Wireless Sensor Network Recent advances in electronics and wireless communications have enabled the design and manufacture of sensors with low power consumption, small size, reasonable cost, and diverse applications. These small sensors can collect various
Wireless Sensor Network
Recent advances in electronics and wireless communications have enabled the design and manufacture of sensors with low power consumption, small size, reasonable cost, and diverse applications. These small sensors can collect various environmental data according to sensor type, process it, and transmit the information, giving rise to the idea of creating and expanding Wireless Sensor Networks (WSNs).
A sensor network consists of a large number of sensor nodes distributed widely throughout an environment to collect information from it. The locations of sensor nodes are not necessarily predetermined or known. This makes it possible to deploy them in dangerous or inaccessible locations.
This also means sensor-network protocols and algorithms must support self-organization. Another unique characteristic of sensor networks is the ability of sensor nodes to cooperate and coordinate with one another. Each sensor node has an onboard processor. Instead of sending all raw data to a central point or processing node, it first performs simple preliminary processing on the collected data and then transmits the partially processed information.
Although each individual sensor has limited capability, combining hundreds of small sensors creates new possibilities. The strength of wireless sensor networks lies in their ability to deploy large numbers of small nodes that can assemble and organize themselves and be used for tasks such as coordinated routing, environmental monitoring, and monitoring the health of structures or system equipment.
Wireless sensor networks have a very broad range of applications, from agriculture, medicine, and industry to military uses. For example, one common application of this technology is monitoring a remote environment. A leak in a large chemical plant, for example, can be monitored by hundreds of sensors that automatically form a wireless network and quickly notify a control center when a chemical leak occurs.
Unlike older wired systems, these systems reduce network configuration and deployment costs and replace thousands of meters of wiring with small devices approximately the size of a coin.
A Wireless Sensor Network (WSN) is a wireless network of autonomous sensors distributed across an area to collectively measure physical quantities or environmental conditions such as temperature, sound, vibration, pressure, motion, or pollutants at different locations. Sensor networks were initially developed for military applications such as battlefield surveillance. Today, however, wireless sensor networks are used in industry and many civilian applications, including industrial process monitoring and control, equipment health monitoring, environmental or home monitoring, healthcare, smart homes, agriculture, and traffic control.
In addition to one or more sensors, each network node is generally equipped with a radio transmitter and receiver (or another wireless communication device), a small microcontroller, and an energy source, usually a battery. The size of a sensor node varies with its packaging and can potentially be reduced to the size of a grain of sand; although components at that microscopic scale still need to be manufactured. Similarly, the cost of each sensor node can range from several hundred dollars to a few cents, depending on the required size and complexity. Cost and size constraints on sensor nodes lead to limitations in resources such as energy, memory, processing speed, and bandwidth.
A sensor network generally forms a wireless ad-hoc network, meaning each node uses a multi-hop routing algorithm. (Many nodes forward an information packet until it reaches the base station.) Wireless sensor networks are currently an active research area in computer science and communications, with numerous workshops and conferences held on the subject each year.
One important requirement of a sensor network is time synchronization. The importance of timing in sensor networks makes disruption of sensor synchronization one of the primary targets for attacks on these networks. An attacker may try to prevent correct synchronization in various ways, such as disrupting synchronization messages, altering or forging them, delaying time-sensitive messages, compromising some nodes, and sending false synchronization messages through them. Although several synchronization methods for sensor networks have been introduced in recent years, no comprehensive method has yet been presented that simultaneously satisfies both security and efficiency requirements.
Applications
Wireless sensor networks have many and varied uses. For example, in commercial and industrial applications they are used for data monitoring and in situations where wired sensors are difficult or expensive to deploy. These networks can also be deployed in desert environments and remain operational for years. Another application is detecting an intruder entering a controlled area and then tracking the intruder.
Other applications include residential monitoring, tracking moving targets, nuclear reactor control, fire detection, traffic monitoring, and more.
Environmental-area monitoring
Environmental control or monitoring is one use of wireless sensor networks. In environmental monitoring, wireless sensor nodes are distributed across an area where phenomena or events need to be observed. For example, large numbers of transmitter/receiver nodes can be deployed on a battlefield to detect enemy intrusion instead of using landmines.
When a sensor detects an event being monitored—such as heat, pressure, sound, light, magnetic properties, vibration, and so on—the event must be reported to a base station. Depending on the network application, the base station performs an appropriate action, such as sending a message to the internet or satellite or processing the data locally.
Healthcare monitoring
This can involve implanted sensors, wearable sensors, or environmental sensors used for healthcare, each of which detects and monitors health-related data from living organisms.
Environmental monitoring
Air pollution monitoring, forest fire monitoring, landslide monitoring, water pollution monitoring, and monitoring weather changes to prevent or reduce the consequences of natural disasters such as floods and storms.
Industrial monitoring
Monitoring machine performance and health, system operation—including monitoring another network sensor—data center operation, and the structural health of engineering structures, which itself is a broad field.
Characteristics
Unique characteristics of wireless sensor nodes:
Small-scale sensor nodes
Limited power that can be stored or replenished
Harsh environmental conditions
Node failures
Node loss
Dynamic network topology
Communication failures
Node heterogeneity
Large-scale deployment
Autonomous operation
Sensor nodes can be thought of as small computers. At a basic level, their common structure and components generally include a processor with limited computational capability and limited memory, sensors with application-specific circuitry, a communication device—usually a radio transceiver or sometimes optical communication—and an energy source, usually a battery. Base stations are one or more prominent components of a wireless sensor network (WSN) with greater computing, energy, and communication resources. They act as gateways between sensor nodes and the end user.
