You are on page 1of 31


Wireless technology has expanded the limits of our world. Through this innovation, people have been given freedom to work away from their desks or even outside. The new found freedom that people are beginning to enjoy with their computers has started making the world of technology and nature blend. Wireless Sensor Networks are the next stage of this technology-nature cohesion. Although a young technology, the applications have been varied and promise to be even more varied. These networks are collections of small devices, known as nodes, with limited computational power. Each node has approximately 1-100th of the computing power of a PDA, but when combined with hundreds of other nodes, they combine to form an extremely capable system. Wireless Sensor Networks, or WSNs, have been used to enable better data collection in scientific studies, create more effective strategic military defences, pinpoint the origin of a gunshot, and monitor factory machinery. All of these uses depend on the ability to collect data such as light, vibration, moisture, temperature, and more, as well as the ability to communicate with each other. This last ability is what makes a collection of nodes so much more powerful than any node in particular. The purpose of this report is to explain how wireless sensor networks work, including the workings for each individual node. The way the multi-function chips are made, as well as the communication system will be discussed. Specialized nodes will not be discussed, as multifunction nodes are more appropriate for this report. Past uses and potential applications will be touched upon briefly to further illustrate the faculty of these machines. The introduction of these collections of computing devices has brought forth changes in factory safety, machine maintenance, data collection, and military effectiveness
Smart environments represent the next evolutionary development step in building, utilities, industrial, home, shipboard, and transportation systems automation. Like any sentient organism, the smart environment relies first and foremost on sensory data from the real world. Sensory data comes from multiple sensors of different modalities in distributed locations. The smart environment needs information about its surroundings as well as about its internal workings; this is captured in biological systems by the distinction between exteroceptors and proprioceptors.

The challenges in the hierarchy of: detecting the relevant quantities, monitoring and collecting the data, assessing and evaluating the information, formulating meaningful user displays, and performing decision-making and alarm functions are enormous. The information needed by smart environments is provided by Distributed Wireless Sensor Networks, which are responsible for sensing as well as for the first stages of the processing 1

hierarchy. The importance of sensor networks is highlighted by the number of recent funding initiatives, including the DARPA SENSIT program, military programs, and NSF Program Announcements.


Wireless Sensor Networks are collections of nodes. Nodes are the individual computers that work together to form networks. The requirements for nodes are extensive. They must be small, energy efficient, multifunctional, and wireless. Collections of nodes communicate with each other to reach a common goal. For example, if the goal is to collect information about the microclimates around all sections of redwoods in a forest, the nodes are placed in the trees to form a network. Once placed, they collect and transmit data to each other, and eventually to a main computer. Due to interference from the surroundings and the node¶s maximum broadcast range, not all of the nodes placed around trees can communicate will all others. The node¶s radios are designed to save as much power as possible and therefore have a limited broadcast range. This range is approximately 30 meters . If the nodes have a short radio broadcast range, and many nodes are more than 30 meters off the ground, how can one collect data from the nodes farthest away from the computer (or station)? Nodes solve this problem by packaging their information and broadcasting it to multiple other nodes, which then communicate with others, to find the most rapid or successful route for the information to travel to reach the main computer located elsewhere. Nodes communicate with each other using radio transmitters and receivers. They form networks with other nodes that change with the positions of the nodes. They create links with each other in different configurations to maximize the performance for each node. These links all lead to the ³parent´ node, which transmits the information from each of the ³child´ nodes to whatever computer or PDA type device is used to collect and process the data. Figure 1 illustrates one possible path data can travel between the outer nodes and those close to the computer/station. The Nodes 1 through 13 are the children nodes (all the ones in light grey), Node 14 is the parent (in purple). The ³Computer´ (in red) can be any type of computer such as PDA, laptop, etc. as long as it is capable of accessing the internet via a specified ISP (the grey building with yellow windows). The arrows connecting the nodes are not fixed, and to illustrate this, they are purposefully unorganized. When the nodes are linked together, they form parts of a machine with greater computational power than any of the individual parts.



These ³machines´ of nodes change with position and with conditions. Sometimes high moisture and other situations can affect broadcast abilities of many nodes. Changes in conditions can make some connections stronger than they used to be, and others nearly impossible. The thinking capability within the network allows the pieces to reorganize in such a way that all nodes will continue to be functional.

(2.1)SAFETY MECHANISM. In addition to communicating, nodes also adapt to their situation. In the case of a malfunction the remaining nodes will reform the network. For example, there are 500 nodes in place around a system. If 20% or 100 nodes die, the rest of the nodes will reconfigure the network to continue working with the remaining 400. Furthermore, 400 nodes will collect as much data as 400 scientists working non-stop for that period of time. The bigger problems that face these networks are people with destructive intentions, and the potential for nodes to keep ³working´ while spitting out bad information. Corrupt data can sometimes be caught when 4

the information is used and reread by humans, but the times that it goes unnoticed can slightly or significantly alter conclusions drawn from the data. The threat of hackers is a serious problem because the operating system for the nodes is an ³open-source´ system, which allows relatively easy access to codes. Some security systems are in place currently, such as code recognition software imbedded in the operating systems. All security procedures will develop with the growth of the technology in response to a larger number of hackers. The harder people try to break the system, the more the system will be protected

