research article sparco: stochastic performance analysis...

Research ArticleSPARCO Stochastic Performance Analysis with Reliability andCooperation for Underwater Wireless Sensor Networks

Sheeraz Ahmed12 Nadeem Javaid1 Ashfaq Ahmad1 Imran Ahmed2 Mehr Yahya Durrani3

Armughan Ali3 Syed Bilal Haider3 and Manzoor Ilahi1

1COMSATS Institute of Information Technology Islamabad Pakistan2Institute of Management Sciences (IMS) Peshawar Pakistan3COMSATS Institute of Information Technology Attock Pakistan

Correspondence should be addressed to Nadeem Javaid nadeemjavaidqaugmailcom

Received 22 November 2015 Accepted 5 January 2016

Academic Editor Wei Cao

Copyright copy 2016 Sheeraz Ahmed et alThis is an open access article distributed under the Creative CommonsAttribution Licensewhich permits unrestricted use distribution and reproduction in any medium provided the original work is properly cited

Reliability is a key factor for application-oriented Underwater Sensor Networks (UWSNs) which are utilized for gaining certainobjectives and a demand always exists for efficient data routing mechanisms Cooperative routing is a promising technique whichutilizes the broadcast feature ofwirelessmediumand forwards datawith cooperation using sensor nodes as relaysHere we present acooperation-based routing protocol for underwater networks to enhance their performance called Stochastic Performance Analysiswith Reliability andCooperation (SPARCO) Cooperative communication is explored in order to design an energy-efficient routingscheme for UWSNs Each node of the network is assumed to be consisting of a single omnidirectional antenna and multiple nodescooperatively forward their transmissions taking advantage of spatial diversity to reduce energy consumption Both multihopand single-hop schemes are exploited which contribute to lowering of path-losses present in the channels connecting nodesand forwarding of data Simulations demonstrate that SPARCO protocol functions better regarding end-to-end delay networklifetime and energy consumption comparative to noncooperative routing protocolmdashimproved AdaptiveMobility of Courier nodesinThreshold-optimizedDepth-based routing (iAMCTD)The performance is also compared with three cooperation-based routingprotocols for UWSN Cognitive Cooperation (Cog-Coop) Cooperative Depth-Based Routing (CoDBR) and Cooperative PartnerNode Selection Criteria for Cooperative Routing (Coop Re and dth)

1 Introduction

UWSNs consist of sensors and vehicles that are deployedover a given region to perform collaborative monitoringtasksThese networks offer variety of applications like tacticalsurveillance environmental monitoring assisted navigationresource investigation and disaster prevention The physicallayer has a strong influence on UWSNs due to the presenceof acoustic waves Acoustic waves are the most accuratesource of reaching up to desirable range and rate of datatransmission in UW communications Radio waves do notsupport required data rate and range as they get absorbedin water very quickly New achievements in underwateracoustic (UWA) communications however make adequate

data forwarding over long distances Many techniques havebeen investigated on developing networking solutions forUWSNs including acoustic channel modeling physical layertransmission analysis and networking protocols

Acoustic waves experience large delay spreads as theymove at the speed of 1500ms This speed makes one-fifthspeed as that of radio waves Path-loss in the case of radiochannels depends solely on the length of the link but theacoustic waves in UW experience both link length andfrequency dependent path-losses Bubbles and suspendedparticles in water make wide dispersion in acoustic wavesAlso the reflections from surface of water as well as from thebed of the sea increase the channel fading All these factors

Hindawi Publishing CorporationJournal of SensorsVolume 2016 Article ID 7604163 17 pageshttpdxdoiorg10115520167604163

