Introduction

Wireless sensor network- distributed, a set of sensors (sensors) and actuators, interconnected by means of a radio channel. The coverage area of ​​such a network can range from several meters to several kilometers due to the ability to relay messages from one element to another.

The main features of wireless sensor networks are self-organization and adaptability to changes in operating conditions, so it requires minimum costs when deploying the network at the facility and during its subsequent maintenance during operation.

Short story

One of the first prototypes of the sensor network can be considered the SOSUS system, designed to detect and identify submarines. In the mid-1990s, wireless sensor network technologies began to actively develop; in the early 2000s, the development of microelectronics made it possible to produce a fairly cheap element base for such devices. Wireless networks of the early 2010s are mainly based on the .

Purpose

The main purpose is not only to exchange data between nodes via a decentralized self-organizing network, but also to collect transmitted information (mainly data) from sensors (temperature, pressure, humidity, radiation levels, acoustic vibrations) to a central node for the purpose of its subsequent analysis or processing.

The demand for wireless sensor networks in the market is also closely related to the concept of intellectualization of such objects as home, office and industrial premises, where an urban person spends up to 90% of his time, as well as the concept of creating cybernetic industries (fully equipped with robots), the primary task of which is to introduce wireless technologies at the APCS level.

Sensor network technology is designed to solve the widest range of industrial monitoring and control tasks and has the following undeniable advantages over other existing wireless and wired systems:

• the ability to install sensors on an existing and operated facility without additional work for laying a wired network;
• low cost a separate control element;
• low cost installation, commissioning and maintenance of the system;
• minimal restrictions on the placement of wireless devices;
• high fault tolerance sensory network as a whole.

Description

The hardware of the wireless nodes and the protocols of network interaction between them are optimized for power consumption to ensure a long service life of the system with offline sources nutrition. Depending on the mode of operation, the lifetime of a node can reach several years.

Each sensor network node usually contains data input/output ports with various sensors control external environment(or the sensors themselves), a microcontroller and a radio transceiver, as well as an autonomous or external power source. This allows the device to receive measurement results, perform initial data processing, and communicate with an external information system. The microcontroller can be used to implement intelligent distributed data processing. In an intelligent wireless sensor network, devices are able to exchange information at the local level, analyze it and transmit processed information to a certain depth, rather than "raw" data. This can significantly reduce the requirements for bandwidth network, increase the scalability and service life of the system. However, adding "intelligence" to the network requires taking into account the specifics of the applied task, so this approach is usually effective when developing a custom highly specialized system.

In this way key features of sensor networks are:

• the ability of self-organization of the information transmission network and its adaptation to the number of devices;
• the ability to relay messages from one element to another;
• the possibility of having sensors in each element;
• long term battery life(1 year or more)

Today, the technology of wireless sensor networks is the only one that can be used to solve monitoring and control tasks that are critical to the requirements for battery life of devices, their reliability, automatic or semi-automatic configuration of each of them, the possibility simple addition or exclusion of the device from the network, propagation of signals through walls and ceilings at a low system cost. And the technology of relayed short-range radio communication, known as "Sensor networks", is one of the modern directions in the development of self-organizing fault-tolerant distributed systems for industrial monitoring and resource and process control.

The advantages of wireless sensor networks technologies can be effectively used to solve various applied problems related to the distributed collection, analysis and transmission of information.

Building automation

In some building automation applications, the use of traditional wired communication systems is not feasible for economic reasons.

For example, you need to introduce a new or expand existing system in a used building. In this case, the use of wireless solutions is the most acceptable option, because. no additional installation work is required with a violation of the interior decoration of the premises, practically no inconvenience is caused to employees or residents of the building, etc. As a result, the cost of implementing the system is significantly reduced.

Another example would be open-plan office buildings for which it is not possible to specify the exact locations of sensors at the design and construction stage. At the same time, the layout of offices can change many times during the operation of the building, therefore, the time and money spent on reconfiguring the system should be minimal, which can be achieved by using wireless solutions.