(2.2)CONFIGRATION OF NODES. Nodes, the individual computers that make up a Wireless Sensor Network, are very small and simple. While we associate a computer with a PC, the technical definition of a computer is a thing that computes, be it human or machine, thus nodes are computers. Most nodes consist of five crucial components. These components include a number of sensors, such as temperature, moisture, and vibration sensors, a power source, in the case of older nodes, 2 AA batteries, a radio transmitter/receiver, and an electric ³brain.´(Figure 2)


Sensors: When nodes are under construction, their intended purpose often dictates the sensors that are added to the node. The node in Figure 2 contains three types of sensors: temperature, moisture, and vibration. This is a fairly typical node, but some nodes have many more functions. There are nodes that take photographs of the surroundings, sense motion, measure light intensity, and much more. The sensors are attached to the node base and communicate readings to the electronic brain.


Power Source: The power source for the node also depends the node¶s intended use. If the node is designed to last a very long time, say one year, it will have a larger power source than a node that is only meant to run for a month. The power sources usually range between a couple of AA batteries, and a watch battery, but with the new smart-dust nodes, also called ³Spec,´ they can collect enough energy to sustain themselves from ambient light, or even vibrations. The power source is connected to the node base and provides energy required to run the sensors, electronic brain, and radio. Electronic brain ,Power source .Sensors ,Radio 5


The Electronic Brain: The older nodes¶ brains consist of a microprocessor and some flash memory. Many of them have connectors to add other processes and sensors with ease. The MEMS nodes also contain an analog-digital converter. The basic functions of the electronic brain are to make decisions and deal with collected data. The electronic brain stores collected data in its memory until enough information has been collected. Once this point is reached, the microprocessor portion of the electronic brain then puts the data in ³envelopes,´ or packages of data formatted for greatest transferring efficiency. These envelopes are then sent to the radio for broadcast. The brain also communicates with other nodes to maintain the most effective network in much the same way it deals with data. The electronic brain is connected to the base and interacts with the sensors and radio.


Node base: The node base is simply the base on which the node is built. In the case of MEMS nodes, the base is the inactive metal layer in the chip, and in the case of multichip nodes, the base consists of a circuit board that provides connections between the node¶s pieces.

Figure 2. This figure illustrates the construction of a first generation node, usually known as a Mica-node.



Radio: The radio consists of a radio transmitter and a radio receiver. Both of these parts must exist for any node to fully communicate with the other nodes. The radio, when transmitting, receives information from the electronic brain and broadcasts the data to other nodes according to the network connections. In the other direction, when receiving, the radio receives information from another node¶s radio and transmits it to the electronic brain. The radio is connected to the node base.

(2.3)Node: Theory of Operation Nodes collect and transfer data using four stages: collecting the data, processing the data, packaging the data, and communicated the data. Each node collects data using its various types of sensors. After collecting the data, the node processes the data using its electronic brain. Once the data has been processed, the brain packages the data into an easily handled . This process is known as enveloping. Once the data has been collected and processed to this point, the node then begins to interact with other nodes. This process can be seen in Figure 3.

Fig. 3. Ordered processes a node undergoes when collecting data and communicating data to other nodes.

This process is one of many ways nodes differ from traditional computers. Another way is nodes spend as much as 99% of their time ³sleeping´ to conserve energy, only waking up to record data, send and receive information (either data or instructions), or when instructed to by its programming.



Transducers and Physical Transduction Principles A transducer is a device that converts energy from one domain to another. In our application, it converts the quantity to be sensed into a useful signal that can be directly measured and processed. Since much signal conditioning (SC) and digital signal processing (DSP) is carried out by electronic circuits, the outputs of transducers that are useful for sensor networks are generally voltages or currents. Mechanical Sensors include those that rely on direct physical contact. The Piezoresistive Sensors converts an applied strain to a change in resistance that can be sensed using electronic circuits such as the Wheatstone Bridge , the relationship is , with R the resistance, _ the strain, and S the gauge factor which depends on quantities such as the resistivity and the Poisson¶ ratio of the material. There may be a quadratic term in _ for some materials. Metals and semiconductors exhibit piezoresistivity. The piezoresistive effect in silicon is enhanced by doping with boron (p-type silicon can have a gauge factor up to 200). With semiconductor strain gauges, temperature compensation is important. The Piezoelectric Sensors, discovered by the Curies in 1880, converts an applied stress (force) to a charge separation or potential difference. Piezoelectric materials include barium titanate, PZT, and single-crystal quartz. The piezoelectric effect is reversible, so that a change in voltage also generates a force and a corresponding change in thickness. Thus the same device can be both a sensor and an actuator. Combined sensor/actuators are an intriguing topic of current research. Tunneling Sensors depends on the exponential relationship between the tunneling current I and the tip/surface separation z given by , where k depends on the tunnel barrier height in ev. Tunneling is an extremely accurate method of sensing nanometer-scale displacements, but its highly nonlinear nature requires the use of feedback control to make it useful. Capacitive Sensors typically have one fixed plate and one movable plate. When a force is applied to the movble plate, the change in capacitance C is given as , with the resulting displacement, A the area, and _ the dielectric constant. Changes in capacitance can be detected using a variety of electric circuits and converted to a voltage or current change for further processing. Magnetic and Electromagnetic Sensors do not require direct physical contact and are useful 8

for detecting proximity effects The Hall Effect, discovered by Edwin Hall in 1879, relies on the fact that the Lorentz Force deflects flowing charge carriers in a direction perpendicular to both their direction of flow and an applied magnetic field (i.e. vector cross product).