2 Journal of Sensors

have to be taken into account for the design of UWAwirelesssystems

Having now sufficient technological advancements madein the field of radio communications researches are tryingto enhance the working of UW systems using modern tech-niques adopted from radio communications A promisingtechnique is cooperative communication already being usedin terrestrial WSNs It is a potential approach for distributedUWSNs to upgrade the quality of link connecting sensors aswell as the reliability in both point-to-point and multipointenvironments having multiple relays doing cooperationWireless network designs take into account the diversityto improve the overall successful transmissions by allowingduplicate signals at the receiver In contrast to this approachMultiple-Input Multiple-Output (MIMO) technique alsouses a promising mechanism to improve Signal-to-NoiseRatio (SNR) by enhancing diversity gain But the techniqueneeds extra equipment cost for each sensor with muchcomplexity A different approach for gaining diversity is toutilize several sensors cooperatingwith each other to upgradethe quality of communication channel In variation to anindividual sensor having an antenna array duplicate data isforwarded by an array of distributed antennas comprised ofseveral sensors to reach the end-point but introducing somedelay Spatial diversity concept has directed regular efforts toits use in wireless networks particularly in WSNs If variouspaths existing between two end devices are not dependent oneach other and have adequate working channel efficiency canbe improved by forwarding multiple duplicates of data alongthese links and merging them at the destination The totalerror probability decreases since the paths are independentwhich makes the channel and system performance increaseHowever in general case of WSNs nodes are very small tothat needed for the support of such distributed antennasHence to combat such issues the idea of cooperative routingis proposed Cooperation is defined as a group of entitiesworking together to achieve a common goal while sharingeach otherrsquos resources In these systems transmitter forwardsone copy of data packets to a node acting as relay The relaythen decodes or amplifies each data packet as the schemesuggests and reforwards it to the final receiver Relay nodeuses a path which is generally different from the direct pathThe destinationmerges or utilizes both of the received signalsto extract the forwarded information

Cooperative diversity an alternative to combat fadingin wireless channels allows distributed users to help relayinformation of each other to explore inherent spatial diver-sity present in channels Various cooperation-based pro-tocols have been proposed in literature like fixed relayingprotocol adaptive relaying protocol user cooperation pro-tocol and coded cooperation schemes In fixed relayingschemes such as Amplify-and-Forward (AF) and Decode-and-Forward (DF) relays provide assistance to providethe source information In AF scheme the relays amplifyand forward the information whereas in DF scheme therelays decode the received information and then transmitit to the receiver Although in DF the relay forwardsthe decoded information to the receiver however thisscheme has adequate degradation in performance if the

relay does not decode the transmitters information prop-erly

2 Related Work

Earlier efforts to investigate the underwater behavior werejust depending on the technology available for terrestrialsensor networks However UWSNs show several structuraldesign differences in comparison to terrestrial networksespecially due to the water used as the transmission mediumand the signal required for data transmission Design of asuitable structure for UWSNs is also complex due to thebehavior of communication system

An efficient scheme for UWSN Coop (Re and dth) [1]employs cooperative routing which involves data transmis-sion via partner noderelay towards sink In this paper twodifferent partner node selection criteria are implementedand compared The authors have considered source nodedepth-threshold (dth) potential relays depth and residualenergy (Re) as one criterion and SNR of the link connectingsource node with relay or destination as another criterionfor selection parameters Cog-Coop [2] is an efficient schemefor maximization of network lifetime using residual energyof the nodes It improves the spectrum sensing performancealong with having better energy consumption of the sensorsOptimal conditions are attained based on the standard opti-mizationmethods to find the priority of sensors for spectrumsensingThey showed that cooperation among cognitive sen-sors is necessary for lowering the fading as well as shadowingeffects and hence correct sensing Cooperative Depth-BasedRouting (CoDBR) for UWSNs [3] is a cooperation-basedrouting protocol proposed to enhance network performanceEfficient nodes for relaying are selected on the basis ofinformation of depth Source node transmits data to thesink node by relay sensors using cooperation In [4] authorspropose forwarding function based routing scheme iAMCTDfor UWSNs The reactive protocol enhances the underwaternetwork lifetime by the best possible mobility design of sink

A communication path based routing scheme calledRelative Distance Based Forwarding (RDBF) is presented in[5] The authors have used a mathematical factor to judgeand measure the level of suitability for a sensor to relaythe data packets Few nodes are involved in forwardingmechanism helping in reduction of end-to-end delay aswell as energy consumption The protocol also controls theforwarding time of the multiple transmitters to minimizethe duplicate data transmission In [6] the authors havepresented routing protocols for cylindrical networks anddeveloped mathematical models to achieve best selection ofpath for information forwarding Their results showed thatthe suggested 4-chain based protocol performs efficiently interms of lifetime of the network path-loss end-to-end delayand packet sending rate

The problem of tracking UW moving targets is tackledin [7] For 3-dimensional UW maneuvering target track-ing the cooperating model technique is added with theparticle filter to handle with qualms Simulation outcomesexplain that the suggested scheme is a good alternatefor customary sensor-based or imaging-based approaches

