Wireless sensor network and its Applications

Introduction
Wireless sensor networks use sensing techniques to gather information about a phenomenon and react to the events in a specified environment by the means of Sensors. These small, inexpensive, smart devices, which are connected through wireless links, provide unique opportunities for controlling and monitoring environments. Technically, a sensor translates the information from the physical world into signals and prepares them for analysis and processing.
The terms, Wireless Nodes, Sensor nodes and motes can be used interchangeably in different contexts. Here we refer to them as motes. Motes are typically produced in large quantities and are usually densely distributed in the network. Their size(or their components size) varies from macroscopic- scale to microscopic or even sometimes nanoscopic-scale. “Micro-sensors with on-board processing and wireless interfaces can be utilized to study and monitor a variety of phenomena and environments at close proximity.”

A mote is consisted of four major components:
Processing Unit: For data processing and “managing the procedures that make the motes collaborate with other nodes to carry out the assigned sensing tasks.”
Sensing Unit: To sense the physical world and convert the data into digital signal ready for processing.
Transceiver Unit: To provide the connection of nodes in the network.
Power Unit: To supply energy for the device components.
Based on the application, motes may have some additional components such as location finding system, mobilizer and power generator.
These components should be put together in a way to fit in a small size module, be adaptive to different environments and consume as little power as possible.
The components of a mote
Figure is a representation of data acquisition about a phenomena (Process) in the real world which can be sensed by a sensor. The sensed signal needs often needs some changing in order to be processed (Signal Conditioning). For example in order to make the signal range appropriate for conversion some changes on signal magnitude is needed through signal amplification. Unwanted noise can also be removed through this stage.
The analog signal is then transformed to digital signal by using ADC and is ready for further processing or storage.

Data acquisition and actuation
Applications:
Wireless sensor networks can be used in places where wired systems cannot be deployed (e.g., a remote or dangerous area). It can also be used in commercial products to improve the performance or quality of them or provide convenience for their users.
Sensor can sense many different variables such as: temperature, humidity, pressure and movement. They can sense an environment continuously or they can be event driven and sense an event when it occurs.
Wireless sensor networks can support a wide range of applications.
Battlefield surveillance, Bridge and highway monitoring, Earthquake detection, Habitat Monitoring, Health care, Industrial monitoring and control, Tracking wildfires, Traffic flow and surveillance, Video surveillance and Weather monitoring are few examples of its applications.
Military Applications
One of the first applications of sensor network was military sensing. WSN could be used for monitoring the critical equipment, vehicle or weapons to make sure they are in a proper condition. Terrains, paths and roads could be monitored to sense the presence of opposing forces. They also can be used to enhance the targeting system of ammunitions. Human teams can be replaced by sensor networks in places affected by biological and chemical warfare or incidents in order to perform nuclear reconnaissance and prevent humans to be exposed to radiations.
Traffic surveillance
Traffic surveillance is another example of WSN applications. Sensors are placed in predefined places to gather data and send it via wireless links to data centres for further processing. This data can be beneficial for statistical purposes such as vehicle count per day, the number of cars per lane and the average speed of vehicles. It can also be useful for real time applications such as traffic flow monitoring, incident reporting and managing the traffic lights in order to prevent heavy traffics.

Real-time traffic flow control
Medical Applications:
Wireless sensor network benefits are being explored by many hospitals and medical centres around the world. As it can be seen in Figure sensors can be implanted in patient body or connected to him in order to collect information about his vital signs such as heart beat, blood pressure and oxygen level in blood. This information can be transferred patient’s medical record for future examinations and long-term inspections. It also can be displayed in real-time or alert physicians based on the sensor program in case of any sudden change in under-care patient condition.

http://www.infotech.oulu.fi/Annual/2007/opme.html
Realization of these various applications requires wireless ad hoc networking techniques. However they are not suitably designed for special features and applications of sensor networks.