Thermo-Mechanical Sensors is used for temperature sensing and regulation in homes and automobiles. On changes in temperature T, all materials exhibit thermal expansion of the form TLL , with L the length and _ the coefficient of linear expansion. One can fabricate a strip of two joined materials with different thermal expansions. Then, the radius of curvature of this thermal bimorph depends on the temperature change. Thermoresistive Sensors are based on the fact that the resistance R changes with temperature T. For moderate changes, the relation is approximately given by for many metals by TRRR with the temperature coefficient of resistance. The relationship for silicon is more complicated but is well understood. Hence, silicon is useful for detecting temperature changes. Thermocouples are based on the thermoelectric Seebeck effect, whereby if a circuit consists of two different materials joined together at each end, with one junction hotter than the other, a current flows in the circuit. Thermocouples are inexpensive and reliable, and so are much

used. Typical thermocouples have outputs on the order of 50 V/ C and some are effective for
o o

temperature ranges of -270 C to 2700 C. Optical Sensors convert light to various quantities that can be detected . These are based on one of several mechanisms. In the photoelectric effect one electron is emitted at the negative end of a pair of charged plates for each light photon of sufficient energy. This causes a current to flow. In photoconductive sensors, photons generate carriers that lower the resistance of the material. In junction-based photo sensors, photons generate electron-hole pairs in a semiconductor junction that causes current flow. This is often misnamed the photovoltaic effect. These devices include photodiodes and phototransistors. Chemical And Biological Sensors cover a very wide range of devices that interact with solids, liquids, and gases of all types. Potential applications include environmental monitoring, biochemical warfare monitoring, security area surveillance, medical diagnostics, implantable biosensors, and food monitoring. Effective use has been shown for NOx (from pollution), organo phosphorus pesticides, nerve gases (Sarin, etc), hydrogen cyanide, smallpox, anthrax, COx, SOx, and others. 9

Chemiresistors have two interdigitated finger electrodes coated with specialized chemical coatings that change their resistance when exposed to certain chemical challenge agents. The electrodes may be connected directly to an FET, which amplifies the resulting signals in situ for good noise rejection. This device is known as an interdigitated-gate electrode FET (IGEFET). Arrays of chemiresistors, each device with a different chemically active coating, can be used to increase specificity for specific challenge agents. Digital signal processing, including neural network classification techniques, is important in correct identification of the agent. Electrochemical Sensors rely on currents induced by oxidation or reduction of a chemical species at an electrode surface. These are among the simplest and most useful of chemical sensors. An electron transfer reaction occurs that is described by O, with O the oxidized species, R the reduced species, and z the charge on the ion involved. Bio Sensors of a wide variety of types depend on the high selectivity of many biomolecular reactions, e.g. molecular binding sites of the detector may only admit certain species of analyte molecules. Unfortunately, such reactions are not usually reversible so the sensor is not reusable. These devices have a biochemically active thin film deposited on a platform device that converts induced property changes (e.g. mass, resistance) into detectable electric or optical signals.



Operating systems for wireless sensor network nodes are typically less complex than generalpurpose operating systems. They more strongly resemble embedded systems, for two reasons. First, wireless sensor networks are typically deployed with a particular application in mind, rather than as a general platform. Second, a need for low costs and low power leads most wireless sensor nodes to have low-power microcontrollers ensuring that mechanisms such as virtual memory are either unnecessary or too expensive to implement. It is therefore possible to use embedded operating systems such as eCos or uC/OS for sensor networks. However, such operating systems are often designed with real-time properties.

TinyOS is perhaps the first operating system specifically designed for wireless sensor networks. TinyOS is based on an event-driven programming model instead of multithreading. TinyOS programs are composed of event handlers and tasks with run-to-completion semantics. When an external event occurs, such as an incoming data packet or a sensor reading, TinyOS signals the appropriate event handler to handle the event. Event handlers can post tasks that are scheduled by the TinyOS kernel some time later. TinyOS is a free and open source component-based operating system and platform targeting wireless sensor networks (WSNs). TinyOS is an embedded operating system written in the nesC programming language as a set of cooperating tasks and processes. It is intended to be incorporated into smart dust. TinyOS applications are written in nesC, a dialect of the C language optimized for the memory limits of sensor networks. Its supplementary tools are mainly in the form of Java and shell script front-ends. Associated libraries and tools, such as the NesC compiler and Atmel AVR binutils toolchains, are mostly written in C. TinyOS programs are built out of software components, some of which present hardware abstractions. Components are connected to each other using interfaces. TinyOS provides


