A Review Routing Techniques in Wireless Sensor Networks

8/12/2019 A Review Routing Techniques in Wireless Sensor Networks

1/23

A Review Routing Techniques in Wireless Sensor NetworksProf. Partha Ghosh Amit Jain

Dept. of Information TechnologyNetaji Subhash Engineering College

Garia, Kolkata:-700132, IndiaEmail: - {partha1812, itskingjain} @gmail.com

Abstract:

The rapid progress of wireless communication and the availability of many small-sized, light-weighted and low-cost communication and computing devices nowadays have greatly impactedthe development of wireless sensor network. Localization using sensor network has attractedmuch attention for its comparable low-cost and potential use with monitoring and targetingpurposes in real and hostile application scenarios. Most of the attention, however, has beengiven to the routing protocols since they might differ depending on the application and networkarchitecture. This paper surveys recent routing protocols for sensor networks and presents aclassification for the various approaches pursued. The three main categories explored in thispaper are data-centric, hierarchical and location-based. Each routing protocol is described anddiscussed under the appropriate category.

Introduction:AWireless Sensor Networkconsists of a number of sensors spread across a geographicalarea. Each sensor has wireless communication capability and some level of intelligence forsignal processing and networking of the data. These sensor nodes can sense, measure, andgather information from the environment and, based on some local decision process, they cantransmit the sensed data to the user. Smart sensor nodes are low power devices equipped withone or more sensors, a processor, memory, a power supply, a radio, and an actuator 1.

Available sensors in the market include generic (multi-purpose) nodes and gateway (bridge)nodes. A generic (multi-purpose) sensor nodes task is to take measurements from themonitored environment. It may be equipped with a variety of devices which can measure

various physical attributes such as light, temperature, humidity, barometric pressure, velocity,acceleration, acoustics, magnetic field, etc. Gateway (bridge) nodes gather data from genericsensors and relay them to the base station. Gateway nodes have higher processing capability,

battery power, and transmission (radio) range. A combination of generic and gateway nodes istypically deployed to form a WSN.1An actuator is an electro-mechanical device that can be used to control different components in a system. In a sensor node,actuators can actuate different sensing devices, adjust sensor parameters, move the sensor, or monitor power in the sensor node.

Sensors

Generic Gateway


2/23

Components of Sensor Node.

Architecture of Wireless Sensor Network:A WSN typically has little or no infrastructure. It consists of a number of sensor nodes (few tensto thousands) working together to monitor a region to obtain data about the environment.

There are two types of WSNs:

Unstructured Structured.

An unstructured WSN is one that contains a dense collection of sensor nodes. Sensor nodesmay be deployed in an ad hoc manner2into the field. Once deployed, the network is leftunattended to perform monitoring and reporting functions. In an unstructured WSN, networkmaintenance such as managing connectivity and detecting failures is difficult since there are somany nodes.In a structured WSN, all or some of the sensor nodes are deployed in a pre-planned manner 3.

The advantage of a structured network is that fewer nodes can be deployed with lower network

maintenance and management cost. Fewer nodes can be deployed now since nodes are placed atspecific locations to provide coverage while ad hoc deployment can have uncovered regions.

Routing in Wireless Sensor Networks:Routingis the process of selecting paths in a network along which to send network traffic. A

Wireless Sensor Network (WSN) contain hundreds or thousands of sensor nodes. These sensorshave the ability to communicate either among each other or directly to an external base-station(BS).

2 In ad hoc deployment, sensor nodes may be randomly placed into the field3 In pre-planned deployment, sensor nodes are pre-determined to be placed at fixed locations.


3/23

Challenges in WSN Routing:-

In WSN, the routing protocols are application specific, data centric, capable of aggregatingdata and capable of optimizing energy consumption. The important characteristics of a goodrouting protocol for WSN are simplicity, energy awareness, adaptability and scalability due tolimited energy supply, limited computation power, limited memory and limited bandwidth of

WSN. The main design goal of WSNs is to carry out data communication while trying toprolong the lifetime of the network .The design of routing protocol in WSNs is influenced

by many challenging factors as summarized below.

Node deployment: Node deployment in WSNs is application dependent and affectsthe performance of the routing protocol. The deployment is either deterministic(manual) or self-organizing (random). In deterministic situations, the sensors aremanually placed and data is routed through pre-determined paths. Whereas in self-organizing systems, the sensor nodes are scattered randomly creating an infrastructurein an ad hoc manner. The position of the sink or the cluster-head is very crucial in

terms of energy efficiency and performance. When the distribution of nodes is notuniform, optimal clustering becomes a necessity to enable energy efficient networkoperation. In some applications like battle field and wildlife monitoring, sensor nodesare randomly deployed like being dropped from an airplane.

Network dynamics: Most of the network architectures assume that sensor nodes arestationary, because there are very few setups that utilize mobile sensors. It issometimes necessary to support the mobility of sinks or cluster-heads (gateways).Route stability becomes an important optimization factor, in addition to energy,

bandwidth etc. as communication from moving nodes is more challenging. Further,the sensed event can also be either dynamic or static depending on the application.

Energy Conservation: During the creation of an infrastructure, the process of settingup the routes is greatly influenced by energy considerations. Since the transmissionpower of a wireless radio is proportional to distance squared or even higher order inthe presence of obstacles, multi-hop routing will consume less energy than directcommunication. However, multi-hop routing introduces significant overhead fortopology management and medium access control. Direct routing would perform wellenough if all the nodes were very close to the sink. Most of the time sensors arescattered randomly over an area of interest and multi-hop routing becomesunavoidable.

Fault Tolerance: If sensor nodes fail, MAC and routing protocols mustaccommodate formation of new links so that sensor node failure should not affect theoverall task of the sensor network.

Scalability:The number of sensor node in the target area may be on the order ofhundreds or thousands, or more so protocols should be able to scale to such high


4/23

degree and take advantage of the high density of such networks.

Production Costs:The cost of a single node must be low.

Hardware Constraint:All Subunits of sensor node (e.g. sensing, processing,communication, power, location finding system and mobilizer) must consume

extremely low power and be contained within an extremely small volume.