Journal of Sensors 3

In [8] the authors propose cooperation-based scheme basedon both Quality-of-Service (QoS) and energy consumptionTo amalgamate the two strictures in the scheme a com-petitive approach is adopted at each sensor by utilizingMultiagent Reinforcement-Learning (MRL) algorithm Thesuggested scheme guarantees improved working regardingpacket loss rate and end-to-end delay while consideringthe energy consumption of the network The main idea ofcooperative communication is to utilize the resources ofmorethan one node to transmit data Hence by resource sharingbetween sensors quality of transmission is upgraded ACOA-AFSA forwarding protocol is presented in [9] which has thepositive characteristics of Artificial Fish Swarm Algorithm(AFSA) and Ant Colony Optimization Algorithm (ACOA)The fusion algorithm reduces the energy consumption andtransmission delay of the existing routing protocols andpromises to improve the robustness The Remotely PoweredUWA Sensor Networks (RPUASN) protocol proposed in [10]shows that sensors harvest and store the power given by anexternal acoustic source upgrading their stability period to alarge extent The desired number of sensors and the regionwhich is trusted to be sensed by the sensors are examinedregarding range power directivity and frequency of theexternal source

Authors in [11] investigate a 3-dimensional underwaternetwork which aims to cover up the coverage hole problemand reduces the energy consumption of the sensors alongwith enhancing the collection of data They used an apple-slice technique to build multiple sectors to enclose the holeand to guarantee the continuity of routing path Results showthe working upgradation of successful packet delivery ratioand reduction of message overhead and power consumptionIn [12] the authors propose a time-based priority forwardingmethod to prevent flooding Result outcomes indicate thatthe routing protocol attains outstanding working regardingpacket delivery ratio energy consumption and end-to-enddelay

In [13] authors have presented a detailed measurementof fading and path-loss characteristics for sensors in bothflat and irregular outdoor land using mathematical path-lossmodel Research in [14] addresses the diverse challenges facedin an underwater environment and the advancements beingin progress According to authors due to the cost of seatests and the lack of standards there are no operational UWnetworks but only experimental demonstrations Efficientand scalable protocols are needed if bigger deployments areto be expected Authors in [15] have focused on a surveyof existing routing protocols in UWSNs Firstly they haveclassified routing protocols into two categories based ona route decision maker Then the performance of existingrouting protocols is compared Furthermore future researchissues of routing protocols in UWSNs are carefully analyzed

An approach to mapping of a wireless connected UWRobotic Fish (URF) is presented in [16] It is based on bothCooperative Localization Particle Filter (CLPF) scheme andOccupancy GridMapping Algorithm (OGMA) CLPF showsthat no prior information about the kinematic model of URFis needed to attain correct 3D localization Simulations showthe effectiveness of the suggested scheme In [17] researchers

present a contention-freemultichannelMediaAccessControl(MAC) scheme for underwater networks that performs welleven when sensors go through uneven and heavy trafficconditions Simulations show that the scheme saves energyand is quite reasonable for a bursty-loaded environment In[18] the authors describe a new acoustic modem ITACA forUWSN which includes a low-power asynchronous wake-upsystem implementation This modem for UWA forwardingis based on a low-cost radio frequency integrated circuitThe characteristic enables a much less power dissipation of10W in stand-bymode In [19] the authors introduce an UWsensor node equipped with an embedded camera Utilizingthis platform the authors present a speedy and much correctdebris detection algorithm based on compressive sensingtheory to consider the challenges of UWA environmentsExperimental results identify that their approach is trustwor-thy and suitable for debris detection using camera sensors inUW environments

In [20] the authors present localization schemes usingcolor filtering technology called Projection-Color FilteringLocalization (PCFL) and Anchor-Color Filtering Localiza-tion (ACFL) Both algorithms focus at jointly accomplishingcorrect localization of UWA sensors with much less energyconsumption They both accept the overlapping signal areaof task anchors which communicate with the mobile nodedirectly as the current sampling region Comparison of thenearness degrees of the RGB sequences between samples andthe mobile node filters out the samples Simulations indicatethat the suggested techniques have quite better localizationperformance and can timely localize the mobile node In[21] the authors have presented a scheme Ultrasonic FrogCalling Algorithm (UFCA) for UWSNs that targets achievingenergy-efficient routing under harsh UW conditions In thisscheme the selection of forwarding nodes is adopted in thesame way of calling behavior of frogs for mating The sensornodes located in worse places go into sleep mode for energyconservation In [22] the authors proposed a balanced trans-mission mechanism for decreasing energy consumptionThey divided the data transmission phase into two phases andthen determined single-hop or multihop data transmissionof the node to the sink depending on the residual energyof the node The authors in [23] presented a cooperation-based Harvest-Then-Cooperate (HTC) scheme for WSNs inwhich the source and relay harvest energy from the accesspoint in the downlink and work cooperatively in the uplinkfor the source information transmission The impacts ofthe system parameters like relay number time allocationand relay position on the throughput performance wereinvestigated Modified Double-Threshold Energy Detection(MDTED) scheme is presented in [24] which is a cooperativespectrum sensing scheme for WSNs The paper incorporateslocation and channel to improve the clustering mechanismand hence collaborative sensing ability