In addition, the following examples of systems based on wireless sensor networks can be given:

• monitoring of temperature, airflow, presence of people and control of heating, ventilation and air conditioning equipment in order to maintain the microclimate;
• lighting control;
• energy management;
• collection of readings from apartment meters for gas, water, electricity, etc.;
• monitoring of the state of load-bearing structures of buildings and structures.

industrial automation

Until now, the widespread use of wireless communication in the field of industrial automation has been held back by the poor reliability of radio links compared to hardwired connections in harsh environments. industrial operation, but wireless sensor networks are fundamentally changing the current situation, because inherently resistant to various kinds of disturbances (for example, physical damage to the node, the appearance of interference, changing obstacles, etc.). Moreover, under some conditions, a wireless sensor network can provide even greater reliability than a wired communication system.

Solutions based on wireless sensor networks fully meet the requirements of the industry:

• fault tolerance;
• scalability;
• adaptability to operating conditions;
• energy efficiency;
• taking into account the specifics of the applied task;
• economic profitability.

Wireless sensor network technologies can be used in the following industrial automation tasks:

• remote control and diagnostics of industrial equipment;
• maintenance of equipment according to the current state (prediction of the safety margin);
• monitoring of production processes;
• telemetry for research and testing.

Other applications

The unique features and differences of wireless sensor networks from traditional wired and wireless data transmission systems make their application effective in the most various areas. For example:

• security and defense:
  • control over the movement of people and equipment;
  • funds operational communications and intelligence;
  • perimeter control and remote monitoring;
  • assistance in rescue operations;
  • monitoring of property and valuables;
  • security and fire alarm;
• monitoring environment:
  • pollution monitoring;
  • Agriculture;
• healthcare:
  • monitoring the physiological state of patients;
  • location control and notification of medical personnel.

The corporate version of the Internet of Things (IoT) technology is actively used in industry today. The Enterprise Internet of Things (EIoT) uses wireless sensor networks and controls to provide enterprises with new ways to control machines and equipment. Wireless sensors, powered by a small battery and not connected to a wired power supply, can be placed in industrial environments in places completely inaccessible to previous generation controls.

EIoT has increased the reliability, security, and interoperability of systems and equipment to meet the most stringent requirements for the implementation of wireless technologies in this area, not only in industry, but also in healthcare, financial services, etc. EIoT addresses the needs of these areas by what specifications and design elements of this new technology are far superior to similar IoT technologies of traditional devices designed for less critical consumer or commercial applications.

EIoT issues

EIoT-enabled sensors and controls can work almost anywhere in an industrial environment, but so far it has been more of a matter of luck, as not every industrial equipment is ideal for use in wireless networks. This is because there are two interrelated but seemingly contradictory elements in an IoT deployment:

1. The wireless network of devices itself, which is installed using sensors and controls associated with short-range technology with low power consumption.
2. A network of IoT sensors interacting with other equipment, controllers and parts of the network already at a greater distance.

Rice. 1. Applications far from urban centers and traditional telecommunications services can use energy-efficient communication protocol such as LoRa to organize a global network

It is the impossibility of reliable communication over long distances that is often the most significant obstacle in an industrial environment. This problem has a simple cause: telecommunications, which is carried out over wired cable lines or by using signal transmission through towers. cellular communication, is not always available at industrial equipment locations. In addition, the cost of using cellular services only to deliver several packets of data from sensors in one communication session does not make much sense both from an economic point of view and from purely technical considerations. In addition, quite often there is a problem of power supply for sensors and communication devices, which is very difficult to organize in remote places where the equipment or infrastructure is not powered directly from the industrial network.

Despite the wide coverage of cellular communications in settlements, in some places there is no reliable service for organizing wireless communications. This is a common problem in rural areas and remote locations of industrial equipment, such as isolated oil and gas equipment or pipeline transport, water supply and wastewater systems (Fig. 1), etc. Such sites are also often far from the nearest technical service personnel who checks the proper functioning of the devices. Sometimes it takes an engineer a whole day, or even several, to get to the equipment and inspect it. It is often difficult and easy to find specialists willing to work in such remote areas. Since, due to limited communication coverage, EIoT-enabled sensors and controls are quite rare in remote sites, low-power wide area networks (LPWAN) come to the rescue here.