Sensor network topology: It must be maintained even with very high node density

Environment: Nodes should be operating in inaccessible location because of hostileenvironment.

Transmission Media: Generally, Transmission Media is wireless (RF or Infrared),which is affected by fading and high error rate and affect the operation of WSNs.

Data delivery models: Data delivery model to the sink can be continuous, eventdriven, query-driven and hybrid, depending on the application of the sensor network.In the continuous delivery model, each sensor sends data periodically. In event-drivenand query-driven models, the transmission of data is triggered when an event occurs

or the sink generates a query. Some networks apply a hybrid model using acombination of continuous, event-driven and query-driven data delivery. The routingprotocol is highly influenced by the data delivery model, especially with regard to theminimization of energy consumption and route stability.

Node capabilities: In a sensor network, different functionalities can be associatedwith the sensor nodes. Depending on the application a node can be dedicated to aparticular special function such as relaying, sensing and aggregation since engaging thethree functionalities at the same time on a node might quickly drain the energy of that

node.

Data aggregation/fusion: Data aggregation is the combination of data fromdifferent sources by using functions such as suppression (eliminating duplicates), min,max and average. Similar packets from multiple nodes can be aggregated to reduce thetransmission.

Classification of Routing Technics in WSN:-In general, routing in WSNs can be divided into Flat-Based Routing, Hierarchical-BasedRouting, and Location-Based Routingdepending on the network structure.In flat-based routing, all nodes are typically assigned equal roles or functionality.


5/23

In hierarchical-based routing, however, nodes will play different roles in the network.In location-based routing, sensor nodes' positions are exploited to route data in the network. Arouting protocol is considered adaptive if certain system parameters can be controlled in orderto adapt to the current network conditions and available energy levels. Furthermore, theseprotocols can be classified into multipath-based, query-based, negotiation-based, QoS-based, orcoherent-based routing techniques depending on the protocol operation. In addition to theabove, routing protocols can be classified into three categories, namely, proactive, reactive, andhybrid protocols depending on how the source finds a route to the destination. In proactive

protocols, all routes are computed before they are really needed, while in reactive protocols,routes are computed on demand. Hybrid protocols use a combination of these two ideas. Whensensor nodes are static, it is preferable to have table driven routing protocols rather than usingreactive protocols. A significant amount of energy is used in route discovery and setup ofreactive protocols. Another class of routing protocols is called the cooperative routingprotocols. In cooperative routing, nodes send data to a central node where data can beaggregated and may be subject to further processing, hence reducing route cost in terms ofenergy use. Many other protocols rely on timing and position information. We also shed somelight on these types of protocols in this paper. In order to streamline this survey, we use a

classification according to the network structure and protocol operation (routing criteria).


6/23

WSN Routing

Network BasedRouting

Flat NetworkRouting

Flooding &Gossiping

SPIN

Directed Diffusion

Rumor Routing

Energy AwareRouting

MCFA

COUGAR

ACQUIRE

HierarchicalNetwork Routing

LEACH

PEGASIS

TEEN &APTEEN

SOP

Location BasedRouting

GAF

GEAR

MECN

SMECN

Protocol BasedRouting

Multi-Path Based

Negotiation Based

Query Based

QoS Based

Coherent &

Non-CoherentBased


7/23

1. Network Structure Based Protocol: -The underlying network structure can playsignificant role in the operation of the routing protocol in WSNs. In this section, we surveyin details most of the protocols that fall below this category.1.1.Flat Network Routing:- The first category of routing protocols are the multihop flat

routing protocols. In flat networks, each node typically plays the same role and sensornodes collaborate together to perform the sensing task. Due to the large number of suchnodes, it is not feasible to assign a global identifier to each node. This consideration hasled to data centric routing, where the BS sends queries to certain regions and waits for

data from the sensors located in the selected regions. Since data is being requestedthrough queries, attribute-based naming is necessary to specify the properties of data.Early works on data centric routing, e.g., SPIN and directed diffusion were shown tosave energy through data negotiation and elimination of redundant data. These twoprotocols motivated the design of many other protocols which follow a similar concept.In the rest of this subsection, we summarize these protocols and highlight theiradvantages and their performance issues.

1.1.1. Flooding & Gossiping:- Flooding and gossipingare two classical mechanisms to relay data in sensor

networks without the need for any routing algorithmsand topology maintenance. In flooding, each sensorreceiving a data packet broadcasts it to all of itsneighbors and this process continues until the packetarrives at the destination or the maximum number ofhops for the packet is reached. On the other hand,gossiping is a slightly enhanced version of flooding

where the receiving node sends the packet to arandomly selected neighbor, which picks anotherrandom neighbor to forward the packet to and so on.

Although flooding is very easy to implement, it hasseveral drawbacks.Such drawbacks (shown in thefigure) include implosion caused by duplicatedmessages sent to same node (Node A starts byflooding its data to all of its neighbors. D gets twosame copies of data eventually, which is notnecessary.), overlap when two nodes sensing thesame region send similar packets to the same neighbor and resource blindness byconsuming large amount of energy without consideration for the energy

constraints(Two sensors cover an overlapping geographic region and C gets samecopy of data form these sensors.). Gossiping avoids the problem of implosion by justselecting a random node to send the packet rather than broadcasting. However, thiscause delays in propagation of data through the nodes.

1.1.2. SPIN:-Sensor Protocol for Information via Negotiation (SPIN) protocol wasdesigned to improve classic flooding protocols and overcome the problems they maycause, for example, implosion and overlap. The SPIN protocols are resource awareand resource adaptive. The sensors running the SPIN protocols are able to computethe energy consumption required to compute, send, and receive data over thenetwork. Thus, they can make informed decisions for efficient use of their ownresources. The SPIN protocols are based on two key mechanisms namely negotiationand resource adaptation. SPIN enables the sensors to negotiate with each other


8/23

before any data dissemination can occur in order to avoid injecting non-useful andredundant information in the network. SPIN uses meta-data as the descriptors of thedata that the sensors want to disseminate. The notion of meta-data avoids theoccurrence of overlap given sensors can name the interesting portion of the datathey want to get. It may be noted here that the size of the meta-data should definitelybe less than that of the corresponding sensor data. Contrary to the floodingtechnique, each sensor is aware of its resource consumption with the help of its ownresource manager that is probed by the application before any data processing or

transmission. This helps the sensors to monitor and adapt to any change in their ownresources.