WSN vs. Mobile Adhoc Netoworks
[12] Although there are lots of similarities between Mobile ad networks (MANET) and WSN for instance their lack of network infrastructure, use of multi-hop routing and wireless channel, there are some major differences to point out.
Nodes in MANET are designed for human interaction such as laptop and PDAs, whereas in WSN motes are usually left unattended in remote or dangerous locations with the least possible interactions.
In WSN “the topology of the network may change dynamically” due to node failure. It can happen because sometimes motes in some specific areas may be damaged and fail. In some network topologies motes have a sleep/awake cycle in order to save energy, so the topology needs to change when a mote is not available at a specific time.
In WSN unlike MANETs the source of energy is limited and the nodes are sometimes left unattended in places where there is no access to them to change or recharge their batteries. “The range of communications is typically within a few meters and at low rates (some kilobits per second); there are typically a few kilobytes of memory and the processor may operate at speeds of only some megahertz.”
Mote design and communication aspect of WSN is totally application dependent and changes based on different application requirements.
Motes in some wireless sensor applications remain sleep for the most of their lifetime and transfer their information in a timely basis in order to save energy. So the traffic flow in the network is almost infrequent and delay time is usually higher than MANET networks.
Overview of 802.15.4
1-http://www.ieee802.org/15/pub/TG4.html -> IEEE 802.15 WPAN™ Task Group 4 (TG4)
2-http://www.zigbee.org/Specifications.aspx a ZigBee Alliance the Official Website
3-http://www.eetimes.com/showArticle.jhtml?articleID=173600329 -> EE Times: The global electronics engineering community
The IEEE 802.15.4 and the Zigbee alliance have been working together in order to improve WSN efficiency, safety, security, reliability and convenience of this technology. IEEE 802.15.4 focuses on physical layer and MAC layer at the 868MHz (Europe), 915MHz (US) and 2.4GHz (worldwide) ISM bands whereas Zigbee alliances work on higher level protocols.
“The IEEE 802.15 was chartered to investigate a low data rate solution with multi-month to multi-year battery life and very low complexity. It is operating in an unlicensed, international frequency band.”
“Some of the characteristics of IEEE 802.15.4 include:
Data rates of 250 kbps, 40 kbps, and 20 kbps
CSMA-CA(Carrier sense multiple access with collision avoidance) channel access
Fully handshaked protocol for transfer reliability
Power management to ensure low power consumption
16 channels in the 2.4GHz ISM band, 10 channels in the 915MHz I and one channel in the 868MHz band.”
“The ZigBee specification enhances the IEEE 802.15.4 standard by adding network and security layers and an application framework. From this foundation, Alliance developed standards, technically referred to as public application profiles, can be used to create a multi-vendor interoperable solutions. For custom application where interoperability is not required, manufacturers can create their own manufacturer specific profiles.”
[2]Some of the characteristics of ZigBee include:
Global operation in the 2.4GHz frequency band according to IEEE 802.15.4
Regional operation in the 915Mhz (Americas) and 868Mhz (Europe).
Frequency agile solution operating over 16 channels in the 2.4GHz frequency
Incorporates power saving mechanisms for all device classes
[802]IEEE 802.15.4 standard defines PHY (physical layer) and MAC (medium access control) layer for the purpose of low data rate wireless communications which consume very low power.
Physical Layer
Some of the main characteristics of the PHY are the processes of sensing the environment, turning on/off the transceiver, estimating the receiver power/link quality indication and transmitting/receiving the information between two nodes. It finally sends the result of channel assessment to the MAC layer. The PHY is responsible for providing two services:
PHY Data Service: “Enables the transmission and reception of PHY protocol data units (PPDUs) across the physical radio channel.
PHY management service
There are different frequency bands and data rates which a device should be able to operate with which are summarized in Table ?.
Table – Frequency bands and data rates
Mac Layer
MAC layer provides access to the physical radio channel to transmit MAC frames.
Some of the main characteristics of MAC sublayer are network beaconing, frame validation, Guarantees time slots (GTS) and handles node associations.
The MAC layer is responsible for providing two services:
MAC Data Service: “Enables the transmission and reception of MAC protocol data units (MPDUs) across the PHY data service.”
MAC Management Service
IEEE 802.15.4 MAC can work with both beacon enabled and non-beacon models. When it is on non-beacon model it is a simple CSMA/CA protocol but in beacon enable mode it works with super frame structure, shown in Fig. The frame starts with a Beacon which is sent by coordinator periodically. The frame also contains inactive period and active period. During the inactive period the device switches to low power mode and communicate with others during active period. The Beacon Interval is calculated based different attributes. In Active period the portion is divided into 16 slots which consist of three parts: Connection Access Period (CAP), Collision Free Period (CFP) (the GTS sections within it is for specific nodes) and the beacon.
Fig Superframe structure
Network Topologies
ZigBee supports 3 types of topologies: Star, Mesh(peer-to-peer) and Cluster tree as shown in Fig .
– Star topology:
In this topology the communication is only between the single central controller called Personal Area Network (PAN) coordinator and other devices in the network which is mostly suitable for small networks such as single hop networks. A PAN coordinator usually has a unique identifier which is only used by this specific coordinator and allows different star networks to operate separately in the same area.
– Mesh topology:
This topology also has a PAN coordinator like Star topology but with the difference of having communication not only between coordinator and devices but between devices as well when they are in the range of one another. Although it makes the network structure more complex, but as a result of allowing multi-hop routing it is suitable for large networks. It also can be an adhoc network with self-healing and self-organizing characteristics.
– Cluster tree topology:
Cluster tree network is a form of peer-to-peer network. One coordinator operates as a PAN coordinator which has the responsibility of defining Cluster Heads (CH). The CH is a kind of Full Function Device (FFD) which can act as a coordinator. Each Reduced Function Device (RFD) then can selects its CH and joins that cluster. This kind of structure has a great impact on energy saving in the network which will be discussed later.
Fig Topology Model
Energy Conservation and measurement:
[24]A wireless sensor network is created with hundreds or thousands of sensor motes, distributed independently in a remote area with the responsibility of sensing the environment, processing information and communicating with other motes in the network for years with a limited source of energy provided by a small battery which is almost impossible to be changed or recharged during motes life time. Therefore the concept of energy consumption management in the network has become one of the most important aspects of wireless sensor network design and implementation. The power saving approach has affected the mote design, power management strategies, communication and routing protocols of the WSN.
Generally energy saving methods are divided in two major categories:
Energy saving at Mote level; aims to selects the most energy efficient components of the device and trade off unnecessary operations in order to save energy based on the application requirements.
Energy saving at Communication level; selecting the most efficient communication methods and protocols to conserve energy at this level.
Power saving at mote level:
The first step in saving energy at mote level is to find out where the energy is consumed in the mote. As it was mentioned before, a mote consists of 4 components: Processing Unit, Sensor, Transceiver and a Power supply to provide energy for the mentioned parts.
Based on the experimental measurements in [40] data transmission is more energy consuming that data processing. Passive sensors such as temperature sensors on the other hand consume a small amount of power compared to other components which is almost usually negligible. Table shows a power model of a Mica2 mote in different states.