BLE and LPWAN

The most widely used wireless technology short range in EIoT systems is Bluetooth low energy technology - BLE (English Bluetooth low energy, also known as Bluetooth Smart). The main reason for the high popularity of BLE for EIoT is its energy efficiency, which allows sensors and controls to work for a long time with very low battery consumption. BLE manages sleep cycles, standby, and active cycles. BLE is also widely used due to the strength of its RF signal, which allows this technology to work effectively even in difficult environments with increased levels of high-frequency noise, digital signals from computer equipment, and even in the presence of physical obstacles to the propagation of radio waves. But, as you know, all these factors are familiar to the industrial environment.

In projects for the implementation of EIoT, it is BLE technology that is the basis for organizing short-range communications. Moreover, it can be used both on already operated and on industrial equipment complexes that are still being designed. However, such a network of BLE-enabled devices needs a way to receive instructions and relay data over longer distances. Relying on a traditional telecommunications infrastructure that allows for bi-directional Wi-Fi or cellular signals is not possible due to the barrier that limits the application of these sensor and control networks. By combining BLE with the ultra-range and energy efficiency of LoRa technology, companies have been able to deploy EIoT in places where telecommunications infrastructure and power infrastructure are not available, and this, in turn, has expanded the geography of the implementation of the Internet of Things technology.

Rice. 2. Sensors are first connected to the LoRa client and then through the LoRa gateway

The LoRa WAN protocol is often LPWAN as it provides secure bi-directional data transmission and communication with IoT networks over long distances for many years without battery replacement. When using LoRa technology, it is possible to send and receive signals at a distance of up to about 16 km, and repeaters (repeaters) installed if necessary can increase this distance to hundreds of kilometers. On fig. Figure 2 shows how LoRa works. For IoT applications, LoRa has many advantages precisely because of its economic characteristics and capabilities:

• Since LoRa, like BLE, is an ultra-low power technology, it is able to operate on battery-powered IoT device networks and can provide long battery life without requiring frequent maintenance.
• LoRa nodes are inexpensive and allow companies to reduce the cost of data transmission over cellular systems, as well as eliminate the installation of fiber optic or copper cables. This removes a major financial barrier to linking remotely located sensors and equipment.
• LoRa technology works well with network devices placed indoors, including in complex industrial environments.
• LoRa is highly scalable and interoperable by supporting millions of nodes, and can be connected to public and private data networks and bi-directional communication systems.

So, while other LPWAN technologies will only be able to solve the communication range problem in the implementation of IoT solutions in the long term, LoRa technology offers bi-directional communication, anti-jamming and high information content for this.

LoRa also has a significant drawback - low bandwidth. This makes it unsuitable for applications requiring streaming data. However, this limitation does not prevent its use for a wide range of IoT applications where only small data packets are transmitted from time to time.

Interaction

Rice. 3. RM1xx module from Laird, which includes communication capabilities for LoRa and Bluetooth wireless network protocols

The potential of LoRa is doubled when it is combined with technology like BLE. Together, they provide a set of ultra-low power wireless capabilities for short and long range communications that enhance the capabilities of EIoT networks. For example, the central part of urban areas can be covered with just a few LoRaWAN gateways, which are the basis for BLE sensor networks, which are now independent of traditional telecommunications infrastructures. Thus, the symbiosis of LoRa and BLE removes a number of barriers to the expansion of IoT both in megacities and in small cities that have barriers to the widespread implementation of the Internet of Things. However, the biggest beneficiaries of the combination of LoRA and BLE are wireless sensors, controls and other equipment, which can now be installed without any restrictions literally anywhere (Fig. 3). This is a special merit of BLE. BLE also allows these devices to work together in an integrated, short-range network controlled, for example, from smartphones or tablets, which in this case are used as remote wireless displays. In this bundle, LoRa technology, based on the mobile capabilities of BLE, acts as a kind of radio relay station that can send and receive data over long distances. Moreover, these distances can be increased by simple gateways for signal transmission.

There are already many good examples, demonstrating how LoRa and BLE pairing allows EIoT networks to reach a completely different technical level and increase your expansion.