There are two protocols in the SPIN family: SPIN-l (or SPIN-PP) and SPIN-2 (orSPIN-EC). While SPIN-l uses a negotiation mechanism to reduce the consumptionof the sensors, SPIN-2 uses a resource-aware mechanism for energy savings. Bothprotocols allow the sensors to exchange information about their sensed data, thushelping them to obtain the data they are interested in. SPIN-l is a three-stagehandshake protocol by which the sensors can disseminate their data. This protocolapplies for those networks using point-to-point transmission media (or point-to-point networks), in which two sensors can communicate exclusively with each other

without interfering with other sensors. SPIN-BC improves SPIN-PP by using one-to-many communication instead of many one-to-one communications. It is a three-stage handshake protocol for broadcast transmission media, where the sensors in anetwork communicate with each other using a single shared channel. SPIN-2 differsfrom SPIN-l in that it takes into account the residual energy of sensors. If thesensors have plenty of energy, SPIN-2 is identical to SPIN-l, and hence has the same

three stages. However, when a sensor has low residual energy, it controls itsparticipation in a data dissemination process. While the family of SPIN protocolsapplies to lossless networks, it can be slightly updated to apply to lossy or mobilenetworks.

1.1.3. Directed Diffusion (DD):-Directed diffusion is a data-centric routing protocolfor sensor query dissemination and processing. It meets the main requirements of

WSNs such as energy efficiency, scalability, and robustness. Directed diffusion hasseveral key elements namely data naming, interests and gradients, data propagation,

and reinforcement. A sensing task can be described by a list of attribute-value pairs.At the beginning of the directed diffusion process, the sink specifies a low data ratefor incoming events. After that, the sink can reinforce one particular sensor to sendevents with a higher data rate by resending the original interest message with a


9/23

smaller interval. Likewise, if a neighboring sensorreceives this interest message and finds that thesender's interest has a higher data rate thanbefore, and this data rate is higher than that ofany existing gradient, it will reinforce one ormore of its neighbors.Directed Diffusion differs from SPIN in termsof the on demand data querying mechanism it

has. In Directed Diffusion the sink queries thesensor nodes if a specific data is available byflooding some tasks. In SPIN, sensors advertisethe availability of data allowing interested nodesto query that data. Directed Diffusion has manyadvantages. Since it is data centric, allcommunication is neighbor-to-neighbor with noneed for a node addressing mechanism. Eachnode can do aggregation and caching, in addition

to sensing. Caching is a big advantage in terms ofenergy efficiency and delay. In addition, DirectDiffusion is highly energy efficient since it is ondemand and there is no need for maintainingglobal network topology. However, DirectedDiffusion cannot be applied to all sensornetwork applications since it is based on a query-driven data delivery model. The applications thatrequire continuous data delivery to the sink willnot work efficiently with a Sink query-driven ondemand data model Therefore, DirectedDiffusion is not a good choice as a routingprotocol for the applications such asenvironmental monitoring.

1.1.4. Rumor Routing (RR):-Rumor routing is a logical compromise between queryflooding and event flooding app schemes. Rumor routing is an efficient protocol ifthe number of queries is between the two intersection points of the curve of rumor

routing with those of query flooding and event flooding. Rumor routing is based onthe concept of agent, which is a long-lived packet that traverses a network andinforms each sensor it encounters about the events that it has learned during itsnetwork traverse. An agent will travel the network for a certain number of hops andthen die. Each sensor, including the agent, maintains an event list that has event-distance pairs, where every entry in the list contains the event and the actual distancein the number of hops to that event from the currently visited sensor. Therefore,

when the agent encounters a sensor on its path, it synchronizes its event list with thatof the sensor it has encountered. Also, the sensors that hear the agent update theirevent lists according to that of the agent in order to maintain the shortest paths to


10/23

the events that occur in the network.

1.1.5. Energy Aware Routing: -The objective of energy-aware routing protocol, a

destination initiated reactive protocol, is to increase the network lifetime. Althoughthis protocol is similar to directed diffusion, it differs in the sense that it maintains aset of paths instead of maintaining or enforcing one optimal path at higher rates.

These paths are maintained and chosen by means of a certain probability. The valueof this probability depends on how low the energy consumption of each path can beachieved. By having paths chosen at different times, the energy of any single path

will not deplete quickly. This can achieve longer network lifetime as energy isdissipated more equally among all nodes. Network survivability is the main metric ofthis protocol. The protocol assumes that each node is addressable through a class-based addressing which includes the location and types of the nodes. The protocolinitiates a connection through localized flooding, which is used to discover all routesbetween source/destination pair and their costs; thus building up the routing tables.

The high-cost paths are discarded and a forwarding table is built by choosingneighboring nodes in a manner that is proportional to their cost. Then, forwardingtables are used to send data to the destination with a probability that is inversely

proportional to the node cost. Localized flooding is performed by the destinationnode to keep the paths alive. When compared to directed diffusion, this protocolprovides an overall improvement of 21.5% energy saving and a 44% increase innetwork lifetime. However, the approach requires gathering the location informationand setting up the addressing mechanism for the nodes, which complicate routesetup compared to the directed diffusion.

1.1.6. COUGAR: -The cougar routing protocol is a database approach to tasking sensornetworks. The Cougar approach provides a user and application programs withdeclarative queries of the sensed data generated by the source sensors. These queriesare suitable for WSNs in that they abstract the user from knowing the execution planof its queries. In other words, the user does not know which sensors are contacted,


11/23

how sensed data are processed to compute the queries, and how final results are sentto the user. The Cougar approach uses a query layer where every sensor is associated

with a query proxy that lies between the network layer and application layer of thesensor. This query proxy provides higher level services through queries that can beissued from a gateway node. Furthermore, the Cougar approach employs in-networkprocessing to reduce the total energy consumption and enhance the network lifetime..Cougar is more beneficial if a set of sensed data could be aggregated or fused into asingle one that is more representative and thus significant to the user. The cougar

being database approach, it faces few challenges. A network can be viewed as a hugedistributed database stem, where every sensor possesses a subset of data. Hence,current distributed management approaches cannot be applied directly, but need tobe modified accordingly.

