Wireless sensor networks (WSN). How to Deploy Wireless Sensor Networks in Difficult Industrial Environments New Devices in Wireless Sensor Networks

Kireev A.O., Svetlov A.V. WIRELESS SENSOR NETWORKS IN THE FIELD OF PROTECTION TECHNOLOGIES

The well-established term "wireless sensor network" (WSN) denotes a new class of wireless systems, which are a distributed, self-organizing and fault-tolerant network of miniature electronic devices with autonomous power supplies. Intelligent nodes of such a network are capable of relaying messages along the chain, providing a significant system coverage area with low transmitter power and, consequently, high energy efficiency of the system.

Currently, much attention is paid to the organization of automated monitoring of territories in order to obtain operational information about the presence of an intruder, his movement and unauthorized actions in territories adjacent to especially important (nuclear, government, military) facilities, to the state border, or located in the zone of responsibility reconnaissance subdivisions (monitoring of front sections, rear communications of the enemy). To rationally solve these problems, it is necessary to use a new generation technical means and algorithms that are fundamentally different from those currently used. The most promising direction in this area should be recognized as the creation of wireless sensor networks. They make it possible to provide total targeted monitoring of large areas.

With regard to the security systems of objects, the WSS must detect and classify the intruder, determine the coordinates, and predict the trajectories of his movement. Possessing distributed intelligence, the system independently ensures a change in the direction of information flows, for example, bypassing failed or temporarily non-functioning nodes, organizes reliable transmission of information throughout the controlled territory and to the central point.

WSNs are also promising, in which the transceiver of each sensor will actually be an object detection sensor (the effect of lowering the carrier level in the radio channel due to the appearance of an object in the network coverage area).

To ensure high reliability and protection of transmitted information in the WSN, it is necessary to develop own radio protocols that are resistant to changes in the characteristics of the communication channel, radio jamming, interception and data imitation. In this case, it is advisable to use spread spectrum technologies - DSSS (direct numerical sequence) and FHSS (frequency hopping) methods.

As for the mechanisms of access to the data transmission medium, there appear mutually exclusive requirements for high energy efficiency of the system and minimum time delays for data propagation in the WSN. Using CSMA/CA (Carrier Sense Multiple Media Access with Collision Avoidance) as the basic algorithm has its own drawback - network devices must be in the mode of constant listening to the air, which leads to an increase in power consumption. In completely asynchronous networks, this algorithm is inefficient.

The most acceptable in such a situation is the "slot" CSMA / CA algorithm, which combines the principles of synchronized access (TDMA time division) and access on a competitive basis.

Among the open standards in the field of wireless sensor networks, only the ZigBee standard has been ratified to date, based on the previously adopted standard 802.15.4, which describes physical layer(PHY) and media access layer (MAC) for wireless personal area networks (WPAN). This technology was originally developed for tasks that do not require high data transfer rates. The devices of such networks should be as cheap as possible, with ultra-low power consumption.

Among the undoubted advantages of ZigBee solutions, significant disadvantages should also be noted. For example, the presence of three different classes of devices (coordinators, routers, and terminal devices) significantly reduces the fault tolerance of the network in the event of failure of its individual elements. In addition, such a construction requires planning the placement of devices at the design stage of the system; accordingly, the network's resistance to changes in the topology is sharply reduced.

All of these shortcomings are deprived of Mesh networks - multi-cell peer-to-peer networks in which each node can relay packets during the delivery process. The nodes of such a network are equal and interchangeable - as a result, the scalability of the system improves, and its fault tolerance increases.

The wireless sensor network of the security system should monitor the maximum possible area. In this regard, one of the main requirements for choosing the element base for creating a radio channel between individual network nodes is the maximum communication range. Operation in the 433 MHz frequency band (open for free use in Russia) has a number of advantages compared to operation in the 2.4 GHz microwave band (for which the main range of ZigBee devices is produced). So, in the 433 MHz band, the reliable communication range is several times greater than in the 2.4 GHz band, with the same transmitter power. In addition, devices operating in the 433 MHz band have a fairly good resistance to the action of obstacles in the path of radio waves, such as precipitation, terrain changes, trees, etc. 433 MHz radio waves propagate much better in confined spaces such as subway tunnels , city streets, etc. than 2.4GHz radio waves. The advantage of the 2.4 GHz band in data transfer rate is not critical in the field of security technologies, since the amount of information transmitted is usually insignificant and limited to tens of bytes (with the exception of telemetry).

Thus, the choice of a transceiver for the WSN site for the protection of objects will be carried out in the 433 MHz band. Transceivers must have high energy efficiency (supply voltage no more than

3.3 V, low consumption currents), operate in the temperature range of minus 40 ... +85 °С.

Among the many ISM transceiver ICs, XE-MICS transceivers occupy a special place. For use in wireless sensor networks, 2 chips from this company are suitable: XE1203F and

These are integrated single-chip half-duplex transceivers built according to the direct (Zero-IF) conversion scheme, providing 2-level frequency shift keying without phase break (CPFSK) and NRZ coding. Thus, the type of carrier modulation implemented in XEMICS transceivers makes it possible to rationally use the operating frequency band.

Common for XE1203F and XE1205F transceivers are ultra-low power consumption: operation in the supply voltage range of 2.4 ... 3.6 V, consumption currents:

0.2 µA in sleep mode;

14 mA in receive mode;

62 mA in transmit mode (+15 dBm) .

Operating frequency band: 433-435 MHz. Temperature range: minus 40. +85°С. Transceiver receivers

The beliefs are identical to each other and are built according to a direct frequency conversion scheme. Built into these modules is a frequency synthesizer based on a sigma-delta PLL with 500 Hz steps.

The receivers have an indicator of the level of the received signal RSSI (Received Signal Strength Indicator), which, combined with the ability to program the output power, allows you to implement the idea of ​​​​adaptive power management. The transceiver includes a frequency control device FEI (Frequency Error Indicator), which allows you to obtain information about the frequency offset of the receiver local oscillator and organize the AFC.

The transceivers also have a pattern recognition feature that allows the transceiver to detect a programmable word (up to 4 bytes) in the received data stream. last feature can be used to identify modules in the WSN, which will reduce the number of overhead bytes in the transmitted packet.

The main differences between the two modules are manifested in the use of different spectrum spreading methods.

The XE1203F transceiver has a hardware Direct Sequence Spread Spectrum (DSSS) block. When DSSS mode is activated, each data bit is encoded with an 11-bit Barker code: 101 1011 1000 or 0x5B8h. The autocorrelation function of the Barker code has a pronounced autocorrelation peak.

Unlike the XE1203F, the XE1205F transceiver (and the DP1205F module based on it) is a narrowband device. The smallest value of the internal band pass filter that can be set by the 2-bit configuration register is 10 kHz (using special additional settings, this value can even be reduced to 7 kHz!). Quantity possible channels in this case

This capability allows the XE1205F to be used for specific narrowband applications. Band narrowing can be used if the data rate and frequency deviation do not exceed 4800 bits and 5 kHz, respectively, and provided that clock frequency the reference oscillator is stabilized by a resonator having high stability, or frequency correction is used.

The transceiver uses a 16-byte FIFO to store data bytes transmitted or received. Data bytes are transmitted and received from the FIFO buffer over an external standard 3-wire serial interface SPI.

Narrowband, as well as low transmitter recovery time when switching between channels (~150 µs), make it possible to use the XE1205F transceiver to build radio systems using the frequency hopping method (FHSS). The frequency hopping method implies that the entire bandwidth allocated for transmission is divided into a certain number of frequency channels. Jumps from channel to channel occur synchronously in some sequence (for example, linear or pseudo-random).