interfaces and components for common abstractions such as packet communication, routing, sensing, actuation and storage. TinyOS is completely non-blocking: it has one stack. Therefore, all I/O operations that last longer than a few hundred microseconds are asynchronous and have a call-back. To enable the native compiler to better optimize across call boundaries, TinyOS uses nesC's features to link these callbacks, called events, statically. While being non-blocking enables TinyOS to maintain high concurrency with one stack, it forces programmers to write complex logic by stitching together many small event handlers. To support larger computations, TinyOS provides tasks, which are similar to a Deferred Procedure Call and interrupt handler bottom halves. A TinyOS component can post a task, which the OS will schedule to run later. Tasks are non-pre-emptive and run in FIFO order. This simple concurrency model is typically sufficient for I/O centric applications, but its difficulty with CPU-heavy applications has led to the development of a thread library for the OS, named TOS Threads. TinyOS code is statically linked with program code, and compiled into a small binary, using a custom GNU toolchain. Associated utilities are provided to complete a development platform for working with TinyOS LiteOS LiteOS is a newly developed OS for wireless sensor networks, which provides UNIX-like abstraction and support for the C programming language. Contiki is an OS which uses a simpler programming style in C while providing advances such as 6LoWPAN and protothreads. This operating system is multithreaded and comes bundled with a UNIX-like file system. It supports C programming natively, and allows online debugging to locate application bugs. The purpose of LiteOS is to significantly reduce such a learning curve. The rapid advances of sensor networks in the past few years created many exciting systems and applications. The most obvious challenge in sensor network development has been to fit within extremely constrained platform resources.

LiteOS differs from both current sensor network operating systems and more conventional embedded operating systems. Compared to the former category, such as TinyOS, LiteOS provides a more familiar environment to the user. Its features are either not available in 12

existing sensor network operating systems, such as the shell, and the hierarchical file system, or are only partially supported. The kernel subsystem of LiteOS takes the thread approach. We implement two different scheduling policies: priority-based scheduling, where the thread with the highest priority always gets dispatched, and round-robin scheduling, where each thread gets a slice of CPU, whose length is decided by its priority. By default, eight threads are supported. Unlike Tiny Thread, which implements cooperative scheduling, the LiteOS kernel takes back the CPU every 0.1 seconds. Of course, the correct execution of this design choice relies on an assumption that user threads are not malicious, and will not disable interrupts to enter an infinite loop.



Since a distributed network has multiple nodes and services many messages, and each node is a shared resource, many decisions must be made. There may be multiple paths from the source to the destination. Therefore, message routing is an important topic. The main performance measures affected by the routing scheme are throughput (quantity of service) and average packet delay (quality of service). Routing schemes should also avoid both deadlock and livelock . Routing methods can be fixed , adaptive, centralized, distributed, broadcast, etc. Perhaps the simplest routing scheme is the token ring. Here, a simple topology and a straightforward fixed protocol result in very good reliability and precomputable QoS. A token passes continuously around a ring topology. When a node desires to transmit, it captures the token and attaches the message. As the token passes, the destination reads the header, and captures the message. In some schemes, it attaches a µmessage received¶ signal to the token, which is then received by the original source node. Then, the token is released and can accept further messages. The token ring is a completely decentralized scheme that effectively uses TDMA. Though this scheme is very reliable, one can see that it results in a waste of network capacity. The token must pass once around the ring for each message. Therefore, there are various modifications of this scheme, including using several tokens, etc.

Fixed routing schemes often use Routing Tables that dictate the next node to be routed to, given the current message location and the destination node. Routing tables can be very large for large networks, and cannot take into account real-time effects such as failed links, nodes with backed up queues, or congested links.

Adaptive routing schemes depend on the current network status and can take into account various performance measures, including cost of transmission over a given link, congestion of a given link, reliability of a path, and time of transmission. They can also account for link or node failures. Routing algorithms can be based on various network analysis and graph theoretic concepts in Computer Science (e.g. A-star tree search), or in Operations Research including shortestroute, maximal flow, and minimum-span problems. Routing is closely associated with 14

dynamic programming and the optimal control problem in feedback control theory Shortest Path routing schemes find the shortest path from a given node to the destination node. If the cost, instead of the link length, is associated with each link, these algorithms can also compute minimum cost routes. These algorithms can be centralized (find the shortest path from a given node to all other nodes) or decentralized (find the shortest path from all nodes to a given node).


With the increasing ubiquity of WSNs, environmental data will be available almost everywhere in our environment. We believe that in the future the current temperature, humidity, etc. at a particular location will be available on demand from a surrounding WSN. f course, accessing this data will in general not be for free since deployment of WSNs induces some costs. This means that the deployment agencies of some of these services will make them available only to ³authorized´ people (i.e., paying customers). In this case, a WSN must be able to distinguish legitimate users from illegitimate users, resulting in the problem of access control. Access control is an old problem from classical computer science but has not received much attention in the context of WSNs. This is unfortunate since WSNs define an environment which naturally calls for security solutions but²due to the resourceconstraintswith respect to computational and battery power for example ² also defines an environment in which security solutions are extremely hard to implement. This extended abstract investigates the problem of access control in WSNs. More specifically, we focus on the problem of user authentication in WSNs, an important subproblem of access control.