1.1.7.AQUIRE: -ACtive QUery forwarding In sensoR nEtworks (ACQUIRE) isanother data-centric querying mechanism used for querying named data. It providessuperior query optimization to answer specific types of queries, called one-shot

complex queries for replicated data. ACQUIRE query (i.e., interest for named data)consists of several sub queries for which several simple responses are provided byseveral relevant sensors. Each sub-query is answered based on the currently storeddata at its relevant sensor. ACQUIRE allows a sensor to inject an active query in anetwork following either a random or a specified trajectory until the query getsanswered by some sensors on the path using a localized update mechanism. Unlikeother query techniques, ACQUIRE allows the querier to inject a complex query intothe network to be forwarded stepwise through a sequence of sensors.

1.1.8. MCFA: - The Minimum Cost Finding Algorithm (MCFA) algorithm exploits thefact that the direction of routing is always known, that is, towards the fixed externalbase-station. Hence, a sensor node need not have a unique ID nor maintain a routingtable. Instead, each node maintains the least cost estimate from itself to the base-station. Each message to be forwarded by the sensor node is broadcast to itsneighbors. When a node receives the message, it checks if it is on the least cost pathbetween the source sensor node and the base-station. If this is the case, it re-broadcasts the message to its neighbors. This process repeats until the base-station isreached. In MCFA, each node should know the least cost path estimate from itself to

the base-station. This is obtained as follows. The base-station broadcasts a messagewith the cost set to zero while every node initially set its least cost to the base-stationto infinity (). Each node, upon receiving the broadcast message originated at thebase-station, checks to see if the estimate in the message plus the link on which it isreceived is less than the current estimate. If yes, the current estimate and the estimatein the broadcast message are updated. If the received broadcast message is updated,then it is re-sent; otherwise, it is purged and nothing further is done. However, theprevious procedure may result in some nodes having multiple updates and thosenodes far away from the base-station will get more updates from those closer to thebase-station. To avoid this, the MCFA was modified to run a backoff algorithm atthe setup phase. The backoff algorithm dictates that a node will not send the updatedmessage until a*lctime units have elapsed from the time at which the message isupdated, where ais a constant and lcis the link cost from which the message was


12/23

received.

1.2.Hierarchical Network Routing: -Hierarchical or cluster-based routing, originallyproposed in wire line networks, are well-known techniques with special advantagesrelated to scalability and efficient communication. As such, the concept of hierarchicalrouting is also utilized to perform energy-efficient routing in WSNs. In a hierarchicalarchitecture, higher energy nodes can be used to process and send the information while

low energy nodes can be used to perform the sensing in the proximity of the target. Thismeans that creation of clusters and assigning special tasks to cluster heads can greatlycontribute to overall system scalability, lifetime, and energy efficiency. Hierarchicalrouting is an efficient way to lower energy consumption within a cluster and byperforming data aggregation and fusion in order to decrease the number of transmittedmessages to the BS. Hierarchical routing is mainly two-layer routing where one layer isused to select clusterheads and the other layer is used for routing. However, mosttechniques in this category are not about routing, rather on "who and when to send orprocess/aggregate" the information, channel allocation etc., which can be orthogonal to

the multihop routing function.

1.2.1. LEACH: - Low-EnergyAdaptive Clustering Hierarchy (LEACH) is the first andmost popular energy-efficient hierarchical clustering algorithm for WSNs that wasproposed for reducing power consumption. In LEACH, the clustering task is rotatedamong the nodes, based on duration. Direct communication is used by each clusterhead (CH) to forward the data to the base station (BS). It uses clusters to prolong thelife of the wireless sensor network. LEACH is based on an aggregation (or fusion)technique that combines or aggregates the original data into a smaller size of datathat carry only meaningful information to all individual sensors. LEACH divides thea network into several cluster of sensors, which are constructed by using localizedcoordination and control not only to reduce the amount of data that are transmittedto the sink, but also to make routing and data dissemination more scalable androbust. LEACH uses a randomize rotation of high-energy CH position rather thanselecting in static manner, to give a chance to all sensors to act as CHs and avoid thebattery depletion of an individual sensor and dieing quickly.The operation of LEACH is divided into rounds having two phases each namely (i)a setup phase to organize the network into clusters, CH advertisement, andtransmission

schedulecreation and(ii) a steady-state phasefor dataaggregation,compression,andtransmissionto the sink.LEACH iscompletelydistributed


13/23

and requires no global knowledge of network. The duration of the steady state phaseis longer than the duration of the setup phase in order to minimize the overhead.During the setup phase, a predetermined fraction of nodes, p, elect themselves asCHs as follows.

A sensor node chooses a random number, v, between 0 and 1. If this randomnumber is less than a threshold value, T(n), the node becomes a cluster-head for thecurrent round. The threshold value is calculated based on an equation thatincorporates the desired percentage to become a cluster-head in the current round

from the set of nodes that have not been selected as a cluster-in the last (1/P)rounds. The threshold value is given by:

Each elected CH broadcasts an advertisement message to the rest of the nodes in thenetwork that they are the new cluster-heads. All the non-cluster head nodes, afterreceiving this advertisement, decide on the cluster to which they want to belong to.