The XE1205F also benefits from a class-leading -121 dBm receiver sensitivity.

As for data transfer rates, the capabilities of the XE1203F module when using the Barker codec look insufficient even for security systems - only 1.154 kbps. This indicator will not allow implementing an energy-efficient WSN, because the sleep time provided by the CSMA/CA protocol will be too short.

The transceivers of the nodes of the wireless sensor network for the protection of objects should provide the ability to:

creating a mesh network with an increased range;

physical layer implementations of FHSS spectrum spreading technologies;

implementation at the level of access to the environment - "slot" CSMA / CA with access synchronization.

Based on the foregoing, we can conclude that it is preferable to use the XE1205F transceiver module to organize the physical and MAC levels of a wireless sensor network for the protection of objects.

LITERATURE

1. Varaguzin V. Radio networks for collecting data from sensors, monitoring and control based on the IEEE 802.15.4 standard // TeleMultiMedia. - 2005.-№6.- С23-27. - www.telemultimedia.ru

2. Vishnevsky V.M., Lyakhov A.I., Portnoy S.L., Shakhnovich I.V. Broadband wireless networks for information transmission. - M.: Technosphere, 2005 - 592 p.

3. Baskakov S., Oganov V. Wireless sensor networks based on the MeshLogic™ platform // Electronic

Components. - 2006. - No. 8. - P.65-69.

4. Goryunov G. Integrated microwave transceiver XE1203. // World of electronic components. - 2004. - No. 1. -

Wireless sensor networks: an overview

Akuldiz I.F.

Translation from English: Levzhinsky A.S.

annotation

The article describes the concepts of sensor networks, the implementation of which became possible as a result of the combination of microelectromechanical systems, wireless communication and digital electronics. The tasks and potential of sensor networks are studied, a review of the facts influencing their development is made. The architecture of building sensor networks, the developed algorithms and protocols for each layer of the architecture are also considered. The article explores questions about the implementation of sensor networks.

1. Introduction

Recent advances in micro-electro-mechanical systems (MEMS) technologies, wireless communications and digital electronics have made it possible to create low-cost, low-power, multifunctional motes (nodes), they are small and "talk" directly to each other. Sensor networks based on the joint work of a large number of tiny nodes, which consist of modules for collecting and processing data, a transmitter. Such a network has significant advantages over a set of traditional sensors. Here are two key features of traditional sensors: Sensors can be located far from the observed phenomenon. This approach requires many sensors that use some sophisticated techniques to pick out targets from the noise.
You can deploy multiple sensors that only collect data. Carefully design sensor positions and topology. They will transmit observations to the central nodes, where data collection and processing will be performed.
The sensor network consists of a large number of nodes (motes), which are densely located close to the observed phenomenon. The position of the motes does not need to be pre-calculated. This allows them to be randomly placed in hard-to-reach areas or used for relief operations that require a quick response. On the other hand, this means that network protocols and mot algorithms must be self-organizing. Another unique feature of sensor networks is the collaboration of individual nodes. Motes are equipped with a processor. Therefore, instead of passing on the original data, they can process it by performing simple calculations and pass on only the necessary and partially processed data. The features described above provide a wide range of applications for sensor networks. Such networks can be used in healthcare, military and security. For example, physiological data about a patient can be monitored remotely by a doctor. This is convenient both for the patient and allows the doctor to understand his current condition. Sensor networks can be used to detect foreign chemical agents in air and water. They can help determine the type, concentration, and location of contaminants. In essence, sensor networks allow for a better understanding of the environment. We anticipate that in the future, wireless sensor networks will be an integral part of our lives, more so than today's personal computers. The implementation of these and other projects that require the use of wireless sensor networks require special methods. Many protocols and algorithms have been developed for traditional wireless ad hoc networks, so they are not well suited to the unique features and requirements of sensor networks. Here are the differences between sensor and peer-to-peer networks: The number of nodes in a sensor network can be several orders of magnitude higher than nodes in a peer-to-peer network.
The nodes are densely spaced.
Nodes are prone to failure.
The topology of sensor networks can change frequently
Nodes mainly use broadcast messages, while most peer-to-peer networks are based on point-to-point communications.
Nodes are limited in power, processing power, and memory.
Nodes cannot have a global identification number (ID) due to the large amount of overhead and the large number of sensors.
Since the nodes in the network are densely packed, neighboring nodes can be very close to each other. Therefore, multi-hop connections in sensor networks will consume less power than direct connections. In addition, a low data signal power can be used, which is useful in covert surveillance. Multi-hop communications can effectively overcome some of the difficulties of signal propagation over long distances in wireless communications. One of the most important limitations for nodes is low power consumption. Motes have limited energy sources. So, while traditional networks aim to achieve High Quality signal, motes network protocols should focus mainly on energy conservation. They must have mechanisms that allow the user to extend the lifetime of the mote by either reducing the throughput or increasing the data transfer latency. Many researchers are currently involved in the development of circuits that fulfill these requirements. In this article, we will review the protocols and algorithms that currently exist for sensor networks. Our goal is to provide a better understanding of current research issues in this area. We will also try to explore the design constraints and identify tools that can be used to solve design problems. The article is organized like this: in the second section, we describe the potential and usefulness of sensor networks. In Section 3, we discuss the factors that influence the design of such networks. A detailed study of existing methods in this area will be considered in Section 4. And we will summarize in Section 5.

2. Application of wireless sensor networks

Sensor networks can be composed of different types of sensors, such as seismic, magnetic field, thermal, infrared, acoustic, capable of performing a wide variety of measurements of environmental conditions. For example, such as:
temperature,
humidity,
car traffic,
lightning state,
pressure,
soil composition,
noise level,
the presence or absence of certain objects,
mechanical load
dynamic characteristics such as the speed, direction and size of the object.
Motes can be used for continuous probing, event detection and identification. The micro-sensing concept and wireless connectivity promise many new applications for such networks. We have categorized them according to the main areas: military applications, environmental research, healthcare, use in homes and other commercial areas. But it is possible to expand this classification and add more categories, such as space exploration, chemical processing, and disaster relief.

2.1. Military application

Wireless sensor networks can be an integral part of military command, communications, intelligence, surveillance and orientation systems (C4ISRT). Rapid deployment, self-organization, and fault tolerance are characteristics of sensor networks that make them a promising tool for solving problems. Since sensor networks can be based on a dense deployment of disposable and cheap nodes, destroying some of them during hostilities will not affect the military operation in the same way as destroying traditional sensors. Therefore, the use of sensor networks is better suited for battles. We list some more ways to use such networks: monitoring of weapons and ammunition of friendly forces, observation of the battle; orientation on the ground; battle damage assessment; detection of nuclear, biological and chemical attacks. Monitoring of friendly forces, weapons and ammunition: leaders and commanders can constantly monitor the status of their troops, the condition and availability of equipment and ammunition on the battlefield using sensor networks. Each vehicle, equipment and important munitions can have sensors attached to report their status. This data is collected together at key nodes and sent to the leaders. Data can also be forwarded to upper levels command hierarchies to combine with data from other parts. Combat Observations: Critical areas, paths, routes and straits can be quickly covered with sensor networks to study the activities of enemy forces. During operations or after new plans have been developed, sensor networks can be deployed at any time to monitor combat. Enemy Force and Terrain Reconnaissance: Sensor networks can be deployed in critical areas and valuable, detailed and timely data on enemy forces and terrain can be collected within minutes before the enemy can intercept it. Orientation: sensor networks can be used in smart munitions guidance systems. Post-Combat Damage Assessment: Just before or after an attack, sensor networks can be deployed to the target area to collect damage assessment data. Detection of nuclear, biological and chemical attacks: when using chemical or biological weapons, the use of which is close to zero, importance have a timely and accurate determination of chemical agents. Sensor networks can be used as warning systems for chemical or biological attacks, and the data collected in a short time will help to dramatically reduce the number of victims. It is also possible to use sensor networks for detailed reconnaissance after such attacks are detected. For example, it is possible to carry out reconnaissance in the event of radiation contamination without exposing people to radiation.