(5.1).Simple Authentication.
Formal definition of authentication protocols is defined with respect to the two primitive operations of authentication: (1) authenticate(V, I) is invoked by the user P whenever P would like to be authenticated by V using identity I 2 I (2) associate(P, I) is invoked by the verifier whenever it has established the relation between P and some identity I. Intuitively, an authentication protocol is correct if the identity associated to P by V is the ³real´ identity of P. If P is dishonest or claims to have a fake identity this is indicated by a special value ? which is supposed to be distinct from any value in I. Authentication is successful if V invokes associate(P, I) with some I 6= ?. More precisely, a protocol solves authentication if it guarantees two properties: ‡ (Validity) An honest verifier V invokes associate(P, I) with I 2 I only if P in fact has identity I. ‡ (Termination) If P invokes authenticate(V, I) and if V is honest then V will eventually invoke associate(P, I0) for some identity I0 2 I or I0 = ?. 16

We call a protocol which satisfies the above two conditions a simple authentication protocol. Simple authentication is not sufficient in wireless sensor networks if failures and active adversaries are taken into account. If we require that a user (i.e., a user) always authenticates to some particular sensor, then this becomes impossible if that sensor fails. However, if we don¶t care which sensor the user uses for authentication, then taking control of a single sensor is sufficient for an active adversary to gain access to the entire system. What is needed is a more robust notion of authentication.

(5.2). n-Authentication. n-authentication is a robust version of simple authentication. To be robust against failures, this new form of authentication succeeds if the user can successfully authenticate with any subset of sensors out of a set of n sensors (n can be the average number of sensors within broadcast distance of the user). To be robust against active attacks where the adversary can compromise up to t sensors (t < n), we require that the subset of sensors to which the user has to authenticate has at least the size of n í t. More formally, we now consider a set of n verifiers V = {V1, . . . , Vn}. To distinguish the primitive operations of simple authentication from those of n-authentication we denote the latter ones with n-associate(P, I) and n-authenticate(V, I). Note that n-authenticate refers to the entire set of verifiers while n-associate just refers to a single user. A protocol solves nauthentication if it satisfies the following properties: ‡ (Termination) If P invokes n-authenticate(V, I) then eventually all honest Vi 2 V invoke nassociate(P, Ii) for some Ii 2 I or I = ?. ‡ (Validity) An honest verifier Vi invokes n-associate(P, I) only if P in fact has identity I 2 I. ‡ (Agreement) If honest verifier Vi invokes n-associate(P, I0) and honest verifier V invokes n-associate(P, I00) then I0 = I00. If we assume that at most t verifiers fail, then n-authentication ensures that the remaining (at least nít) verifiers eventually successfully authenticate an honest user and that they agree on its identity. If a user is dishonest or claims to have a fake identity then all honest verifiers will return ? so that the user is not authenticated.



The figure showing the IEEE 1451 Smart Sensor includes basic blocks for signal conditioning (SC), digital signal processing (DSP), and A/D conversion.

Signal Conditioning
Signals coming from MEMS sensors can be very noisy, of low amplitude, biased, and dependent on secondary parameters such as temperature. Moreover, one may not always be able to measure the quantity of interest, but only a related quantity. Therefore signal conditioning is usually required. SC is performed using electronic circuitry, which may conveniently be built using standard VLSI fabrication techniques in situ with MEMS sensors. A reference for SC, A/D conversion, and filtering . A real problem with MEMS sensors is undesired sensitivity to secondary quantities such as temperature. Temperature compensation can often be directly built into a MEMS sensor

circuit. In the figure above showing a 3x3 array of IGEFET sensors, there is shown a 10 IGEFET- this is for temperature compensation. Temperature compensation can also be added during the SC stage as discussed below. A basic technique for improving the signal-to-noise ratio (SNR) is low-pass filtering, since noise generally dominates the desirable signals at high frequencies. Shown in the figure is an analog LPF that also amplifies, constructed from an operational amplifier. Such devices are easily fabricated using VLSI semiconductor techniques. The time constant of this circuit is CR2 . The transfer function of this filter is with 3 dB cut-off frequency given by rad. and gain given by . Here, s is the Laplace transform variable. The cut-off frequency should be chosen larger than the highest useful signal frequency of the sensor. Here, z is the z-transform variable treated as a unit delay in the time domain, is the measured signal, and s is the filtered or smoothed variable with reduced noise content. The filter parameters are selected in terms of the desired cut-off frequency and the sampling period It is often the case that one can measure a variable sk (e.g. position), but needs to know its rate of change vk (e.g. velocity). Due to the presence of noise, one cannot simply take the difference between successive values of sk as the velocity. A filtered velocity estimate given by v both filters out noise and gives a smooth velocity estimate. Often, changes in resistance 18