This decision is taken based on the signal strength of the advertisement. Thenoncluster- head nodes inform the appropriate clusterheads that they will be amember of the cluster. After receiving all the messages from the nodes that wouldlike to be included in the cluster and based on the number of nodes in the cluster,

the cluster-head node creates a TDMA (i.e., Time Division Multiple Access) scheduleand assigns each node a time slot when it can transmit. This schedule is broadcast toall the nodes in the cluster. It reduces energy consumption by (a) minimizing thecommunication cost between sensors and their cluster heads and (b) turning off non-head nodes as much as possible. LEACH uses single-hop routing where each nodecan transmit directly to the cluster-head and the sink. Therefore, it is not applicableto networks deployed in large regions. Furthermore, the idea of dynamic clusteringbrings extra overhead, e.g. head changes, advertisements etc., which may diminishthe gain in energy consumption. While LEACH helps the sensors within their cluster

dissipate their energy slowly, the CHs consume a larger amount of energy when theyare located farther away from the sink. Also, LEACH clustering terminates in a finitenumber of iterations, but does not guarantee good CH distribution and assumesuniform energy consumption for CHs.

1.2.2. PEGASIS: -Power-Efficient Gathering in Sensor Information Systems(PEGASIS) is an extension of the LEACH protocol, which forms chains fromsensor nodes so that each node transmits and receives from a neighbor and only onenode is selected from that chain to transmit to the base station (sink). The data isgathered and moves from node to node, aggregated and eventually sent to the base

station. The chain construction is performedin a greedy way. Unlike LEACH, PEGASISavoids cluster formation and uses only one


14/23

node in a chain to transmit to the BS (sink)instead of using multiple nodes. A sensortransmits to its local neighbors in the data fusionphase instead of sending directly to its CH as inthe case of LEACH. In PEGASIS routingprotocol, the construction phase assumes that allthe sensors have global knowledge about thenetwork, particularly, the positions of the

sensors, and use a greedy approach. When asensor fails or dies due to low battery power, thechain is constructed using the same greedy approach by bypassing the failed sensor.In each round, a randomly chosen sensor node from the chain will transmit theaggregated data to the BS, thus reducing the per round energy expenditure comparedto LEACH.Simulation results showed that PEGASIS is able to increase the lifetime of thenetwork twice as much the lifetime of the network under the LEACH protocol. Suchperformance gain is achieved through the elimination of the overhead caused by

dynamic cluster formation in LEACH and through decreasing the number oftransmissions and reception by using data aggregation. Although the clusteringoverhead is avoided, PEGASIS still requires dynamic topology adjustment since asensor node needs to know about energy status of its neighbors in order to know

where to route its data. Such topology adjustment can introduce significant overheadespecially for highly utilized networks.

1.2.3.TEEN & APTEEN: - In Threshold-sensitive Energy Efficient Protocols(TEEN), sensor nodes sense the medium continuously, but the data transmission isdone less frequently. A cluster head sensor sends its members a hard threshold,

which is the threshold value of the sensed attribute and a soft threshold, which is asmall change in the value of the sensed attribute that triggers the node to switch onits transmitter and transmit. Thus the hard threshold tries to reduce the number oftransmissions by allowing the nodes to transmit only when the sensed attribute is inthe range of interest. The soft threshold further reduces the number of transmissionsthat might have otherwise occurred when there is little or no change in the sensedattribute. A smaller value of the soft threshold gives a more accurate picture of thenetwork, at the expense of increased energy consumption. Thus, the user can controlthe trade-off between energy efficiency and data accuracy. When cluster-heads are to

change, new values for the above parameters are broadcast. The main drawback ofthis scheme is that, if the thresholds are not received, the nodes will nevercommunicate, and the user will not get any data from the network at all. The nodessense their environment continuously. The first time a parameter from the attributeset reaches its hard threshold value, the node switches its transmitter on and sendsthe sensed data. The sensed value is stored in an internal variable, called Sensed

Value (SV). The nodes will transmit data in the current cluster period only when thefollowing conditions are true: (1) The current value of the sensed attribute is greaterthan the hard threshold (2) The current value of the sensed attribute differs from SVby an amount equal to or greater than the soft threshold. Important features of

TEEN include its suitability for time critical sensing applications. Also, sincemessage transmission consumes more energy than data sensing, so the energyconsumption in this scheme is less than the proactive networks. The soft threshold


15/23

can be varied. At every cluster change time, a fresh parameters are broadcast and so,the user can change them as required.

In Adaptive Periodic Threshold Sensitive Energy Efficient Sensor Network

Protocol(APTEEN), the cluster-heads broadcasts the following parameters (see

above figure (b)):

1. Attributes (A): this is a set of physical parameters which the user is interested in

obtaining information about.

2. Thresholds: this parameter consists of the Hard Threshold (HT) and the SoftThreshold (ST).

3. Schedule: this is a TDMA schedule, assigning a slot to each node.

4. Count Time (CT): it is the maximum time period between two successive

reports sent by a node.The node senses the environment continuously, and only those nodes which sense a

data value at or beyond the hard threshold transmit. Once a node senses a valuebeyond HT, it transmits data only when the value of that attribute changes by an

amount equal to or greater than the ST. If a node does not send data for a time

period equal to the count time, it is forced to sense and retransmit the data. ATDMA schedule is used and each node in the cluster is assigned a transmission slot.

Hence, APTEEN uses a modified TDMA schedule to implement the hybrid

network. The main features of the APTEEN scheme include the following. Itcombines both proactive and reactive policies. It offers a lot of flexibility by allowing

the user to set the count-time interval (CT), and the threshold values for the energyconsumption can be controlled by changing the count time as well as the threshold

values. The main drawback of the scheme is the additional complexity required to

implement the threshold functions and the count time. Simulation of TEEN and

APTEEN has shown that these two protocols outperform LEACH. The

experiments have demonstrated that APTEENs performance is somewhere between

LEACH and TEEN in terms of energy dissipation and network lifetime. TEEN

gives the best performance since it decreases the number of transmissions. The maindrawbacks of the two approaches are the overhead and complexity associated with


16/23

forming clusters at multiple levels, the method of implementing threshold-based

functions, and how to deal with attribute-based naming of queries.

1.2.4 Self-Organizing Protocol (SOP):- Self-organizing protocol (SOP) is

heterogeneity based routing protocol. In this approach, some sensors sense theenvironment and forward the data to a designated set of nodes that act as routers.