2.2. Environmental application

Some of the areas in ecology where sensor networks are used are: tracking the movement of birds, small animals and insects; monitoring the state of the environment in order to identify its impact on crops and livestock; irrigation; large-scale earth monitoring and planetary exploration; chemical / biological detection; detection of forest fires; meteorological or geophysical research; flood detection; and pollution research. Wildfire Detection: Because motes can be strategically and tightly deployed in a forest, they can relay the exact origin of a fire before the fire gets out of control. Millions of sensors can be deployed on a continuous basis. They can be equipped with solar panels, as the nodes can be left unattended for months or even years. Motes will work together to perform distributed sensing tasks and overcome obstacles such as trees and rocks that block wired sensors. Mapping the bio-state of the environment: Requires complex approaches to integrate information across time and space scales. Advances in remote sensing technology and automated data collection have greatly reduced research costs. The advantage of these networks is that the nodes can be connected to the Internet, which allows remote users to control, monitor and observe the environment. Although satellite and airborne sensors are useful in observing the great diversity, such as the spatial complexity of dominant plant species, they do not allow observation of the small elements that make up the majority of an ecosystem. As a result, there is a need to deploy wireless sensor network nodes in the field. One example of an application is the biological mapping of the environment in a reserve in Southern California. Three sites are covered by a network, each of which has 25-100 nodes, which are used for continuous monitoring of the state of the environment. Flood detection: An example of flood detection is the public address system in the United States. Several types of sensors placed in the warning system determine the level of precipitation, water level and weather. Research projects such as the COUGAR Device Database Project at Cornell University and the DataSpace Project at Rutgers University are exploring different approaches to interacting with individual nodes on a network to obtain snapshots and long-term data. Agriculture: The advantage of sensor networks is also the ability to monitor pesticide levels in water, soil erosion levels and air pollution levels in real time.

2.3. Application in medicine

One application in medicine is in devices for the disabled; patient monitoring; diagnostics; monitoring the use of medicines in hospitals; collection of human physiological data; and monitoring doctors and patients in hospitals. Monitoring of human physiological state: physiological data collected by sensor networks can be stored for a long period of time and can be used for medical research. Installed network nodes can also track the movements of the elderly and, for example, prevent falls. These nodes are small and provide the patient with greater freedom of movement, while at the same time allowing doctors to identify the symptoms of the disease in advance. In addition, they contribute to a more comfortable life for patients compared to hospital treatment. To test the feasibility of such a system, the Grenoble-France Faculty of Medicine created the "Healthy Smart Home". . Monitoring doctors and patients in the hospital: each patient has a small and light network node. Each node has its own specific task. For example, one might monitor your heart rate while another takes readings of your blood pressure. Doctors may also have such a node, it will allow other doctors to find them in the hospital. Monitoring of medicines in hospitals: Nodes can be attached to medicines, then the chances of dispensing the wrong medicine can be minimized. So, patients will have nodes that determine their allergies and the necessary medications. Computerized systems as described in have shown that they can help minimize the side effects of erroneous dispensing of drugs.

2.4. Application at home

Home automation: smart nodes can be integrated into Appliances such as vacuum cleaners, microwave ovens, refrigerators and VCRs. They can communicate with each other and with an external network via the Internet or satellite. This will allow end users to easily manage devices at home both locally and remotely. Smart environment: Smart environment design can take two different approaches, i.e. human-centric or technology-centric. In the case of the first approach, the smart environment must adapt to the needs of end users in terms of interaction with them. For technology-centered systems, new hardware technologies, networking solutions, and intermediate applications must be developed. Examples of how nodes can be used to create a smart environment are described in . The nodes can be built into furniture and appliances, they can communicate with each other and the room server. The room server can also communicate with other room servers to learn about the services they can offer, such as printing, scanning, and faxing. These servers and sensor nodes can be integrated into existing embedded devices and constitute self-organizing, self-regulating and adaptive systems based on the control theory model as described in .

3. Factors influencing the development of sensor network models.

The development of sensor networks depends on many factors, which include fault tolerance, scalability, production costs, type of operating environment, sensor network topology, hardware limitations, information transmission model, and power consumption. These factors are considered by many researchers. However, none of these studies fully account for all the factors that influence network design. They are important because they serve as a guideline for the development of a protocol or algorithms for the operation of sensor networks. In addition, these factors can be used to compare different models.

3.1. fault tolerance

Some nodes may fail due to lack of power, physical damage, or third party interference. Node failure should not affect the operation of the sensor network. This is a matter of reliability and fault tolerance. Fault tolerance - the ability to maintain the functionality of the sensor network without failure when a node fails. Reliability Rk(t) or node fault tolerance is modeled in using a Poisson distribution to determine the probability of no node failure in the time period (0; t) It is worth noting that protocols and algorithms can be oriented to the level of fault tolerance required to build sensor networks . If the environment in which the nodes are placed is less prone to interference, then the protocols may be less fault-tolerant. For example, if nodes are introduced into a home to monitor humidity and temperature levels, the requirements for fault tolerance may be low, since such sensor networks cannot fail and the “noise” of the environment does not affect their operation. On the other hand, if the nodes are used on the battlefield for surveillance, then the fault tolerance should be high, since surveillance is critical and the nodes can be destroyed during hostilities. As a result, the level of fault tolerance depends on the application of sensor networks and models must be developed with this in mind.

3.2. Scalability

The number of nodes deployed to study the phenomenon can be in the order of hundreds or thousands. Depending on the application, the number can reach extreme values ​​(millions). New models should be able to handle this number of nodes. They also need to use a high density of sensor networks, which can range from a few nodes to several hundred in an area that can be less than 10m in diameter. Density can be calculated according to ,

3.3. Production costs

Since sensor networks consist of a large number of nodes, the cost per node must be such as to justify the total cost of the network. If the cost of the network is higher than the deployment of traditional sensors, then it is not economically viable. As a result, the cost of each node must be low. Now the cost of a node using a Bluetooth transmitter is less than $10. The price for PicoNode is around $1. Therefore, the cost of a sensor network node should be much less than $1 for the economic justification of their use. The cost of a Bluetooth node, which is considered a cheap device, is 10 times higher than the average price of sensor network nodes. Please note that the node also has some additional modules such as a data acquisition module and a data processing module (described in section 3.4.) In addition, they can be equipped with a positioning system or a power generator, depending on the application of sensor networks. As a result, the cost of a node is a complex issue, given the amount of functionality, even at less than $1.00.