3 Motivation

Majority of network applications consist of battery-powerednodes having limited transmission-reception range Thuscooperative communication or alternately cooperative


routing is particularly needed for such networks in whichsensor nodes share their resources among each otherMigrating from lengthy but weak links to shorter as wellas strengthened links can minimize the load on the pathconnecting sensors Different available paths present betweensensors and sinks provide means for the new design optionsregarding scheduling and routing

Cog-Coop uses both spectrum sensing information andresidual energy of sensors to select the optimal forwardernodes It improves the spectrum sensing performance alongwith having better energy consumption of the sensorsOptimal conditions are attained centered on the typicaloptimization approaches to find the priority of sensors forspectrum sensing They showed that cooperation amongcognitive nodes is necessary for lowering the fading as wellas shadowing effects and hence correct sensing Also theforwarding node selection criteria are suggested to saveenergy and deal with the issue of direct communications withthe fusion center It is a cooperative routing protocol andredundant transmissions consume a lot of energy Thereforepackets are forwarded over a single lossy channel in amultihop order Due to noise and multipath fading in UWenvironment signal suffers high bit error rate In iAMCTDa routing scheme is proposed to maximize the lifetime ofreactive UWSNs iAMCTD considers signal quality as well asresidual energy for routing metrics It is a prototype schemein localization-free and flooding centered data forwardingfor UW scenarios It improves the network throughput andminimizes packet drop ratio to a larger extent by usingformulated forwarding functions iAMCTD faces redundanttransmissions resulting in major energy consumption

Coop (Re and dth) aims to solve the issues of EEDBR andiAMCTD via cooperative diversity This protocol involvesdata transmission through the use of partner nodesrelaysthat use cooperation to route the data to the sink Thisimproves the rate of fruitful packets delivery to the sink asin the chances of link disaster at least one link is present forforwarding of packets reliably to the final end The schemeconsiders a node link state information alongwith its residualenergy and depth as selection criteria So Coop (Re and dth)is consuming more broadcasting energy than iAMCTD andEEDBRThis presents uswith the trade-off between reliabilityand energy conservationAlso the protocol does not considerany transmission impairments present in the underwaterenvironment CoDBR aims to solve the issues of EEDBRvia cooperative diversity The protocol chooses the forwardersensor using two relays selected on the basis of least depth thatroute their packets to sink cooperatively This enhances thesuccessful data delivery rate because if a link fails then at leastanother path is present for forwarding packets successfullyto the sink As CoDBR considers one source node andtwo relay sensors to forward packets to next hop henceCoDBR consumes three times greater broadcasting energythan EEDBR It is also a trade-off between link availabilityand energy conservation In order to address the issues of allthese four protocols we have tried to propose a new protocolby the name of SPARCO

In SPARCO scheme we suggest an approach to forwardinformation throughUWSNswithmuch lower path-loss over

a link using the features of single-hop and multihop Theprotocol uses a cost function to find the most suitable pathto sinkThis function is computed using node distances fromthe destination and their residual energies Simulations depictthat SPARCO scheme enhances the network stability periodwith much reduced path-loss to a large extent

The research takes into account an underwater acousticenvironment where the channel is largely attenuated byfading and other noise effects The sensed signal by thesensors is modeled by a Rayleigh random variable Theproposed mechanism leads to enhancing the reliability ofthe underwater channel by the use of cooperative routingIn this work we consider the technique of FRC for signalcombining Cooperative diversity is obtained without usingvarious antennas This is particularly useful when frequencytime and diversity by the use of several antennas are notappropriate This motivated us not only to the utility ofcooperation in underwater environment but also to evaluateits impact on system performance