Router nodes are stationary and form a backbone for communication. Collected

data are forwarded through the routers tithe more powerful BS nodes. Sensing

nodes can be identified through the address of the router node they are connectedto. The routing architecture is hierarchical where groups of nodes are formed andmerged when needed. Local Markov Loops (LML) algorithm, which performs a

random walk on spanning trees of a graph, is used to support fault tolerance and as

a medium for broadcasting. Here sensor nodes can be addressed individually, andhence it is suitable for applications where communication to a particular node is

required. The algorithm for self-organizing the router nodes and creating the

routing tables consists of four phases: Discovery Phase: The nodes in the neighborhood of each sensor are discovered.

Organization phase: Groups are formed and merged by forming a hierarchy.Each node is allocated an address based on its position in the hierarchy. Routing

tables of size O(log N) are created for each node. Broadcast trees that span all thenodes are constructed.

Maintenance phase: Updating of routing tables and energy levels of nodes is

made in this phase. Each node informs the neighbors about its routing table andenergy level. LML are used to maintain broadcast trees.

Self-reorganization phase: In case of partition or node failures, group

reorganizations are performed.

The proposed algorithm utilizes the router nodes to keep all the sensors

connected by forming a dominating set. The major advantage of using thealgorithm is the small cost of maintaining routing tables and keeping routing

balanced. The disadvantage is in the organization phase of algorithm, which is not

on-demand. Furthermore, this algorithm incurs a small cost for maintainingrouting tables and maintaining a balanced routing hierarchy. Therefore, it may

cause extra overhead.

1.3 Location/Geographic Based Routing: -In location-based protocols, sensor nodes are

addressed by means of their locations. Location information for sensor nodes is required

for sensor networks by most of the routing protocols to calculate the distance between

two particular nodes so that energy consumption can be estimated. In this section, we

present a sample of location-aware routing protocols proposed for WSNs.

1.3.1 Geographic and Energy-Aware Routing (GEAR):GEAR uses energy aware andgeographically informed neighbor selection heuristics to route a packet towards the targetregion. The idea is to restrict the number of interests in Directed Diffusion by onlyconsidering a certain region rather than sending the interests to the whole network.


17/23

GEAR compliments Directed

Diffusion in this way and thusconserves more energy. In GEAR,

each node keeps an estimated cost

and a learning cost of reaching the

destination through its neighbors.

The estimated cost is a combination

of residual energy and distance todestination. The learned cost is a

refinement of the estimated costthat accounts for routing around

holes in the network. A hole occurs

when a node does not have any

closer neighbor to the target region

than itself. If there are no holes, the

estimated cost is equal to the

learned cost. The learned cost is propagated one hop back every time a packet reaches thedestination so that route setup for next packet will be adjusted. There are two phases in

the algorithm:

1. Forwarding packets towards the target region:Upon receiving a packet, a node

checks its neighbors to see if there is one neighbor, which is closer to the target region

than itself. If there is more than one, the nearest neighbor to the target region is selected

as the next hop. If they are all further than the node itself, this means there is a hole. Inthis case, one of the neighbors is picked to forward the packet based on the learning cost

function. This choice can then be updated according to the convergence of the learnedcost during the delivery of packets.

2. Forwarding the packets within the region: If the packet has reached the region, it

can be diffused in that region by either recursive geographic forwarding or restricted

flooding. Restricted flooding is good when the sensors are not densely deployed. In high-

density networks, recursive geographic flooding is more energy efficient than restricted

flooding. In that case, the region is divided into four sub regions and four copies of thepacket are created. This splitting and forwarding process continues until the regions with

only one node are left. An example is depicted in figure above.

1.3.2 Geographic Adaptive Fidelity (GAF):- GAF is an energy-aware routing protocol

primarily proposed for MANETs, but can also be used for WSNs because it favorsenergy conservation. The design of GAF is motivated based on an energy model that

considers energy consumption due to the reception and transmission of packets as well as

idle (or listening) time when the radio of a sensor is on to detect the presence ofincoming packets. GAF is based on mechanism of turning off unnecessary sensors while

keeping a constant level of routing fidelity (or uninterrupted connectivity between

communicating sensors). In GAF, sensor field is divided into grid squares and every

sensor uses its location information, which can be provided by GPS or other locationsystems, to associate itself with a particular grid in which it resides. This kind of

association is exploited by GAF to identify the sensors that are equivalent from the


18/23

perspective of packet forwarding. As

shown in below, the state transitiondiagram of GAF has three states,

namely, discovery, active, and sleeping.

When a sensor enters the sleeping state,

it turns off its radio for energy savings.

In the discovery state, a sensor

exchanges discovery messages to learnabout other sensors in the same grid.

Even in the active state, a sensorperiodically broadcasts its discovery

message to inform equivalent sensors

about its state. The time spent in each of these states can be tuned by the application

depending on several factors, such as its needs and sensor mobility. GAF aims to

maximize the network lifetime by reaching a state where each grid has only one active

sensor based on sensor ranking rules. The ranking of sensors is based on their residual

energy levels. Thus, a sensor with a higher rank will be able to handle routing within theircorresponding grids. For example, a sensor in the active state has a higher rank than a

sensor in the discovery state. A sensor with longer expected lifetime has a higher rank.

1.3.3 MECN: -Minimum Energy Communication Network (MECN) is a location-basedprotocol for achieving minimum energy for randomly deployed ad hoc networks, whichattempts to set up and maintain a minimum energy network with mobile sensors. It is

self-reconfiguring protocol that maintains network connectivity in spite of sensor

mobility. It computes an optimal spanning tree rooted at the sink, called minimum powertopology, which contains only the minimum power paths from ach sensor to the sink. It

is based on the positions of sensors on the plane and consists of two main phases,

namely, enclosure graph construction and cost distribution. For a stationary network, in

the first phase (enclosure graph construction), MECN constructs a sparse graph, called an

enclosure graph, based on the immediate locality of the sensors. An enclosure graph is a

directed graph that includes all the sensors as its vertex set and whose edge set is theunion of all edges between the sensors and the neighbors located in their enclosure

regions. In other words, a sensor will not consider the sensors located in its relay regionsas potential candidate forwarders of its sensed data to the sink. In the second phase (cost

distribution), non-optimal links of the enclosure graph are simply eliminated and the

resulting graph is a minimum power topology. This graph has a directed path from each

sensor to the sink and consumes the least total power among all graphs having directed

paths from each sensor to the sink. Each sensor broadcasts its cost to its neighbors,

where the cost of a node is the minimum power required for this sensor to establish adirected path to the sink.