A key goal for sensor nodes is to achieve small size at low cost. With these goals in mind, current sensor nodes are largely prototypes. Miniaturization and cost reduction are ongoing and future goals driven by advances in MEMS and NEMS; some very small sensor nodes have been demonstrated, while others remain in the research stage. A general overview of network usage, base stations, components, and related SNM topics is available.
Standards
While mainstream computer networking follows established standards, formal standards adopted for wireless sensor networks include ISO 18000-7, 6LoWPAN, and WirelessHART, along with several other standards investigated by researchers in this field.
ZigBee
Wibree
IEEE 802.15.4-2006
Software
Energy is a scarce resource in wireless sensor nodes and determines the lifetime of a wireless sensor network (WSN). Such networks may contain large numbers of nodes deployed in diverse environments, including remote or hostile areas, with ad-hoc communications. Therefore algorithms and protocols must address the following requirements:
Maximizing lifetime.
Robustness and fault tolerance
Self-configuration; and automatic deployment
Some active research topics in WSN software include:
Security
Mobility and node departure (when sensor nodes or base stations are moving)
Middleware, an intermediate software layer between software and hardware.
Operating system
Operating systems for wireless sensor network nodes are generally less complex than general-purpose operating systems. This is due both to the specific requirements of sensor-network applications and to resource constraints in sensor-node hardware. For example, sensor-network applications generally do not require interactive use like a conventional computer. Therefore, the operating system does not need to support the same user applications; moreover, memory constraints and hardware memory mapping make features such as virtual memory both unnecessary and impractical. Wireless sensor network hardware is not fundamentally different from traditional embedded systems, so embedded operating systems such as eCos or µC/OS can be used. However, operating systems designed specifically for sensor networks may differ from traditional embedded real-time systems. They often do not provide real-time support. TinyOS was perhaps the first operating system designed specifically for sensor networks. Unlike many other operating systems, TinyOS uses an event-driven programming model rather than a multithreaded model with several logical execution paths running concurrently.
TinyOS applications are composed of components and tasks whose execution continues in response to external events such as incoming data or sensor readings.
TinyOS signals the appropriate component when an event occurs, and components can post tasks to be scheduled later by the TinyOS kernel. Both TinyOS itself and applications written for TinyOS are implemented using C-related programming technologies. NesC was designed to detect race conditions—situations arising when instructions execute concurrently and it is uncertain which one will finish first—between tasks and event handlers.
There are also operating systems that allow programming in C, including Contiki, MANTIS, BTnut, SOS, and Nano-RK.
Contiki is designed to support dynamic loading over a network and runtime loading using standard ELF files. The Contiki kernel is event-driven, but the system also supports multithreading when required. It includes lightweight threading mechanisms that provide programming similar to threads while using very little additional memory.
Unlike event-driven systems, MANTIS and Nano-RK use preemptive multithreaded kernels, while Contiki can support both models. With preemptive multithreading, users do not need to explicitly yield the processor to other processes. Instead, the kernel divides processor time among active processes and decides which process runs, which can simplify programming.
Sensor-network operating systems such as TinyOS, Contiki, and SOS use lightweight designs; SOS in particular supports dynamically loadable modules. A complete SOS operating system is built from smaller components and also focuses on support for dynamic memory management.
Middleware
Considerable recent research has focused on designing middleware for wireless sensor networks. These approaches can generally be classified as distributed databases, mobile agents, and event-based systems.
Programming language
Programming sensor nodes is difficult compared with conventional computer systems. The natural resource constraints of these nodes have led to new programming models. However, most current nodes are programmed in C.
C@t: time-and-space point computing
DSL: distributed temporal compositions
Galsc
Nec C
Proto thread
SNACK
SQTL
Algorithm
A WSN consists of a large number of sensor nodes. Therefore, algorithms for WSNs are distributed algorithms. In a WSN, energy is the scarce resource; and one of the most energy-intensive operations is data transmission. For this reason, algorithmic research in WSNs focuses heavily on energy efficiency. Energy-aware algorithms for transferring information from sensor nodes to a base station commonly use multi-hop communication—node to node toward the base station—because radio transmission energy cost grows significantly with transmission distance.
Algorithmic approaches differ from protocol approaches in WSN research in that the mathematical models used are more abstract. They are more general but sometimes less realistic. Protocol-oriented models include simulators designed specifically to evaluate sensor-network performance, such as TOSSIM, which is part of TinyOS, older network simulators such as NS-2, the visual OPNET simulator for analyzing and simulating computer and telecommunications networks from small to global scales, and SENSIM based on OMNeT++ for wireless sensor-network analysis. A wide range of WSN simulation tools is available, including database-oriented tools such as CRUISE WSN.
Simulation
Data visualization
Data collected from wireless sensor networks is usually stored in numerical form in a central database. Various applications such as TosGUI, Sensor, MonSense, and GSN help users explore large volumes of data. Geospatial consortium standards can also improve interoperability and data encoding, allowing heterogeneous sites on the internet to be organized and enabling users to monitor wireless sensor networks individually through web-based software.