4 SPARCO System Model

Figure 1 shows a 5-nodeUWSNmodel where 119878 and119863 are thesource and destination nodes respectively Data transferredby 119878 is received by nodes 1198771 1198772 and 1198773 Having donewith the initial transmission nodes 119878 1198771 and 1198772 have thedata and cooperatively transmit the data to 1198773 and then 119863Let the minimum energy path from 119878 to 119863 is determinedthrough 1198772 that is 119878 rarr 1198772 rarr 119863 Node 1198771 presentwithin the transmission range of 119878 to 1198772 receives the dataforwarded from 119878 The dark lines in Figure 1 show the directcommunication paths whereas the dotted lines indicate thecooperative routes in case the direct transfer path is notavailable or not feasible to use

It is assumed that each sensor dynamically adjusts itstransmission power to manage its range and further thatvarious sensors cooperating in forwarding the data to a singlereceiving node can delay their transmission for perfect phasesynchronization at119863

It is also assumed that the information is routed from119878 to 119863 in transmission slots one after another For eachslot a sensor is selected to broadcast the data to a group ofsensors that received the informationThiswill help the nodesto cooperate for transmitting the data to another group ofsensors

The routing problem is tackled as a multistage decisionproblem The decision is made at each stage to select thereceiving as well as forwarding group of sensors and thetransmission power level among all sensors of that stage aswell The purpose is to get the information to the final pointwith minimum energy

Let the transmitting set at any stage 119896 be denoted by 119878119899

and let the receiving set be denoted by 119877119898 The link cost

between 119878119899and 119877

119898 denoted by 119862(119878

119899 119877119898) is the minimum

power needed for data transmission from 119878119899to 119877119898 Here

119878119899= 1199041 1199042 119904

119899 and 119877

119898= 1199031 1199032 119903

119898

To derive the equations for the costs on the link connect-ing nodes we consider 4 different possibilities [24]


S

D

R3

R1R2 Transmissionradius

Figure 1 Multihop routing

(i) Point-to-Point Link That Is (119899 = 1 119898 = 1) A singlesource node is transmitting to a single receiving node withina time-slot

(ii) Point-to-MultipointBroadcast LinkThat Is (119899 = 1 119898 gt

1) One node is transmitting to various receiving nodes

(iii) Multipoint-to-PointCooperative 119871119894119899119896 That Is (119899 gt

1 119898 = 1) Various nodes are cooperating in forwarding thesame data to one receiver The signal is aggregated at thereceiver and full decoding is only possible if the received SNRis more than the threshold SNRmin

(iv) Multipoint-to-Multipoint Link That Is (119899 gt 1 119898 gt

1) Several nodes transmitting data to various other nodesmakes redundant transmission at various receivers which isnot feasible Hence it is dealt with as a MIMO case

The preceding four cases are discussed individually withtheir link-cost formulations

Case 1 (point-to-point link ie (119899 = 1 119898 = 1)) Here 119878119899=

1199041 and 119877

119898= 1199031 Let the wireless channel in underwater

between these nodes be described by 2 factors magnitudeor attenuation factor 120572

119899119898and phase delay 120601

119899119898as shown in

Figure 2 The generated signal is managed by a scaling factor119891 In UW applications 119891 is a complex factor adopting bothphase and power adjustment by the sender As there is a singlereceiver in this case hence the phase can be ignored

The received signal is computed as follows [24]

119909 (119905) = 120572119890119895120601

119891 (119905) + 119873 (119905) (1)

where (119905) is the unit-power generated signal and 119873(119905) isthe cumulative receiver noise in UW having power 119875

119873 Net

generated power is 119875119879= 1198912 and SNR at the receiving node

is 1205722|1198912|119875119873 SNR needs to be more than the threshold value

SNRmin Minimum power needed is 119875119879and hence the point-

to-point 119862119901minus119901

(1199041 1199031) is given by [24]

119862119901minus119901

(1199041 1199031) = 119875119879=SNRmin 119875119873

1205722

119875119879prop

1

1205722=

1

1198892

(2)

s1 r112057211

12060111

Figure 2 Single-input-single-output linkage

12057211

12057212

12057213

s1

1205721m

r1

r2

r3

rm

Figure 3 Broadcast linkage

that is119862119901minus119901

is inversely proportional to square of the distance119889 between 119904

1and 1199031

Case 2 (point-to-multipoint link ie (119899 = 1 119898 gt 1)) Inthis case 119878

119899= 1199041 and 119877

119898= 1199031 1199032 119903

119898 as shown in

Figure 3 thus 119898 similar SNR limitations should be satisfiedat receiver nodes The signal transmitted by 119904

1is received

by all those sensors that are within the transmission radiusand proportional to the transmission power 119875

119879 Minimum

power needed for this broadcast transmission is representedby 119862119901minus119898