While MECN is a self-reconfiguring protocol, and hence is fault tolerant (in the

case of mobile networks), it suffers from a severe battery depletion problem when appliedto static networks. MECN does not take into consideration the available energy at each

sensor, and hence the optimal cost links are static. In other words, a sensor will always


19/23

use the same neighbor to transmit or forward sensed data to the sink. For this reason, this

neighbor would die very quickly and the network thus becomes disconnected. To addressthis problem, the enclosure graph and thus the minimum power topology should be

dynamic based on the residual energy of the sensors.

1.3.4 SMECN: - Small Minimum-Energy Communication Network (SMECN) is a routing

protocol proposed to improve MECN, in which a minimal graph is characterized with

regard to the minimum energy property. This property implies that for any pair of sensors

in a graph associated with a network, there is a minimum energy-efficient path betweenthem; that is, a path that has the smallest cost in terms of energy consumption over allpossible paths between this pair of sensors. Their characterization of a graph with respect

to the minimum energy property is intuitive. In SMECN protocol, every sensor discovers

its immediate neighbors by broadcasting a neighbor discovery message using some initialpower that is updated incrementally. Specifically, the immediate neighbors of a given

sensor are computed analytically. Then, a sensor starts broadcasting a neighbor discovery

message with some initial power p and checks whether the theoretical set of immediateneighbors is a subset of the set of sensors that replied to that neighbor discovery message.

If this is the case, the sensor will use the corresponding power p to communicate with its

immediate neighbors. Otherwise, it increments p and rebroadcasts its neighbor discoverymessage.

2. Protocol Based Routing: -In this section, we review routing protocols that differentrouting functionality. It should be noted that some of these protocols may fall below one or

more of the above routing categories.

2.1.1. Multi-Path Based Routing Protocol: -Considering data transmission betweensource sensors and the sink, there are two routing paradigms: single-path routing and

multipath routing. In single-path routing, each source sensor sends its data to the

sink via the shortest path. In multipath routing, each source sensor finds the first kshortest paths to the sink and divides its load evenly among these paths. In this

section, we review a sample of multipath routing protocols for WSNs. Example: -

Directed Diffusion.2.1.1.1. Disjoint Paths: Sensor-disjoint multipath routing is a multipath protocol

that helps find a small number of alternate paths that have no sensor in

common with each other and with the primary path. In sensor-disjoint pathrouting, the primary path is best available whereas the alternate paths are less

desirable as they have longer latency. The disjoint makes those alternate pathsindependent of the primary path. Thus, if a failure occurs on the primary path, it

remains local and does not affect any of those alternate paths. The sink can

determine which of its neighbors can provide it with the highest quality datacharacterized by the lowest loss or lowest delay after the network has been

flooded with some low-rate samples. Although disjoint paths are more resilient

to sensor failures, they can be potentially longer than the primary path and thus

less energy efficient.


20/23

2.1.1.2. Braided Paths: Braided multipath is a partially disjoint path from primary

one after relaxing the disjointedness constraint. To construct the braidedmultipath, first primary path is computed. Then, for each node (or sensor) on

the primary path, the best path from a source sensor to the sink that does not

include that node is computed. Those best alternate paths are not necessarily

disjoint from the primary path and are called idealized braided multipath.

Moreover, the links of each of the alternate paths lie either on or geographically

close to the primary path. Therefore, the energy consumption on the primaryand alternate paths seems to be comparable as opposed to the scenario of

mutually ternate and primary paths. The braided multipath can also beconstructed in a localized manner in which case the sink sends out a primary-

path reinforcement to its first preferred neighbor and alternate-path

reinforcement to its second preferred neighbor.

2.1.1.3. N-to-1 Multipath Discovery: N-to-1 multipath discovery is based on the

simple flooding originated from the sink and is composed of two phases,

namely, branch aware flooding (or phase 1) and multipath extension of flooding(or phase 2). Both phases use the same routing messages whose format is given

by {mtype, mid, nid, bid, cst, path}, where mtype refers to the type of amessage. This multipath discovery protocol generates multiple node-disjoint

paths for every sensor. In multihop routing, an active per-hop packet salvagingstrategy can be adopted to handle sensor failures and enhance network

reliability.

2.1.2. Negotiation Based Routing: - In order to eliminate redundant datatransmissions, these use high level data descriptors through negotiation. Based on

the resources that are available to them, communication decisions are taken. The

motivation is that the use of flooding to disseminate data will produce implosion and

overlap between the sent data; hence nodes will receive duplicate copies of the samedata. This consumes more energy and more processing by sending the same data to

different sensor nodes. So, the main idea of negotiation based routing in WSNs is tosuppress duplicate information and prevent redundant data from being sent to the

next sensor node or the base-station by conducting a series of negotiation messagesbefore the real data transmission begins. Example:- SPIN

2.1.3. Query Based Routing: -The destination nodes propagate a query for data(sensing task) from a node through the network and a node having this data sendsback the data to the node that matches the query to the query that initiates. Usually

these queries are described in natural language, or in high-level query languages.

Example: - Rumor Routing.

2.1.4. QoS Based Routing: - In order to satisfy certain QoS (Quality of Service) metrics,e.g., delay, energy, bandwidth, etc. when delivering data to the Base Station, thenetwork has to balance between energy consumption and data quality.


21/23

2.1.4.1. SPEED: - SPEED is a QoS routing protocol for sensor networks that

provides soft real-time end-to-end guarantees. The protocol requires each nodeto maintain information about its neighbors and uses geographic forwarding to

find the paths. In addition, SPEED strive to ensure a certain speed for each

packet in the network so that each application can estimate the end-to-end delay

for the packets by dividing the distance to the sink by the speed of the packet

before making the admission decision. Moreover, SPEED can provide

congestion avoidance when the network is congested. The routing module inSPEED is called Stateless Geographic Non-Deterministic forwarding (SNFG)

and works with four other modules at the network layer. The beacon exchangemechanism collects information about the nodes and their location. Delay

estimation at each node is basically made by calculating the elapsed time when

an ACK is received from a neighbor as a response to a transmitted data packet.