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 has become possible as a result of the combination of microelectromechanical systems, wireless communications 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 transferring 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 peer-to-peer networks, so they are not well suited for 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 mostly use broadcast messages, while most peer-to-peer networks are based on point-to-point communication.
Nodes are limited in power, processing power, and memory.
Nodes cannot have a global an identification number(IN) 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 are focused on achieving high signal quality, mot network protocols should focus mainly on saving energy. They must have mechanisms that allow the user to extend the lifetime of the mote by either reducing the throughput or increasing the latency of the data transfer. 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 various 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 concept of micro sensing and wireless connection promise many new applications for such networks. We have categorized them by major areas: military, environmental research, healthcare, home use, and other commercial applications. 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 in key nodes and sent to the leaders. Data can also be redirected to higher levels of the command hierarchy to be combined 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, it is important to have timely and accurate identification of chemical agents. Sensor networks can be used as warning systems for chemical or biological attacks and data collected in short time help to drastically 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 the 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 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 collection. 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 House"". . 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 home 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 must be developed, network solutions, and intermediate applications. 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, communication 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 observation, then the fault tolerance should be high, since the observation is critical and the nodes can be destroyed during military operations. 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 number functionality even if the price is less than $1.

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). The analog signal generated by the sensor based on the observed phenomenon is converted into digital signal using the ADC, and then fed into the processing unit. The processing module, which uses the integrated memory, controls 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, a location module is included in the overall scheme. 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 a large number of nodes 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 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 because 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. This module, which has 3500 bytes for the OS and 4500 bytes of free memory for the code, uses the TinyOS operating system. 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.
Although great amount sensors and their automatic deployment usually precludes placing them according to a carefully designed plan, schemes for initial deployment 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 different tasks and can 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 compromise 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 transmitter models for these two zones are discussed 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 an IR interface for data transfer. 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 Pt and Pr values ​​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 a 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 efficiently 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 task of collecting data, different kinds 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 determines 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 works on the protocol stack, which is shown in Figure 3. The WINS model discussed in the source, in which the nodes are connected in 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 a 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 usually communicate directly with the base station transmitter, but one-to-one communication 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, for example, 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. BUT various designs nodes will lead to different types of 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.

Distributed sensor networks

What are wireless sensor networks?

Sensors and receiving device

Wireless sensor networks are built from nodes called moty (mote) - small autonomous devices powered by batteries and microchips with radio communication at a frequency - for example 2.4 GHz. Special software allows motes to organize themselves into distributed networks, communicate with each other, interrogate and exchange data with the nearest nodes, the distance to which usually does not exceed 100 meters.

In English literature, such a network is called wireless sensor network(WSN) is a wireless network consisting of geographically distributed autonomous devices that use sensors to jointly monitor the physical or environmental conditions in different areas.

They can measure parameters such as temperature, sound, vibration, pressure, movement of objects or air. The development of wireless sensor networks was initially motivated by military tasks such as battlefield surveillance. Currently, wireless sensor networks are being used increasingly in many areas of civil life, including industrial and environmental monitoring, healthcare, and object movement control. The scope is getting wider.

Basic principles of work

3-level network diagram. 1st Level of sensors and gateway. 2nd server level. Tier 3 thin client

Each network node: mot equipped with a radio transceiver or other wireless communication device, a small microcontroller, and a power source, usually a battery. Can be used with solar panels or other alternative energy sources

Data from distant elements are transmitted over the network between the closest ones from node to node, via a radio channel. As a result, a data packet is transmitted from the nearest mote to the gateway. The gateway is connected, as a rule, with a USB cable to the server. On the server - the collected data is processed, stored and can be accessed through the WEB shell to a wide range of users.

The cost of a sensor node varies from hundreds of dollars to a few cents, depending on the size of the sensor network and its complexity.

Hardware and standards

Gateway (2pcs), connected to a laptop with a USB cable. The laptop is connected to the Internet via UTP and acts as a server

Sensor devices with radio antenna

The hardware of the wireless node and the protocols of network interaction between the nodes are optimized for power consumption to ensure a long service life of the system with autonomous power supplies. Depending on the mode of operation, the lifetime of a node can reach several years.

A number of standards are currently either ratified or under development for wireless sensor networks. ZigBee is a standard for things like industrial control, embedded sensing, medical data collection, building automation. The development of Zigbee is facilitated by a large consortium of industrial companies.

• WirelessHART is an extension of the HART protocol for industrial automation. WirelessHART was added to the generic HART protocol as part of the HART 7 specification, which was approved by the HART Communications Foundation in June 2007.
• 6lowpan is the declared standard for the network layer, but it hasn't been adopted yet.
• ISA100 is another work in an attempt to enter the WSN technology, but is built more widely to include feedback control in their field. Implementation of the ISA100 based on ANSI standards is expected to be completed by the end of the year 2008.