3.4. Hardware features

A sensor network node consists of four main components, as shown in Fig. 1: data acquisition unit, processing unit, transmitter and power supply. The presence of additional modules depends on the network application, for example, there may be location modules, a power generator and a mobilizer (MAC). The data acquisition module usually consists of two parts: sensors and analog-to-digital converters (ADCs). analog signal, generated by the sensor based on the observed phenomenon, is converted to digital signal using the ADC, and then fed into the processing unit. The processing module, which uses the integrated memory, manages the procedures that allow, in conjunction with other nodes, to perform the assigned monitoring tasks. The transmitter unit (transceiver) connects the node to the network. One of the most important components of the node is the power supply. The power supply may be rechargeable, for example using solar panels.

Most nodes transmitting data and collecting data need to know their location with high accuracy. Therefore, in general scheme location module enabled. Sometimes you may need a mobilizer that, if necessary, moves the node when it is necessary to complete the tasks. All of these modules may need to be housed in a matchbox-sized enclosure. The knot size can be less than a cubic centimeter and light enough to stay in the air. Apart from size, there are some other hard limits for nodes. They must :
consume very little energy
work with large quantity knots at short distances
have a low production cost
be autonomous and work without supervision,
adapt to the environment.
Since nodes can become unavailable, the life of the sensor network depends on the power of individual nodes. Food is a limited resource and due to size restrictions. For example, the total energy storage of a smart node is on the order of 1 J. For Wireless Integrated Sensor Network (WINS), the average charge level should be less than 30 LA to ensure long runtime. It is possible to extend the life of sensor networks by using rechargeable batteries, for example, by obtaining energy from the environment. Solar panels are a prime example of the use of recharging. The node communication module can be a passive or active optical device, as in a smart node, or a radio frequency (RF) transmitter. RF transmission needs a modulation module that uses a certain bandwidth, a filtering module, a demodulation module, which makes them more complex and expensive. In addition, there may be loss in data transmission between two nodes due to the fact that the antennas are located close to the ground. However, radio communication is preferred in most existing sensor network designs because data rates are low (typically less than 1 Hz) and transmission cycle rates are high due to short distances. These characteristics allow the use of low radio frequencies. However, designing energy-efficient and low-frequency radio transmitters is still a technical challenge, and the existing technologies that are used in the manufacture of Bluetooth devices are not efficient enough for sensor networks, as they consume a lot of energy. Although processors are constantly shrinking in size and increasing in power, the processing and storage of data by the node is still its weak point. For example, the smart node processing module consists of a 4 MHz Atmel AVR8535 processor, a microcontroller with 8 KB for instructions, flash memory, 512 bytes of RAM, and 512 bytes of EEPROM. In this module, which has 3500 bytes under OS and 4500 bytes free memory under the code, the TinyOS operating system is used. The processing module of another lAMPS node prototype has a 59-206 MHz SA-1110 processor. IAMPS nodes use a multithreaded operating system. L-OS system. Most data collection tasks require knowledge of the node's position. Since the nodes are usually located randomly and without supervision, they must cooperate using a positioning system. Location determination is used in many sensor network routing protocols (more details in Section 4). Some have suggested that each node should have a Global Positioning System (GPS) module that works to within 5 meters. The paper argues that equipping all nodes with GPS is not necessary for the operation of sensor networks. There is an alternative approach where only some nodes use GPS and help other nodes to determine their position on the ground.

3.5. Network topology

The fact that nodes can become unavailable and subject to frequent failures makes network maintenance a challenging task. From hundreds to several thousand nodes can be placed on the territory of the sensor network. They deploy ten meters apart. The density of knots can be higher than 20 knots per cubic meter. The dense arrangement of many nodes requires careful maintenance of the network. We will cover issues related to maintaining and changing the topology of the network in three stages:

3.5.1. The pre-deployment and deployment of the nodes itself can consist in the mass dispersion of the nodes or the installation of each separately. They can be deployed:

Scattered from an airplane,
by being placed in a rocket or projectile
thrown by means of a catapult (for example, from a ship, etc.),
placement in the factory
each node is placed individually by a human or a robot.
While the sheer number of sensors and their automated deployment usually precludes placing them according to a carefully designed plan, initial deployment schemes should:
reduce installation costs
eliminate the need for any prior organization and advance planning,
increase placement flexibility,
promote self-organization and fault tolerance.

3.5.2. Phase after network deployment

After the network is deployed, the change in its topology is associated with a change in the characteristics of the nodes. Let's list them:
position,
accessibility (due to interference, noise, moving obstacles, etc.),
battery charge,
malfunctions
changing tasks.
Nodes can be statically deployed. However, device failure is common due to battery drain or destruction. Sensor networks with high node mobility are possible. In addition, nodes and networks perform various tasks and may be subject to deliberate interference. Thus, the structure of the sensor network is prone to frequent changes after deployment.

3.5.3. Additional Node Deployment Phase

Additional nodes can be added at any time to replace faulty nodes or due to changing tasks. Adding new nodes creates the need to reorganize the network. Dealing with frequent changes in the topology of a peer-to-peer network that contains many nodes and has very tight power limits requires special routing protocols. This issue is discussed in more detail in Section 4.

3.6. Environment

The nodes are densely located very close to or directly within the observed phenomenon. Thus, they operate unsupervised in remote geographic areas. They can work
at busy intersections
inside big cars
at the bottom of the ocean
inside a tornado
on the surface of the ocean during a tornado,
in biologically and chemically contaminated areas
in the battlefield
in a house or a large building,
in a large warehouse
attached to animals
attached to fast moving vehicles
in a sewer or river along with the flow of water.
This list gives an idea of ​​the conditions under which nodes can operate. They can operate under high pressure on the ocean floor, in harsh environments, among debris or in the battlefield, in extreme temperatures, such as in the nozzle of an aircraft engine or in arctic regions, in very noisy places where there is a lot of interference.

3.7. Data transfer methods

In a multi-hop sensor network, nodes communicate wirelessly. Communication can be via radio, infrared or optical media. In order to use these methods globally, the transmission medium must be available worldwide. One option for radio communications is to use the Industrial, Scientific and Medical (ISM) bands, which are available without a license in most countries. Some of the frequencies that can be used are described in the international frequency table contained in Article S5 on the Radio Regulations (Volume 1). Some of these frequencies are already in use in wireless telephony and wireless local networks(WLAN). For sensor networks of small size and low cost, a signal amplifier is not required. According to , hardware limitations and trade-offs between antenna efficiency and power consumption impose certain restrictions on the choice of transmission frequency in the microwave frequency range. They also offer 433 MHz ISM in Europe and 915 MHz ISM in North America. Possible models transmitters for these two zones are covered in . The main advantages of using ISM radio frequencies are the wide spectrum of frequencies and worldwide availability. They are not tied to a specific standard, thus giving more freedom to implement energy-saving strategies in sensor networks. On the other hand, there are various rules and restrictions, such as various laws and interference from existing applications. These frequency bands are also called unregulated frequencies. Most of today's node equipment is based on the use of radio transmitters. The wireless nodes of IAMPS, described in , use Bluetooth-enabled 2.4 GHz transmitters and have an integrated frequency synthesizer. The device of low-power nodes is described in the work, they use one radio transmission channel, which operates at a frequency of 916 MHz. The WINS architecture also uses radio. Another possible way communication in sensor networks is infrared. IR communication is available without a license and is immune to electrical interference. IR transmitters are cheaper and easier to manufacture. Many of today's laptops, PDAs and mobile phones use the IR interface for data transmission. The main disadvantage of such communication is the requirement of direct visibility between the sender and the recipient. This makes IR communications undesirable for use in sensor networks due to the transmission medium. An interesting transmission method is using smart nodes, which are modules for automatic monitoring and data processing. They use an optical medium for transmission. There are two transmission schemes, passive using a corner-cube retroreflector (CCR) and active using a laser diode and controlled mirrors (discussed in ). In the first case, an integrated light source is not required, a three-mirror (CCR) configuration is used for signal transmission. The active method uses a laser diode and an active laser communication system to send beams of light to the intended receiver. The unusual application requirements of sensor networks make the choice of transmission medium difficult. For example, marine applications require the use of an aquatic transmission medium. Here you need to use long-wave radiation, which can penetrate the surface of the water. In difficult terrain or on the battlefield, errors and more interference may occur. In addition, it may turn out that the node antennas do not have the necessary height and radiation power for communication with other devices. Therefore, the choice of transmission medium must be accompanied by reliable modulation and coding schemes, which depend on the characteristics of the transmission channel.