must be converted to voltages for further processing. This may be accomplished by using a Wheatstone bridge . Suppose R1= R in the figure is the resistance that changes depending on the measurend (e.g. strain gauge), and the other three resistances are constant (quarter bridge configuration). Then the output voltage changes according to . We assume a balanced bridge so that Sensitivity can be improved by having two sensors in situ, such that the changes in each are opposite (e.g. two strain gauges on opposite sides of a flexing bar). This is known as a half bridge. If R1 and R2 are two such sensors and , then the output voltage doubles. The Wheatstone bridge may also be used for differential measurements (e.g. for insensitivity to common changes of two sensors), to improve sensitivity, to remove zero offsets, for temperature compensation, and to perform other signal conditioning. Specially designed operational amplifier circuits are useful for general signal conditioning . Instrumentation Amplifiers provide differential input and common mode rejection, impedance matching between sensors and processing devices, calibration, etc. SLEEPMODE amplifiers (Semiconductor Components Ind., LLC) consume minimum power while asleep, and activate automatically when the sensor signal exceeds a prescribed threshold.


With the advent of ad hoc networks of geographically distributed sensors in renode site environments (e.g. sensors dropped from aircraft for personnel/vehicle surveillance), there is a focus on increasing the lifetimes of sensor nodes through power generation, power conservation, and power management. Current research is in designing small MEMS (microelectromechanical systems) RF components for transceivers, including capacitors, inductors, etc. The limiting factor now is in fabricating micro-sized inductors. Another thrust is in designing MEMS power generators using technologies including solar, vibration (electromagnetic and electrostatic), thermal, etc. RF-ID (RF identification) devices are transponder microcircuits having an L-C tank circuit that stores power from received interrogation signals, and then uses that power to transmit a response. Passive tags have no onboard power source and limited onboard data storage, while active tags have a battery and up to 1Mb of data storage. RF-ID operates in a low frequency range of 100kHz-1.5MHz or a high frequency range of 900 MHz-2.4GHz, which has an operating range up to 30m. RF-ID tags are very inexpensive, and are used in manufacturing and sales inventory control, container shipping control, etc. RF-ID tags are installed on water meters in some cities, allowing a metering vehicle to simply drive by and renodely read the current readings. They are also be used in automobiles for automatic toll collection. Meanwhile, software power management techniques can greatly decrease the power consumed by RF sensor nodes. TDMA is especially useful for power conservation, since a node can power down or µsleep¶ between its assigned time slots, waking up in time to receive and transmit messages. The required transmission power increases as the square of the distance between source and destination. Therefore, multiple short message transmission hops require less power than one long hop. In fact, if the distance between source and destination is R, the power required for

single-hop transmission is proportional to R . If nodes between source and destination are taken advantage of to transmit n short hops instead, the power required by each node is
2 2

proportional to R /n . This is a strong argument in favour of distributed networks with multiple nodes, i.e. nets of the mesh variety. A current topic of research is active power control, whereby each node cooperates with all other nodes in selecting its individual transmission power level . This is a decentralized 20

feedback control problem. Congestion is increased if any node uses too much power, but each node must select a large enough transmission range that the network remains connected. For n nodes randomly distributed in a disk, the network is asymptotically connected with probability one if the transmission range r of all nodes is selected using function that goes to infinity as n becomes large. and electromagnetic method



In queuing networks, each node has an associated queue or buffer that can stack messages. In such networks, flow control and resource assignment are important. The objectives of flow control are to protect the network from problems related to overload and speed mismatches, and to maintain QoS, efficiency, fairness, and freedom from deadlock. If a given node A has high priority, its messages might be preferentially routed in every case, so that competing nodes are choked off as the traffic of A increases. Fair routing schemes avoid this. There are several techniques for flow control: In buffer management, certain portions of the buffer space are assigned for certain purposes. In choke packet schemes, any node sensing congestion sends choke packets to other nodes telling them to reduce their transmissions. Isarithmic schemes have a fixed number of µpermits¶ for the network. A message can be sent only if a permit is available. In window or kanban schemes, the receiver grants µcredits¶ to the sender only if it has free buffer space. Upon receiving a credit, the sender can transmit a message. In Transmission Control Protocol (TCP) schemes (Tahoe and Reno) a source linearly increases its transmission rate as long as all its sent messages are acknowledged for. When it detects a lost packet, it exponentially decreases its transmission rate. Since lost packets depend on congestion, TCP automatically decreases transmissions when congestion is detected



