Position location techniques and applications - Part 2: Ad hoc and sensor network scenarios

[Part one begins a discussion of the basic evolution of wireless networks and how position location could be considered from the networking point of view.]

5.1.4 Ad Hoc and Sensor Network Scenarios
This section introduces a general overview of wireless mobile ad hoc networks (MANETs) and sensor networks, and their evolution, applications, and selected issues such as connectivity and scalability. This section is intended to prepare the reader for a subsequent scenario.

A MANET (see examples in Figure 5.8) is a dynamic network formed by a collection of arbitrarily located wireless mobile nodes without the use of existing network infrastructure or centralized administration. MANETs begin with at least two nodes broadcasting their presence (beaconing) with respective address information; a mobile node communicates with another directly via a single hop whenever a radio channel with adequate propagation characteristics is available between them.

If devices are not directly connected, multihop communication is necessary where one or more intermediate nodes must act as a sequence of relay nodes that forward the traffic of the communicating nodes. As more nodes join the network or some of the existing nodes leave, the topology updates become more numerous and complex, and usually more frequent, thus diminishing the network resources available for exchanging user information [10, 12].

Wireless ad hoc networks have some characteristics similar to wireless networks, but the differences are what make them interesting to study. Some of these differences follow:

Mobility of nodes: An ad hoc wireless network does not depend on a fixed infrastructure. The network is reconfigurable following mobility and communication needs.

Point-to-point communication: In contrast to cellular networks where communication between two cellular devices is made throughout a base station, in ad hoc networks communication between nodes is point to point. This means that two nodes within their sensivity area can communicate directly and without a base station's help.

Self-organizing: An ad hoc wireless network is self-organizing and adaptive to the communication requirements of the moment. This is a fundamental feature of ad hoc networks. A terminal has to be capable of detecting other terminals that are within reach, and it also must learn which terminals are reachable through the network. Autoconfiguration also concerns detecting services available in the networks. Because of this characteristic, most of the protocols for traditional networks cannot easily be adapted to ad hoc networks.

The Department of Defense initiated the DARPA GloMo Information Systems program in 1994, supporting Ethernet-type multimedia connectivity among wireless devices. WINGs at University of California–Santa Cruz deployed a flat peer-to-peer architecture, and the MMWN project from GTE Internetworking used a cluster-based architecture. Tactical Internet implemented by the U.S. Army in 1997 is the largest-scale deployment of a mobile wireless multihop packet radio network; DSSS and TDMA radio are used with data rates in the range of tens of kilobits per second.

In 1999, ELB ACTD demonstrated the feasibility of Marine Corps war-fighting concepts that require over-the-horizon, beyond-LOS communications from ships at sea to Marines on land via an aerial relay. Lucent's WaveLAN and VRC-99A were used to build the access and backbone connections. In 1990, the IEEE, when developing IEEE 802.11, replaced the term packet radio network with ad hoc network. At present, the Bluetooth system is perhaps the most promising technology in the context of MANETs [12,22], but there are problems with scalability, security, and coverage. These problems have been addressed and new technologies such as ZigBee have positioning capabilities.

A MANET can be used to exchange information between the nodes, allowing them to communicate with remote sites that otherwise would not be reachable or where no infrastructure (fixed or cellular) is available. Examples include tactical networks with related applications to improve battlefield communications or survivability, rescue operations, disaster recovery, collaborative work in remote areas, public events, construction sites, and so on [22]. Recently, the introduction of technologies such as Bluetooth and WiFi are helping to enable eventual commercial MANET deployments outside the military domain [10, 12].

Futuristic applications include home networks where devices can communicate directly to exchange information, encompassing computers, microwave ovens, door locks, and other appliances interconnected by a wireless network [22]. Sensor networks have been proposed for environmental monitoring, where the networks could be used to forecast water pollution or to provide early warning of an approaching tsunami. Perhaps the most far-reaching applications are autonomous networks of interconnected home robots that clean, perform security surveillance, and so on.

The multihop nature and the lack of fixed infrastructure add a number of issues and design constraints that are specific to ad hoc networking.

Connectivity issue. In MANETs, because nodes can move arbitrarily, the network topology can change frequently and unpredictably, resulting in connectivity changes, frequent network partitions, and possibly packet losses.

MAC issue. The MAC scheme is also difficult in ad hoc networks due to the time-varying network topology and the lack of centralized control; instead, carefully designed distributed MAC techniques must be used for channel resources [10, 30].

Scalability issue. Many MANET applications involve large networks with tens of thousands of nodes. Scalability is critical to the successful deployment of these networks, presenting many challenges in addressing, routing, location management, security, and so on [12].

Clustering issue. Efforts to maintain a relatively stable effective topology, limit far-reaching reactions to topology dynamics, rearrange clusters, and assign nodes indicate that establishing a good clustering algorithm can be complex, requiring excessive processing and communications overhead [47]. On the other hand, it is known that clustering is useful in improving MAC resource management and in stabilizing network topology [41] (See Lawson and Denison [43] for an excellent overview of this issue.)

All these challenges are potential sources of service impairment in MANETs, and hence may degrade the QoS seen by users of the network, [10].