3.8. Power consumption

The wireless node, being a microelectronic device, can only be equipped with a limited power supply (

3.8.1. Connection

A node spends its maximum energy on communication, which involves both transmitting and receiving data. It can be said that in order to communicate short distances with low transmitting power, transmission and reception require approximately the same amount of energy. Frequency synthesizers, voltage control oscillators, phase blocking (PLL) and power amplifiers all require energy, which is limited. It is important that in this case we do not consider only active power, but also the consumption of electricity when starting transmitters. Starting up the transmitter takes a fraction of a second, so it consumes negligible amounts of power. This value can be compared to the PLL lock time. However, as the transmitted packet decreases, the launch power begins to dominate the power consumption. As a result, it is inefficient to constantly turn the transmitter on and off, because most of the energy will be spent on this. Currently, low power radio transmitters have standard Pt and Pr values ​​of 20 dBm and Pout close to 0 dBm. Note that PicoRadio directed to Pc is -20dBm. The design of small-sized, inexpensive, transmitters is discussed in the source. Based on their results, the authors of this article, given the budget and energy estimates, believe that the values ​​of Pt and Pr should be at least an order of magnitude smaller than the values ​​given above.

3.8.2. Data processing

The power consumption of data processing is much less compared to data transmission. The example described in the paper actually illustrates this discrepancy. Based on Rayleigh's theory that a quarter of the power is lost during transmission, we can conclude that the energy consumption for transmitting 1 KB over a distance of 100 m will be about the same as executing 3 million instructions at a rate of 100 million instructions per second (MIPS )/W by the processor. Therefore, local data processing is critical to minimizing power consumption in a multi-hop sensor network. Therefore, nodes must have built-in computing capabilities and be able to interact with the environment. Cost and size constraints will lead us to choose semiconductors (CMOS) as the main technology for microprocessors. Unfortunately, they have limits on energy efficiency. CMOS requires power every time it changes state. Energy required to change states, proportional to switching frequency, capacitance (depending on area) and voltage fluctuations. Therefore, reducing the supply voltage is an effective means of reducing power consumption in the active state. Dynamic voltage scaling, discussed in , seeks to adapt the power and frequency of the processor according to the workload. When the processing load on the microprocessor is reduced, simply reducing the frequency gives a linear reduction in power consumption, however, reducing the operating voltage gives us a quadratic reduction in power costs. On the other hand, all possible processor performance will not be used. This will give the result if we take into account that peak performance is not always required and therefore, the operating voltage and frequency of the processor can be dynamically adapted to processing requirements. The authors propose workload prediction schemes based on the adaptive processing of existing load profiles and on the analysis of several already created schemes. Other strategies for reducing processor power are discussed in . It should be noted that additional schemes for encoding and decoding data may be used. integrated circuits may also be used in some cases. In all these scenarios, the structure of the sensor network, operation algorithms and protocols depend on the respective energy costs.

4. Architecture of sensor networks

The nodes are usually located randomly throughout the observation area. Each of them can collect data and knows the route of data transfer back to the central node, the end user. Data is transmitted using a multi-hop network architecture. The central node can communicate with the task manager via the Internet or satellite. The protocol stack used by the central node and all other nodes is shown in Fig. 3. The protocol stack includes power information and route information, contains network protocol information, helps to communicate effectively over the wireless environment, and promotes node collaboration. The protocol stack consists of an application layer, a transport layer, a network layer, a data link layer, a physical layer, a power management layer, a mobility management layer, and a task scheduling layer. Depending on the tasks of data collection, various types of application software can be built at the application level. the transport layer helps keep the data flowing if required. The network layer handles the routing of data provided by the transport layer. Since the environment has extraneous noise and nodes can be moved, the MAC protocol must minimize the occurrence of collisions when transmitting data between neighboring nodes. The physical layer is responsible for the ability to transfer information. These protocols help hosts perform tasks while saving power. The power management layer determines how a node should use power. For example, a node may turn off a receiver after receiving a message from one of its neighbors. This will help you avoid getting a duplicate message. Also, when a node is low on battery, it communicates to its neighbors that it cannot participate in message routing. It will use all the remaining energy to collect data. The Mobility Control (MAC) layer detects and registers the movement of nodes, so there is always a route for data transfer to the central node and nodes can determine their neighbors. And knowing its neighbors, the node can balance power consumption by working together with them. The task manager plans and schedules the collection of information for each region separately. Not all nodes in the same region are required to run probing tasks at the same time. As a result, some nodes perform more tasks than others, depending on their capacity. All these layers and modules are necessary for the nodes to work together and strive for maximum energy efficiency, optimize the data transmission route in the network, and also share each other's resources. Without them, each node will work individually. From the point of view of the entire sensor network, it is more efficient if the nodes work together with each other, which helps to extend the lifetime of the networks themselves. Before discussing the need to include modules and control layers in the protocol, we will consider three existing work, dedicated to the protocol stack, which is shown in Figure 3. The WINS model discussed in the source, in which nodes are combined into a distributed network and have access to the Internet. Since a large number of WINS network nodes are located at a small distance from each other, multi-hop communications reduce power consumption to a minimum. The environmental information received by the node is sequentially sent to the central node or WINS gateway through other nodes, as shown in Figure 2 for nodes A, B, C, D, and E. The WINS gateway communicates with the user through common network protocols such as the Internet . The WINS network protocol stack consists of the application layer, the network layer, the MAC layer, and the physical layer. Smart nodes (or specks of dust). These nodes can be attached to objects or even float in the air due to their small size and weight. They use MEMS technology for optical communication and data collection. Dust motes may have solar panels to recharge during the day. They require a line of sight to communicate with an optical base station transmitter or other speck of dust. Comparing the architecture of the dust network with that shown in Figure 2, it can be said that smart nodes, as a rule, communicate directly with the transmitter base station, but a one-to-one connection is also possible. Another approach to the development of protocols and algorithms for sensor networks is due to the requirements of the physical layer. Protocols and algorithms must be designed according to the choice of physical components such as the type of microprocessors and the type of receivers. This bottom-up approach is used in the IAMPS model and also considers the dependence of the application layer, network layer, MAC layer, and physical layer on the host hardware. IAMPS nodes interact with the end user in exactly the same way as in the architecture shown in Figure 2. Various schemes, such as time division division (TDMA) or frequency division channels (FDMA) and binary modulation or M-modulation are compared in the source. The bottom-up approach means that the node's algorithms must know the hardware and use the capabilities of microprocessors and transmitters to minimize power consumption. This may lead to the development of various node designs. And different node designs will lead to different types sensor networks. Which in turn will lead to the development of various algorithms for their work.