When a packet flooding strategy like the one described above is used to deter- mine routing, it is very likely that due to the large number of broadcast packets in circulation, some nodes may not receive the packet in one go due to interference. Hence it becomes essential to re broadcast packets for a certain period of time to ensure with a probability that every node has received a routing packet atleast once. To ensure this the base station keeps sending out these broadcast messages for a fixed period of time (around 10 seconds) with a random back-off. Every node will thus receive much more than just a single routing packet and it must make a decision about which of these packets it will further transmit as otherwise there will be excessive packet flooding in the network and functionality may get severely affected. To counter this, a simple heuristic was adopted. A node re-broadcasts only those packets which lead to an updation of its Parent and/or Hop Distance. Time Synchronisation The routing messages that flood the network also carry time stamps. Whenever a node receives a message and decides to change its Parent and Hop Distance, it also updates its clock. In such a manner time synchronisation may achieved in the network. There are however a couple of issues with this strategy. The first problem is that of the clocks of adjacent nodes in the net work being slightly out of phase . This will happen due to the delay introduced due to message radio transmission, reception and processing. This delay although not the same in every case will roughly be of the same order of magnitude. Therefore in an n-hop network, the worst case delay would be × n. This delay due to phase shifting is almost unavoidable, unless some

correction factors are introduced in th processing of received data. The strategy we use to counter this phase difference is that although we increment the node clocks every 128 1024 milliseconds; the basic unit of time is 1 second. Thus, some of this phase difference is just ironed over. However, if n were to become fairly large, this might pose a problem. The second problem is that of clock skew. This is basically the difference in clock times of two nodes that might occur even though they may have started at the same time value. In case of the hardware used clock skew was of the order of 1 millisecond in every 50,000 milliseconds. Clock skew can be completely taken care of by periodically putting the system into the routing phase; thus re-synchronising clocks. Adding and Deleting nodes. The addition and deletion of nodes can be done as long as re- routing is done. Hence for easy addition and deletion of nodes the routing stage needs to be initiated repeatedly. 23


Wireless sensor networks communicate within themselves as well as with a

user not

necessarily near the network¶s location. How does this work? Wireless sensor networks collect data about what is happening, and perform some action according to that data, be it moving, setting off alarms, or simply recording the data. All of these actions change the world that the node is in, causing other changes, and so on. Because of the connection between the nodes, all of these changes affect each node, and all of the data collected by the node is routed to the parent node. This parent node is connected to a computer of higher power that performs a function for which the nodes are not designed. One such function is to access the internet and transfer the nodes¶ data to the user¶s computer. The user may also communicate with the nodes. If the user gives some directives, the directives will be sent over the internet to the computer/station. The computer/station will communicate the same directives to the parent node, which then disperses the message amongst its ³children´


Wireless sensor networks are ideally suited as a foundation for smart healthcare in AlarmNet, due to several inherent qualities:

1.Portability and unobtrusiveness.
Small devices collect data and communicate wirelessly, operating with minimal patient input. They may be carried on the body or deeply embedded in the environment. Unobtrusiveness helps with patient acceptance and minimizes confounding measurement effects. Since monitoring is done in the living space, the patient travels less often, which is safer and more convenient.

2.Ease of deployment and scalability.
Devices can be deployed in potentially large quantities with dramatically less complexity and cost compared to wired networks. Existing structures, particularly dilapidated ones, can be easily augmented with a WSN network whereas wired installations would be expensive and impractical. Devices are placed in the living space and turned on, self-organizing and calibrating automatically.

3.Real-time and always-on.
Physiological and environmental data can be monitored continuously, allowing real-time response by emergency or healthcare workers. The data collected form a health journal, and are valuable for filling in gaps in the traditional patient history. Even though the network as a whole is always-on, individual sensors still must conserve energy through smart power management.

4.Reconfigurability and self-organization.
Since there is no fixed installation, adding and removing sensors instantly reconfigures the network. Doctors may re-target the mission of the network as medical needs change. Sensors self-organize to form routing paths, collaborate on data processing, and establish hierarchies. The system benefits both the healthcare providers and their patients. For the providers, an 25

automatic monitoring system is valuable for many reasons. Firstly, it frees humans from 24/7 physical monitoring, reducing labor costs and increasing efficiency. Secondly, wearable sensor devices can sense even small changes in vital signals that humans might overlook, for example, heart rate and blood oxygen levels, boosting accuracy. Quickly notifying of these changes may save human lives. Thirdly, the data collected from the wireless sensor network can be stored and integrated into a comprehensive health record of each patient, which helps physicians make more informed diagnoses. Eventually, the analyzing, diagnosis, treatment process may also be semi-automated, so a human physician can be assisted by an electronic physician. Healthcare patients benefit from improved health as a result of faster diagnosis and treatment of diseases. Other quality-of-life issues, such as privacy, dignity, and convenience, are supported and enhanced by the ability to provide services in an environment more comfortable for the patient. Though 24/7 physical presence of caregivers is reduced, the patient is not isolated from contact with the outside world²an important component of mental health. Family members and the system itself become part of the healthcare team. Finally, memory aids and other patient-assistance services can restore some lost independence, while preserving safety.


The development in the communication systems and the networking has given rise to the wireless networks. The ease and flexibility of wireless communication has enabled us to use personal assistance devices to be used anywhere. This has enabled the mankind to excel in every field of the life, but at the same time it has many threats as well.

1.Security Threats to Wireless Networks
Besides all the comforts of the life wireless networks poses serious security threats. The main reason is the signals are spread in the air and it is convenient for the hackers to catch wireless signals. Wireless networks require very tight security so that the unauthorized users cannot exploit the information. As more and users are making use of the wireless technologies, the risk of data being lost is increasing. The current wireless encryption protocols are difficult to handle .moreover the common users are not aware of the ways of addressing wireless security matters. One must build strong security protocols in order to secure the wireless signals like WPA and WPA2. Moreover wireless intrusion prevention system is another better way to build strong security system.