(1199041 119877119898) expressed as

119862119901minus119898

(1199041 119877119898) = maximize 119862

11(1199041 1199031) 11986212(1199041 1199032)

1198621119898

(1199041 119903119898)

(3)

As the receiver nodes can be in various dimensionsin terms of placement so we use Principal ComponentAnalysis (PCA) technique for this maximization problemPCA is a reknowned technique for dimension reduction andexploratory data analysis [25]

For a group of sensed V-dimensional data vectors V119905

119905 isin 1 2 119879 the 119902 principal axes 119908119895 119895 isin 1 2 119902

are those orthonormal axes onto which the retained varianceunder projection is maximal The vectors 119908

119895are expressed

by the 119902 dominant eigenvectors (ie ones having maximumassociated eigenvalues 120582

119895) of the sample covariance matrix

119872 = 119864[(Vminus120583)(Vminus120583)119879] so that119872119908119895

= 120582119895sdot119908119895 The 119902 principal

components of the observed vector V119905are expressed in terms

of the vector 119909119905= 119882119879

(V119905minus120583) where119882119879 = (119908

1 1199082 119908

119902)119879

The variables 119909119895are such decorellated that the covariance

matrix 119864[119909 sdot 119909119879

] is diagonal with elements 120582119895

Case 3 (multipoint-to-point link ie (119899 gt 1 119898 = 1)) Inthis case 119878

119899= 1199041 1199042 119904

119899 and 119877

119898= 1199031 as shown in

Figure 4 Here 119899 transmitting nodes adjust their phases suchthat the signal is received at the receiver in phase


1205721n

r1

sn

12057211

12057212

12057213

s1

s2

s3

Figure 4 Cooperative linkage

As we are considering PCA in this modeling so alatent variable model is required to relate the group of V-dimensional observed data vectors V

119905 to a corresponding

group of 119902-dimensional latent variables 119909119898 such that [24]

119897 = 119910 (119909 120579) + 120598 (4)

where119910(119909 120579) is a function of the latent variable119909with param-eter 120579 and 120598 is an119909-independent noise process Equation (4) isquite similar to (1) which is utilized for a single point-to-pointnode data transfer

Generally 119902 lt V so that the latent variable offers a morepromising description of the data and in Case 3 119902 = 1 thusonly a single receiving node is being used here In standardform the mapping (119909 120579) is linear so

119897 = 119882119909 + 120583 + 120598 (5)

where the latent variables 119909 sim N(0 119868) have a unit Gaussiandistribution The noise model is also Gaussian and 120598 sim

N(0 120595) having 120595 diagonal the (V times 119902) parameter matrix inour case that is (Vtimes1) Hence the model for 119897 is also normalthat isN(120583 119862) where the covariance 119862 = 120595 +119882119882

119879The model is selected because of the diagonality of 120595 and

the sensed variables 119897 are independent of latent variables 119909In this case the latent variables are the parameters 120572

1198991and

1206011198991 Analytic solution for119882 and 120595 does not exist in literature

and their values are to be computed by iterative methodsFor the model of (5) with an isotropic noise 120595 = 120590

2 Thenoise model 120598 sim N(0 120590

2

) in (5) implies that a probabilitydistribution over 119897-space for a given 119909 is expressed by [24]

119901 (119897119909) = (21205871205902

)minus1199052

exp minus1

21205902

1003817100381710038171003817119897 minus 119882119909 minus 1205831003817100381710038171003817

2

(6)

with a Gaussian over latent variables defined by

119901 (119909) = (2120587)minus1199022 exp minus1

2sdot 119909119879

sdot 119909 (7)

Here in Case 3 119902 = 1 so

119901 (119909) = (2120587)minus12 exp minus1

2sdot 119909119879

sdot 119909 (8)

and marginal distribution of 119897 is in the form

119901 (119897) = int119901 (119897119909) sdot 119901 (119909) sdot 119889119909 (9)

119901 (119909)

= (2120587)minus1199022

|119862|minus12 exp minus1

2sdot (119897 minus 120583)

119879

sdot 119862minus1

(119897 minus 120583)

(10)

where 119902 = 1 and the model covariance is

119862 = 1205902

119868 + 119882119882119879

(11)