Literature

1. G.D. Abowd, J.P.G. Sterbenz, Final report on the interagency workshop on research issues for smart environments, IEEE Personal Communications (October 2000) 36–40.
2. J. Agre, L. Clare, An integrated architecture for cooperative sensing networks, IEEE Computer Magazine (May 2000) 106–108.
3. I.F. Akyildiz, W. Su, A power aware enhanced routing (PAER) protocol for sensor networks, Georgia Tech Technical Report, January 2002, submitted for publication.
4. A. Bakre, B.R. Badrinath, I-TCP: indirect TCP for mobile hosts, Proceedings of the 15th International Conference on Distributed Computing Systems, Vancouver, BC, May 1995, pp. 136–143.
5. P. Bauer, M. Sichitiu, R. Istepanian, K. Premaratne, The mobile patient: wireless distributed sensor networks for patient monitoring and care, Proceedings 2000 IEEE EMBS International Conference on Information Technology Applications in Biomedicine, 2000, pp. 17–21.
6. M. Bhardwaj, T. Garnett, A.P. Chandrakasan, Upper bounds on the lifetime of sensor networks, IEEE International Conference on Communications ICC’01, Helsinki, Finland, June 2001.
7. P. Bonnet, J. Gehrke, P. Seshadri, Querying the physical world, IEEE Personal Communications (October 2000) 10–15.

The day is near when hundreds of millions of semiconductor sensors will be integrated into everything that is possible, from a key fob to a baby stroller. And all of them will be able not only to act as intelligent sensors, but also to perform the primary processing of information, as well as interact with each other, forming a single wireless sensor network. At the same time, such sensors will practically not consume electricity, since the built-in miniature batteries will last for several years, that is, for the entire life of the sensors. It will be a conceptually new type of computer system that operates using a wireless sensor network. Such a network is called Ad-hoc Wireless Sensor Networks. The term Ad-hoc is borrowed from modern wireless networks, such as the IEEE 802.11b standard. Such wireless networks have two interaction modes: Infrastructure mode and Ad-hoc. In Infrastructure mode, network nodes do not interact directly with each other, but through an Access Point, which performs wireless network the role of a kind of hub (similar to how it happens in traditional cable networks). In Ad-hoc mode, also called Peer-to-Peer, stations communicate directly with each other. Accordingly, in wireless sensor networks, the Ad-hoc mode means that all sensors directly interact with each other, creating a kind of cellular network.

Wireless sensor networks are a kind of step towards the next era - when computers will be directly connected to the physical world and will be able to guess the desires of users, as well as make decisions for them.
Let's dream a little about what such sensor networks will bring us in the future. Imagine cribs listening to babies breathe; bracelets that monitor the condition of patients in the clinic; smoke detectors, which can not only call firefighters if necessary, but also inform them in advance about the source of fire and the degree of complexity of the fire. Electronic devices will be able to recognize each other, the power sources will remind them that they need to “feed”.

Imagine hundreds of thousands of touch sensors integrated into common network in the forest. In such a forest, it will simply be impossible to get lost, since the movement of a person will be recorded and analyzed by sensors. Another example is sensors in the field, tuned to monitor the state of the soil and, depending on changing conditions, regulate watering and the amount of fertilizer applied.
No less useful will be sensor networks on the roads. Communicating with each other, they will be able to regulate the flow of cars. This is the dream of any driver - roads without traffic jams! Such networks will be able to cope with this task much more effectively than any agency. Control problem
offenses on the roads will be solved by itself.

The use of sensor networks for power management will achieve incredible energy savings. Imagine such a control network in your apartment. Tracking your location, sensors will be able to turn off the lights behind you everywhere and turn them on as needed. Well, if you use such networks to control the lighting of streets and roads, then the problem of lack of electricity will disappear by itself. In order for sensor networks to become a reality of tomorrow, research in this direction is already underway. And the leader in this field is Intel Corporation, which supports all the advanced computer technologies of the future. Particular attention is paid to the development of wireless multi-node sensor networks capable of self-automatic formation and configuration as needed. The implementation of this technology will make it possible to deploy a network of inexpensive, but at the same time very complex semiconductor sensor devices that can independently establish communication with each other, reporting on certain changes in the environment. For example, the Mica sensor is equipped with 128 kilobytes of flash memory software, 256 kilobytes of flash memory for data storage, and a radio transmitter operating at a frequency of 900 MHz.
Some of these devices run an operating system
TinyOS, the code for this operating system is open source and consists of just
8.5 Kb.

Such devices will find applications in revolutionary new areas, such as the development of smart garments, connected blankets that will monitor the health of a newborn and report vital signs of his life, smart farms, in which semiconductor sensors installed in the soil will control irrigation
system and fertilization. Sensor network research at Intel is
famous research laboratory Intel Berkeley Research laboratory, located in California. The experimental sensor networks that exist today only partially satisfy the above requirements. So, to date, networks consist of only hundreds of sensors with a limited coverage area and perform only well-defined tasks. They are able to transmit only a certain type of information from one sensor to another and only in a given bandwidth. Energy consumption is also not negligible
- The battery only lasts a few days. Existing touch sensors are still quite inert, and high reliability and stealth in operation (at least because of their size) are out of the question. And, of course, such sensors are quite expensive, so that a network consisting of hundreds of sensors is not cheap. But we must remember that we are talking about experimental networks and the development of future technology. At the same time, experimental sensor networks are already making a difference. One such sensor network, developed jointly by the Intel Berkeley Research Laboratory, the Institute of the Atlantic and the University of California, operates on Great Duck Island in Maine.

The task of this network is to study the microhabitat of various biological organisms inhabiting the island.
Any human intervention (even for the purpose of learning) is sometimes unnecessary,
this is where sensor networks come to the rescue, allowing you to collect all the necessary information without the direct participation of a person.

The sensor network uses two boards as node elements. The first board contains a temperature sensor, humidity and barometric pressure sensors, and an infrared sensor. The second board contains a microprocessor (frequency 4 MHz), RAM 1KB, flash memory for storing programs and data, power supply (two AA batteries) and radio transmitter/
receiver operating at 900 MHz. Sensors allow you to register all the necessary information and transfer it to the host computer database. All sensors are thoroughly tested beforehand - the board with sensors is immersed in water for two days and its functionality is monitored. All sensor nodes form a single wireless network and are able to exchange information. In this case, the transfer of information from a remote network node to the gateway (Gateway Sensor) occurs in a chain, that is, from one network node to another, which allows you to create a large coverage area.

Through the gateway, the information reaches the main computer. The gateway uses a directional antenna, which allows you to increase the transmission distance up to 300 m. From the host computer, information using satellite communications transmitted over the Internet to a research center located in California.

The laboratory staff is no less actively working on precision biology and the creation of biochips. In addition to sensory perception of the world of solid things, the possibility of "feeling" liquid media and biological, developing objects is being explored. Such research opens up tremendous prospects for medical and pharmaceutical development, the implementation of chemical processes and the manufacture of biological drugs. Since the main purpose of sensor networks is the perception and transmission useful information, experts at the Intel Berkeley Lab are developing a methodology for combining sensors with objects they are responsible for monitoring, and are also exploring the possibility of creating "actuators" - devices based on sensors that allow you to influence the situation, and not just register its state. Sensor networks are obviously useful for military applications, one possible variation of the network was "combat" tested in Afghanistan, where the US military deployed several hundred sensors to track the movements of enemy military equipment. However, about the implementation
real networks in our lives too early to say, the network is vulnerable to fault tolerance. A denial of service (DoS) attack on a sensor network is any event that reduces or eliminates the network's ability to perform its expected function. The authors propose to base sensor network protocols on a layered architecture, which can hurt the efficiency of the network, but increase its reliability. The types of DoS attacks typical for each level and acceptable methods of protection are discussed. Thus, even today, despite the imperfection and still rather narrow range of use, sensor networks are used in science, and later in life.