WirelessHART, ISA100, ZigBee, and they are all based on the same standard: IEEE 802.15.4 - 2005.

Wireless sensor network software

Operating system

Operating systems for wireless sensor networks are less complex than generic operating systems due to limited resources in hardware sensor network. Because of this, the operating system does not need to include support for user interfaces.

Wireless sensor network hardware is no different from traditional embedded systems and therefore an embedded operating system can be used for sensor networks

Visualization Applications

Measurement results visualization and reporting software MoteView v1.1

Data from wireless sensor networks is typically stored as digital data in a central base station. There are many standard programs such as TosGUI MonSense, GNS that make it easy to view these large amounts of data. In addition, the Open Consortium (OGC) specifies standards for interoperability and interoperability of encoding metadata, which will allow real-time monitoring or control of the wireless sensor network by anyone through a Web Browser.

To work with data coming from nodes of the wireless sensor network, programs are used that facilitate viewing and evaluating the data. One such program is MoteView. This program allows you to view data in real time and analyze them, build all kinds of graphs, issue reports in various sections.

Benefits of using

• No need to lay cables for power supply and data transmission;
• Low cost of components, installation, commissioning and maintenance of the system;
• Fast and easy network deployment;
• Reliability and fault tolerance of the entire system as a whole in case of failure of individual nodes or components;
• The possibility of implementing and modifying the network at any object without interfering with the process of functioning of the objects themselves
• Possibility of quick and, if necessary, concealed installation of the entire system as a whole.

Each sensor is about the size of a beer cap (but could be scaled down hundreds of times in the future) and contains a processor, memory, and a radio transmitter. Such covers can be scattered over any territory, and they themselves will establish communication with each other, form a single wireless network and begin transmitting data to the nearest computer.

Combined in a wireless network, the sensors can track environmental parameters: movement, light, temperature, pressure, humidity, etc. Monitoring can be carried out over a very large area, because the sensors transmit information along the chain from neighbor to neighbor. The technology allows them to work for years (even decades) without changing batteries. Sensor networks are the universal sense organs for a computer, and all physical objects in the world equipped with sensors can be recognized by a computer. In the future, each of the billions of sensors will receive an IP address, and they may even form something like a Global Sensor Network. So far, only the military and industry have been interested in the capabilities of sensor networks. According to the latest report from ON World, a specialist in sensor network market research, this year the market is experiencing a significant recovery. Another notable event this year was the release of the world's first single-chip ZigBee system (made by Ember). Among large US industrial companies surveyed by ON World, about 29% already use sensor networks, and another 40% plan to deploy them within 18 months. In America, more than a hundred commercial firms have appeared that are engaged in the creation and maintenance of sensor networks.

By the end of this year, the number of sensors on the planet will exceed 1 million. Now not only the number of networks is growing, but also their size. For the first time, several networks of more than 1,000 nodes have been created and successfully operated, including one for 25,000 nodes.

Source: Web PLANET

Application area

The applications of WSN are many and varied. They are used in commercial and industrial systems to monitor data that is difficult or expensive to control using wired sensors. WSNs can be used in areas that are difficult to reach, where they can remain for many years (environmental environmental monitoring) without the need to change power supplies. They can control the actions of violators of a protected facility

WSN is also used for monitoring, tracking and control. Here are some applications:

• Smoke monitoring and detection of fires from large forests and peatlands
• An additional source of information for the Crisis Centers of the Administration of the Subjects of the Federation of the Russian Federation
• Seismic detection of potential tension
• Military observations
• Acoustic object movement detection in security systems.
• Ecological monitoring of space and environment
• Monitoring of industrial processes, use in MES systems
• Medical monitoring

Building automation:

	monitoring of temperature, airflow, presence of people and control of equipment to maintain the microclimate;
	lighting control;
	energy management;
	collection of readings of apartment meters for gas, water, electricity, etc.;
	security and fire alarm;
	monitoring of the state of load-bearing structures of buildings and structures.

Industrial automation:

	remote control and diagnostics of industrial equipment;
	maintenance of equipment according to the current state (prediction of the safety margin);
	monitoring of production processes;