Net transmitted power is sum119899

119894=1|119891119894|2 and the received

signal power is | sum119899119894=1

1198911198941205721198941|2 Hence for this case the power

allocation problem can be stated as

minimize119899

sum

119894=1

10038161003816100381610038161198911198941003816100381610038161003816

2

subject to1003816100381610038161003816sum119899

119894=11198911198941205721198941

1003816100381610038161003816

2

119875119873

ge SNRmin

1003816100381610038161003816100381610038161003816100381610038161003816

119899

sum

119894=1

1198911198941205721198941

1003816100381610038161003816100381610038161003816100381610038161003816

2

ge SNRmin 119875119873

10038161003816100381610038161198911198941003816100381610038161003816

2

1003816100381610038161003816100381610038161003816100381610038161003816

119899

sum

119894=1

1205721198941

1003816100381610038161003816100381610038161003816100381610038161003816

2

ge SNRmin 119875119873

(12)

Applying Lagrangian multiplier technique for each node

10038161003816100381610038161198911198941003816100381610038161003816 =

1205721198941

1003816100381610038161003816sum119899

119894=11205721198941

1003816100381610038161003816

2

radicSNRmin 119875119873 (13)

Resulting link cost by the use of cooperation 119862119898minus119901

(119878 1199031)

defined as the optimal net power is therefore expressed as

119862119898minus119901

(119878 1199031) = 119875119879=

1

sum119899

119894=1(1205722

1198941 (SNRmin 119875119873))

(14)

Case 4 (multipoint-to-multipoint link ie (119899 gt 1 119898 gt 1))In this case 119878

119899= 1199041 1199042 119904

119899 and 119877

119898= 1199031 1199032 119903

119898

as shown in Figure 5 This is the MIMO case which helpsimplementation in cooperative reception and transmission ofdata among group sensors It is assumed that all sensor nodesare working in half-duplex mode This case is a combinationof working formulations of Cases 2 and 3 In Case 2 multiplenodes are transmitting thus the link cost in (3) applies for119899 transmitters and 119898 receivers Moreover the technique ofPCA is applicable for a group of V-dimensional data vectorsV119905 119905 isin 1 2 119879 Similarly in Case 3 we considered a

single receiver node with 119902 = 1 however here the receivingnodes are more than 1 in distinct dimensions thus 119902 lt 119905 and(7) and (10) apply for any value of 119902 and the formulation offactor analysis applies


r1

r2

r3

sn

s1

s2

s3

rm

Figure 5 MIMO linkages

5 Attenuation Propagation Delay andNoise in UW Channel

Simulation of UWSN communication links requires mod-eling the acoustic waves propagation in the scenario that asensor node in UW struggles to forward data to another oneSound propagates in the underwater environment at approx-imate speed of 119888 = 1500ms As a signal travels towards anode its energy dissipates and it is distorted by noise ForUWA links both link distance 119889 and signal frequency 119891

have dependency on attenuation denoted by 119860(119889 119891) Hencefor a generated signal with low bandwidth centered aroundfrequency 119891 with power of unity the received signal has anSNR expressed as 120588(119889 119891) [15]

Underwater channel is influenced by both spreadingand absorption losses causing attenuation to a large extentFor a separation of 119889 (km) for a source and destinationat a frequency 119891 (kHz) and spreading coefficient 119896 theattenuation 119860(119889 119891) is expressed by Urick [26] given as

119860 (119889 119891) = 1198600119889119896

119886 (119891)119889

(15)

where 1198600is called normalizing constant 119896 is known as the

spreading factor whose value is 119896 = 1 for shallow waterpropagation 119896 = 2 for deep water propagation and forpractical spreading 119896 = 15 The absorption coefficient 119886(119891)is modeled by theThorps formula as [27]

10 log 119886 (119891) =011119891

2

1 + 1198912+

441198912

4200 + 119891+275119891

2

104+ 0003

(for 119891 gt 04) [dBkm]

10 log 119886 (119891) = 0002 +011119891

1 + 119891+ 0011119891

(for 119891 lt 04) [dBkm]

(16)

In the absence of site-specific noises receiver is influ-enced only by ambient noises with overall Power SpectralDensity (PSD) in terms of dB relative to 120583Pa in kHz Thebackground noise in UW has various sources according to

frequency and location like turbulence (119873119905) shipping (119873

119904)

waves (119873119908) and thermal noise (119873th) These noise effects are

modeled by Gaussian statistics The PSD of these ambientnoises as described in [27] is

119873(119891) = 119873119905(119891) + 119873

119904(119891) + 119873

119908(119891) + 119873th (119891) (17)

where

10 log119873119905(119891) = 17 minus 30 log119891 (18)