By looking at the delay values, SNGF selects the node, which meets the speed

requirement. If it fails, the relay ratio of the node is checked, which is calculated

by looking at the miss ratios of the neighbors of a node (the nodes which could

not provide the desired speed) and is fed to the SNGF module. When comparedto Dynamic Source Routing (DSR) and Ad-hoc on-demand vector routing

(AODV), SPEED performs better in terms of end-to-end delay and miss ratio.Moreover, the total transmission energy is less due to the simplicity of the

routing algorithm, i.e. control packet overhead is less, and to the even trafficdistribution. Such load balancing is achieved through the SNGF mechanism of

dispersing packets into a large relay area. SPEED does not consider any further

energy metric in its routing protocol. Therefore, for more realistic understandingof SPEEDs energy consumption, there is a need for comparing it to a routing

protocol, which is energy-aware.

2.1.4.2. Sequential Assignment Routing (SAR): SAR is one of the first routing

protocols for WSNs that introduces the notion of QoS in the routing decisions.

It is a table-driven multi-path approach striving to achieve energy efficiency andfault tolerance. Routing decision in SAR is dependent on three factors: energy

resources, QoS on each path, and the priority level of each packet. The SARprotocol creates trees rooted at one-hop neighbors of the sink by taking QoS

metric, energy resource on each path and priority level of each packet intoconsideration. By using created trees, multiple paths from sink to sensors are

formed. One of these paths is selected according to the energy resources and

QoS on the path. Failure recovery is done by enforcing routing table consistency

between upstream and downstream nodes on each path. Any local failure causesan automatic path restoration procedure locally. The objective of SAR algorithm

is to minimize the average weighted QoS metric throughout the lifetime of the

network. If topology changes due to node failures, a path re-computation is

needed. As a preventive measure, a periodic re-computation of paths is triggered

by the base-station to account for any changes in the topology. A handshakeprocedure based on a local path restoration scheme between neighboring nodesis used to recover from a failure. Failure recovery is done by enforcing routing


22/23

table consistency between upstream and downstream nodes on each path.

Simulation results showed that SAR offers less power consumption than theminimum-energy metric algorithm, which focuses only the energy consumption

of each packet without considering its priority. Although, this ensures fault-

tolerance and easy recovery, the protocol suffers from the overhead of

maintaining the tables and states at each sensor node especially when the

number of nodes is huge.

2.1.4.3. Energy-Aware QoS Routing Protocol:In this QoS aware protocol for

sensor networks, real-time traffic is generated by imaging sensors. The proposedprotocol extends the routing approach in and finds a least cost and energy

efficient path that meets certain end-to-end delay during the connection. The

link cost used is a function that captures the nodes energy reserve, transmission

energy, error rate and other communication parameters. In order to support

both best effort and real-time traffic at the same time, a class-based queuing

model is employed. The queuing model allows service sharing for real-time and

non-real-time traffic. The protocol finds a list of least cost paths by using anextended version of Dijkstras algorithm and picks a path from that list which

meets the end-to-end delay requirement. Simulation results show that theproposed protocol consistently performs well with respect to QoS and energy

metrics, however, it does not provide flexible adjusting of bandwidth sharing fordifferent links.

2.1.5. Coherent and non-coherent processing:Data processing is a major componentin the operation of wireless sensor networks. Hence, routing techniques employ

different data processing techniques. There are two ways of data processing basedrouting.

2.1.5.1. Non-coherent data processing: In this, nodes will locally process the raw

data before being sent to other nodes for further processing. The nodes that

perform further processing are called the aggregators.2.1.5.2. Coherent data processing: In coherent routing, the data is forwarded to

aggregators after minimum processing. The minimum processing typicallyincludes tasks like time stamping, duplicate suppression, etc. When all nodes are

sources and send their data to the central aggregator node, a large amount ofenergy will be consumed and hence this process has a high cost. One way to

lower the energy cost is to limit the number of sources that can send data to the

central aggregator node.

Conclusion and Future Research

One of the main challenges in the design of routing protocols for WSNs is energy efficiency dueto the scarce energy resources of sensors. The ultimate objective behind the routing protocol

design is to keep the sensors operating for as long as possible, thus extending the networklifetime. The energy consumption of the sensors is dominated by data transmission andreception. Therefore, routing protocols designed for WSNs should be as energy efficient aspossible to prolong the lifetime of individual sensors, and hence the network lifetime.


23/23

In this paper, we have surveyed a sample of routing protocols by taking into account severalclassification criteria, including location information, network layering and in-networkprocessing, data centricity, path redundancy, network dynamics, QoS requirements, and networkheterogeneity. For each of these categories, we have discussed a few example protocols. Twoimportant related research directions should receive attention from the researcher namely thedesign of routing protocols for duty-cycled WSNs, and three-dimensional (3D) sensor fields

when designing such protocols. Although most of research work on WSNs, in particular, onrouting, considered two-dimensional (2D) settings, where sensors are deployed on a planar field,

there are some situations where the 2D assumption is not reasonable and the use of a 3D designbecomes a necessity. In fact, 3D settings reflect more accurate network design for real-worldapplications. For example, a network deployed on the trees of different heights in a forest, in abuilding with multiple floors, or underwater, requires design in 3D rather than 2D space.

Although some efforts have been devoted to the design of routing and data disseminationprotocols for 3D sensing applications, we believe that these first-step attempts are in theirinfancy, and more powerful and efficient protocols are required to satisfactorily address allproblems that may occur.

Thank You

Have a Nice Day

Date post:	03-Jun-2018
Category:	Documents
Upload:	itskingjain
View:	220 times
Download:	0 times

A Review Routing Techniques in Wireless Sensor Networks

Documents