Websites used:

Maxim Sergievsky

The latest wireless communication technologies and progress in the field of microchip manufacturing have made it possible over the past few years to move on to the practical development and implementation of a new class of distributed communication systems - sensor networks.

Wireless sensor networks consist of miniature computing and communication devices - motes ( from English. motes - dust particles), or sensors. A mot is a board that is usually no larger than one cubic inch. The board houses a processor, flash and RAM memory, digital-to-analog and analog-to-digital converters, an RF transceiver, a power supply, and sensors. Sensors can be very diverse; they are connected via digital and analog connectors. More often than others, temperature, pressure, humidity, light, vibration sensors are used, less often - magnetoelectric, chemical (for example, measuring the content of CO, CO2), sound and some others. The set of sensors used depends on the functions performed by wireless sensor networks. The motor is powered by a small battery. Motes are used only for collecting, pre-processing and transmitting sensory data. Appearance motors produced by various manufacturers is shown in fig. one.

The main functional processing of data collected by motes is carried out at the node, or gateway, which is a fairly powerful computer. But in order to process the data, they must first be received. For this purpose, the node must be equipped with an antenna. But in any case, only motes that are close enough to it are available to the node; in other words, the node does not receive information directly from each mote. The problem of obtaining sensory information collected by motes is solved as follows. Motes can exchange information with each other using transceivers operating in the radio range. This is, firstly, sensory information read from sensors, and secondly, information about the status of devices and the results of the data transfer process. Information is transmitted from one mote to another along the chain, and as a result, the motes closest to the gateway dump all the accumulated information to it. If some of the motes fail, the operation of the sensor network after the reconfiguration should continue. But in this case, naturally, the number of sources of information decreases.

To perform functions, a specialized operating system is installed on each mot. Currently, most wireless sensor networks use TinyOS, an operating system developed at the University of Berkeley. TinyOS is open source software; it is available at: www.tinyos.net. TinyOS is an event-driven, real-time operating system designed to work with limited computing resources. This OS allows motes to automatically establish connections with neighbors and form a sensor network of a given topology. The last release of TinyOS 2.0 appeared in 2006.

The most important factor in the operation of wireless sensor networks is the limited capacity of the batteries installed on motos. Please note that batteries are often not replaceable. In this regard, it is necessary to perform only the simplest primary processing on motes, aimed at reducing the amount of transmitted information, and, most importantly, minimizing the number of cycles of receiving and transmitting data. To solve this problem, special communication protocols have been developed, the most famous of which are the protocols of the ZigBee alliance. This alliance (website www.zigbee.org) was created in 2002 specifically to coordinate work in the field of wireless sensor networks. It includes the largest developers of hardware and software tools: Philips, Ember, Samsung, IBM, Motorola, Freescale Semiconductor, Texas Instruments, NEC, LG, OKI and many more (more than 200 members in total). Intel Corporation is not included in the alliance, although it supports its activities.

In principle, to develop a standard, including a protocol stack for wireless sensor networks, ZigBee used the previously developed IEEE 802.15.4 standard, which describes the physical layer and the media access layer for wireless data networks over short distances (up to 75 m) with low power consumption, but with a high degree of reliability. Some characteristics of radio data transmission for the IEEE 802.15.4 standard are given in Table. one.

Table 1. Data radio characteristics for IEEE 802.15.4

Frequency band, MHz	Do I need a license	Geographic region	Data transfer rate, Kbps	Number of channels

On the this moment ZigBee has developed the only standard in this area, which is backed by the availability of production of fully compatible hardware and software products. ZigBee protocols allow devices to sleep b about most of the time, greatly extending battery life.

Obviously, it is not so easy to develop data exchange schemes between hundreds and even thousands of motes. Among other things, it is necessary to take into account the fact that sensor networks operate in unlicensed frequency bands, therefore, in some cases, interference from extraneous radio signal sources may occur. It is also desirable to avoid repeated transmission of the same data, and in addition, take into account that due to insufficient power consumption and external influences motes will fail forever or for a while. In all such cases, communication schemes must be modified. Since one of the most important features of TinyOS is the automatic selection of networking schemes and data paths, wireless sensor networks are essentially self-configuring.

More often than not, a mote should be able to determine its own location, at least in relation to the other mote to which it will transmit data. That is, first all the motes are identified, and then the routing scheme is already formed. In general, all motes - ZigBee standard devices - are divided into three classes according to the level of complexity. The highest of them - the coordinator - manages the operation of the network, stores data about its topology and serves as a gateway for transmitting data collected by the entire wireless sensor network for further processing. Sensor networks usually use one coordinator. The mote of average complexity is a router, that is, it can receive and transmit data, as well as determine the direction of transmission. And finally, the simplest mote can only transmit data to the nearest router. Thus, it turns out that the ZigBee standard supports a network with a cluster architecture (Fig. 2). The cluster is formed by a router and the simplest motes from which it requests sensory data. Cluster routers relay data to each other, and eventually the data is sent to the coordinator. The coordinator usually has a connection to the IP network, where the data is sent for final processing.

In Russia, developments are also being carried out related to the creation of wireless sensor networks. Thus, the High-Tech Systems company offers its MeshLogic hardware and software platform for building wireless sensor networks (website www.meshlogic.ru). The main difference between this platform and ZigBee is its focus on building peer-to-peer mesh networks (Fig. 3). In such networks, the functionality of each mote is the same. The possibility of self-organization and self-healing of networks of cellular topology allows, in the event of failure of some of the motes, to spontaneously form new structure networks. True, in any case, you need a central functional node that receives and processes all data, or a gateway for transmitting data to the node for processing. Spontaneously created networks are often referred to by the Latin term Ad Hoc, which means "for a specific occasion."

In MeshLogic networks, each mote can perform packet relay, that is, in its functions it resembles a ZigBee router. MeshLogic networks are fully self-organizing: no coordinator node is provided. As RF transceivers in MeshLogic can be used various devices, in particular Cypress WirelessUSB, which, like ZigBee standard devices, operate in the 2.4 ... 2.4835 GHz frequency range. It should be noted that only the lower layers of the protocol stack exist for the MeshLogic platform. It is believed that the upper layers, in particular the network and application, will be created for specific applications. The configurations and main parameters of two MeshLogic motors and one ZigBee standard motor are shown in Table. 2.

Table 2. Main characteristics of motors from different manufacturers

Options
microcontroller
CPU	Texas Instruments MSP430
Clock frequency	32.768 kHz to 8 MHz
RAM
Flash memory
Transceiver
		Cypress WirelessUSBTM LP
Frequency range	2400-2483.5 MHz		2400-2483.5 MHz
Transfer rate		15.625 to 250 Kbps
output power	-24 to 0 dBm	-35 to 4 dBm	-28 to 3 dBm
Sensitivity
			1 or 2 chips
External interfaces
	12-bit, 7 channels		10-bit, 3 channels
Digital interfaces	I2C/SPI/UART/USB		I2C/SPI/UART/IRQ/JTAG
Other Options
Supply voltage	0.9 to 6.5 V		1.8 to 3.6 V
Temperature Range	-40 to 85 °C	0 to 70°C	0 to 85°C

Note that there are no integrated touch sensors on these boards.

