANETs have a direct correlation with the emerging technology of Internet of Things (IoT). The IoT describes an interconnected network of physical objects (e.g., buildings, vehicles, and devices) that can collect, share, and exchange data. This system allows for better incorporation of the physical world into computer-based systems. Any object within the IoT that moves (vehicles and smart devices as an example) could work as a node in a MANET. Further, many IoT applications may rely on the use of MANETs. IoT application can also rely on machine-to-machine communication protocols.

^[3] Personal Interview with USDOT Subject Matter Expert, March 2017.

^[4] This is not necessarily a limitation because restricted paths can lead to simpler routing protocols.

^[5] Both of these standards use high spectrum flexibility and varying channel bandwidth for high speed wireless communication. For more information, see http://www.sciencedirect.com/science/book/9780123854896. Accessed 4-6-2017.

^[6] A reactive routing protocol that computes routes as necessary, not periodically, and then maintains them.

^[7] A low network-utilizing reactive protocol that, like DSR, computes routes as necessary and maintains them but also uses destination sequence number delivered from the route sender node.

^[8] A proactive routing protocol that constantly sends topological information to all available network nodes and continuously maintains routes. For more information on each of these three protocols, see: Narsimaha, V.B. (2012).

^[9] National Aeronautics and Space Administration, Ames Research Center (July 2017). Meeting Presentation.