Wireless is a public frequency network therefore its interface is highly risky to be used for official private information. The speed and the viability of the wireless signals drop as more and more users use the same frequency. Moreover its original throughput is three times less than it claims to deliver. Wireless technology is available in only three major channels ranging from 2.5 GHz, 11 and 1.6 GHz, Which is much lower than the wired network. They are 50 times slower than the wired network technologies. The wireless signals are also attuned by the barriers such as walls, doors and devices itself. The setup of the wireless technology is difficult maintain. Hence it an unstable network. Though wireless technologies provide flexibility to use and carry your laptop or any other portable device with you anywhere, but the longer the distance the weaker the signals. The extensive use of wireless signals over the mobile phone is dangerous to the health of the human beings. The various health problems that wireless can pose are memory loss and even cancer. Wireless signals are prone to disrupt by the infrared and radio signals. Wireless technologies are four times more expensive than wired technology due to their difficult setup. If one needs to transfer confidential data over the


network wireless technology is a serious risk to be used. Banks, investigation agencies and legal data should be transferred using wired network technologies, because they are more safe and sound. The wireless technology offers public access points which hinder the efficient transfer of data.

3.Unauthorized Access to Wireless Networks .
The unauthorized access to the wireless signals is really common. The various types of unauthorized access are malicious association, accidental association, ad hoc networks, and non traditional networks, man in the middle attack, identity theft and denial of service. When a user turn on its PC, and he or she receives unauthorized signals from a neighboring building. User might not even signals are emerging and make use of it are have a clue from where they are emerging and make use of it , then we can call it accidental association. When a person uses crackers to access the password of the wireless network it is termed as malicious network. When the wireless technology has no centre device to act for it encryption settings are hard to mange therefore security can not be achieved. Bluetooth devices and PDAs are non traditional form of networks; they can be easily hacked by using cracker technology.



1. Military Applications.
Because most of the elemental knowledge of sensor networks is basic on the defense application at the beginning, especially two important programs the Distributed Sensor Networks (DSN) and the Sensor Information Technology (SenIT) form the Defense Advanced Research Project Agency (DARPA), sensor networks are applied very successfully in the military sensing. Now wireless sensor networks can be an integral part of military command, control, communications, computing, intelligence, surveillance, reconnaissance and targeting systems. In the battlefield context, rapid deployment, self-organization, fault tolerance security of the network should be required. The sensor devices or nodes should provide following services: 

Monitoring friendly forces, equipment and ammunition 

Battlefield surveillance  Reconnaissance of opposing forces  Targeting  Battle damage assessment  Nuclear, biological and chemical attack detection reconnaissance

2. Environmental Applications .
Nowadays sensor networks are also widely applied in habitat monitoring, agriculture research, fire detection and traffic control.Because there is no interruption to the environment, sensor networks in environmental area is not that strict as in battlefield. Bush Fire Response: A low cost distributed sensor network for environmental monitor and disaster response. An integrated network of sensors combining on the ground sensors monitoring local moisture levels, humidity, wind speed and direction, together with satellite imagery and longer term meteorological forecasting will enable the determination of fire risk levels in targeted regions as well as valuable information on probable fire direction. Such a network will provide valuable understanding of bushfire development and most importantly assist authorities in organizing a coordinated disaster response that will save lives and 29

property by providing early warning for high risk areas.

3. Health Applications.
Sensor networks are also widely used in health care area. In some modern hospital sensor networks are constructed to monitor patient physiological data, to control the drug administration track and monitor patients and doctors and inside a hospital. In spring 2004 some hospital in Taiwan even use RFID basic of above named applications to get the situation at first hand. Long-term nursing home: this application is focus on nursing of old people. In the town farm cameras, pressure sensors, orientation sensors and sensors for detection of muscle activity construct a complex network. They support fall detection, unconsciousness detection, vital sign monitoring and dietary/exercise monitoring. These applications reduce personnel cost and rapid the reaction of emergence situation.

4. Home Application
Along with developing commercial application of sensor network it is no so hard to image that Home application will step into our normal life in the future. Many concepts are already designed by researcher and architects, like ³Smart Environment: Residential Laboratory´ and ³Smart Kindergarten´



WSNs are still uncommon. This is still a young technology, allowing WSNs great growth potential. It is this potential that captures the attention of those who interact with WSNs. The question that is constantly asked is ³what new use can we come up with for this?´ There have been many answers to this question, including: data collection, strategic mine placement, machinery monitoring, and much more. The way that Nodes are made is a huge part of the success of WSNs in such different situations. Another reason they are successful is the way that the nodes work together to form a network. Using MEMS technology, nodes have been becoming smaller and more efficient. The uses and capabilities of WSNs are varied, and the only way to fully understand the extent to which they can be adapted will be to wait and see what happens. The future of WSNs is bright, as increasing attention is brought to their uses. Who knows, WSNs may become as commonplace as the PC.