We indicate what primarily distinguishes wireless sensor networks from conventional computing (wired and wireless) networks:

• the complete absence of any kind of cables - electrical, communication, etc.;
• the possibility of compact placement or even integration of motes into environmental objects;
• reliability of both individual elements and, more importantly, the entire system as a whole; in some cases, the network can function with only 10-20% of sensors (motes) in good working order;
• no need for personnel for installation and maintenance.

Sensor networks can be used in many application areas. Wireless sensor networks are a promising new technology and all related projects are mostly under development. We indicate the main areas of application of this technology:

• defense and security systems;
• environmental control;
• monitoring of industrial equipment;
• security systems;
• monitoring the state of agricultural land;
• energy management;
• control of ventilation, air conditioning and lighting systems;
• fire alarm;
• inventory control;
• tracking the transportation of goods;
• monitoring of the physiological state of a person;
• personnel control.

From a fairly large number of examples of the use of wireless sensor networks, we single out two. Perhaps the most famous is the deployment of the network aboard a BP oil tanker. There, using a network built on the basis of Intel equipment, the state of the vessel was monitored in order to organize its preventive maintenance. BP has analyzed whether the sensor network can operate on board the ship in the extreme temperatures, high vibration and significant levels of radio frequency interference present in certain areas of the ship. The experiment was successful, the network was reconfigured and restored automatically several times.

An example of another completed pilot project is the deployment of a sensor network at a US Air Force base in Florida. The system demonstrated good opportunities for recognition of various metal objects, including moving ones. The use of a sensor network made it possible to detect the penetration of people and cars into the controlled area and track their movements. To solve these problems, motors equipped with magnetoelectric and temperature sensors were used. The scope of the project is currently expanding and the wireless sensor network is already being installed on a 10,000x500 m test site. software developed by several American universities.

History and scope

One of the first prototypes of the sensor network can be considered the SOSUS system, designed to detect and identify submarines. Technologies of wireless sensor networks began to develop actively relatively recently - in the mid-1990s. However, only at the beginning of the 21st century, the development of microelectronics made it possible to produce a fairly cheap element base for such devices. Modern wireless networks are mainly based on the ZigBee standard. A considerable number of industries and market segments (manufacturing, various modes of transport, life support, security) are ready for the implementation of sensor networks, and this number is constantly increasing. The trend is due to the complication of technological processes, the development of production, the expanding needs of individuals in the segments of security, resource control and the use of inventory. With the development of semiconductor technologies, new practical tasks and theoretical problems appear related to the applications of sensor networks in industry, housing and communal services, and households. The use of low-cost wireless sensor control devices opens up new areas for the application of telemetry and control systems, such as:

• Timely detection of possible failures of actuators, to control such parameters as vibration, temperature, pressure, etc.;
• Real-time access control to remote systems of the monitored object;
• Automation of inspection and maintenance of industrial assets;
• Commercial asset management;
• Application as components in energy and resource saving technologies;
• Control of eco-parameters of the environment.

It should be noted that despite the long history of sensor networks, the concept of building a sensor network has not finally taken shape and has not been expressed in certain software and hardware (platform) solutions. The implementation of sensor networks at the current stage largely depends on the specific requirements of the industrial task. The architecture, software and hardware implementation is at the stage of intensive technology formation, which draws the attention of developers in order to search for a technological niche for future manufacturers.

Technology

Wireless sensor networks (WSNs) are made up of miniature computing devices- motors equipped with sensors (sensors for temperature, pressure, light, vibration level, location, etc.) and signal transceivers operating in a given radio range. Flexible architecture, reduced installation costs distinguish wireless smart sensor networks from other wireless and wired data interfaces, especially when it comes to in large numbers connected devices, the sensor network allows you to connect up to 65,000 devices. The constant reduction in the cost of wireless solutions, the increase in their operational parameters make it possible to gradually reorient from wired solutions in systems for collecting telemetry data, remote diagnostics, and information exchange. "Sensory network" is a well-established term today. Sensor Networks), denoting a distributed, self-organizing, fault-tolerant network of individual elements from unattended and requiring no special installation of devices. Each sensor network node can contain various sensors for environmental control, microcomputer and radio transceiver. This allows the device to take measurements, independently carry out initial data processing and maintain communication with an external information system.

802.15.4/ZigBee relayed short-range radio technology, known as "Sensor Networks" (eng. WSN - Wireless Sensor Network), is one of the modern directions in the development of self-organizing fault-tolerant distributed systems for monitoring and managing resources and processes. Today, wireless sensor network technology is the only wireless technology, with which you can solve the tasks of monitoring and control, which are critical to the operation time of sensors. The sensors combined into a wireless sensor network form a territorially distributed self-organizing system for collecting, processing and transmitting information. The main area of ​​application is the control and monitoring of the measured parameters of physical media and objects.

• radio path;
• processor module;
• battery;
• various sensors.

A typical node can be represented by three types of devices:

• Network coordinator (FFD - Fully Function Device);
  • performs global coordination, organization and setting of network parameters;
  • the most complex of the three device types, requiring the most memory and power supply;

• Device with a full set of functions (FFD - Fully Function Device);
  • support for 802.15.4;
  • additional memory and power consumption allows you to act as a network coordinator;
  • support for all types of topologies ("point-to-point", "star", "tree", "mesh network");
  • the ability to act as a network coordinator;
  • the ability to access other devices on the network;
• (RFD - Reduced Function Device);
  • supports a limited set of 802.15.4 features;
  • support for point-to-point, star topologies;
  • does not act as a coordinator;
  • calls the network coordinator and router;

Companies developers

There are different types of companies on the market:

Notes

Wikimedia Foundation. 2010 .

See what "Wireless Sensor Networks" is in other dictionaries:

- (other names: wireless ad hoc networks, wireless dynamic networks) decentralized wireless networks that do not have a permanent structure. Client devices are connected on the fly, forming a network. Each network node tries to forward ... ... Wikipedia

This page is proposed to be renamed to Wireless ad hoc network. Explanation of reasons and discussion on the Wikipedia page: To be renamed / December 1, 2012. Perhaps its current name does not meet the standards of the modern ... ... Wikipedia

Wireless ad hoc networks are decentralized wireless networks that do not have a permanent structure. Client devices are connected on the fly, forming a network. Each node in the network attempts to forward data destined for other nodes. At the same time ... ... Wikipedia

Wireless ad hoc networks are decentralized wireless networks that do not have a permanent structure. Client devices are connected on the fly, forming a network. Each node in the network attempts to forward data destined for other nodes. At the same time ... ... Wikipedia

The architecture of a typical wireless sensor network A wireless sensor network is a distributed, self-organizing network of many sensors (sensors) and actuators, interconnected via a radio channel. Region ... ... Wikipedia

Would you like to improve this article?: Rework the design in accordance with the rules for writing articles. Check the article for grammatical and spelling errors. Correct the article according to ... Wikipedia

Telemetry, telemetry (from other Greek τῆλε “far” + μέτρεω “I measure”) a set of technologies that allows remote measurements and collection of information to be provided to the operator or user, component… … Wikipedia

Ultra wideband (UWB) signals are radio signals (UHF signals) with "extra large" bandwidth. They are used for ultra-wideband radar and ultra-wideband radio communications. Contents 1 Definition 2 Regulation ... Wikipedia

The first Open Wireless Network Protocol designed for building automation and distributed facility management. One Net can be used with many existing transceivers (transceivers) and ... ... Wikipedia