10 log119873119904(119891) = 40 + 20 (119904 minus 05) + 26 log119891

minus 60 log (119891 + 003)

(19)

10 log119873119908(119891) = 50 + 75radic119908 + 20 log119891

minus 40 log (119891 + 04)

(20)

10 log119873th (119891) = minus15 + 20 log119891 (21)

where 119904 is shipping activity factor whose value rangesbetween 0 and 1 for low and high activity respectively and119908 is the wind velocity ranging from 0 to 10ms

6 SNR in UW Acoustic Channel

The SNR of a generated underwater signal having transmitpower of unity (119905) [watts] at the receiver is expressed by

SNR (119889 119891) = 120588 (119889 119891) = SL minus 119860 (119889 119891) minus 119873 (119891) minus DI (22)

where 119860(119889 119891) is attenuation and 119873(119891) [WHz] is the noisePower Spectral Density (PSD) given in (15) and (18) respec-tively The noise model is the same as considered in (6) withvariance 1205902 which is utilized here

Directivity index is DI = 0 if we assume an omnidi-rectional antenna Source level SL = 20 log 1198681 120583Pa if 119868 isintensity at unity distance from a source node in wattm2given by

119868 = (119905)

2120587119867 (23)

in which119867 is the water depth Signals travelling in underwa-ter channels (119879

119889) normally experience frequency and path-

losses which are much complicated than radio channels andare modeled as [27]

119879119889= 10 log

10119889 + 10

minus3

119886 (119891) 119889 (24)

where 119886(119891) is given in (16) First term of (24) expresses thepower consumption of signals propagating from source toreceiver inwireless channels Second term is the absorption ofpropagating wave power due to acoustic waves [27] Figure 6shows the schematic flow chart for the SPARCO protocol

7 Outage Formulation in UWA Channel

Thenoises in UW followGaussian distribution and the chan-nel is stable for some period known as the coherence time


Reliability computation End

Outage formulation

SNR computation

Follow relay path

No

Direct transfer

Yes

Node elected as relay

Satisfies the cost

Point-to-to-point linkpoint link

Point-to-

linkmultipoint Multipoint-

Relay node selection

Cooperation phase

StartSystem configuration

and initialization

to-multipointMultipoint-

No

function criteria

Routing phase

Yes

SNR lt SNRmin

R E of source gt R Eof relay

link

SNR ge SNRmin

Figure 6 Flow chart of SPARCO protocol

The capacity of a channel with infinite bandwidth determinesthe upper limit on the maximum information which can becommunicated over a channel successfully Shannon-Hartleytheorem [27] expresses this by stating that ldquoA communicationsystem has a maximum information transfer rate 119862 knownas the channel capacity If the information rate 119877 is less than119862 then we can expect arbitrarily small error probabilitiesusing intelligent coding techniques To get minimum errorprobabilities the encoder has to work on longer blocks of sig-nal data This entails longer delays and higher computationalrequirementsrdquo Thus

119862 (119889 120588) = 119861 log2(1 + 120588 (119889 119891)) (25)

where 119862(119889 120588) [bitssec] is the channel capacity dependingon both frequency and distance whose expression makesintuitive sense

(1) When the bandwidth of a channel rises then wecan make rapid variations in the information signalwhich increases the information rate

(2) When the value of 119878119873 rises we can raise theinformation rate while still preventing errors due tonoise

(3) If noise is absent 119878119873 approachesinfin and amaximuminformation transfer is possible independent of band-width


Hence there is a trade-off bandwidth for SNR Howeveras119861 rarr infin channel capacity does not tend to infinity becausewith a rise in bandwidth the noise power also increases

Let the information transfer rate at each sensor be119877 [bitssec] then the signal is expected to be forwardedsuccessfully on lossy channels if the capacity is equal to ormore than the transfer rate that is

119862 (119889 120588) ge 119877 (26)

These criteria assess the quality of incoming signal at thedestination This helps us to approximate the efficiency of alink in wireless systems without any variations in detectiondecoding or complex coding mechanisms Contrary to (26)outage event can take place when the data transfer rate119877 goesmore than 119862 that is

Outage = 119862 (119889 120588) lt 119877 (27)

An error is said to take place when the channel goes inoutage or a decoding error occurs The probability of error isapproximately zero if the channel is not in outage Thereforethe outage probability 119875outage is expressed as

119875outage = 119875 119862 (119889 120588) lt 119877 (28)

119875outage = 119875 119861 log2(1 + 120588 (119889 119891)) lt 119877 (29)

119875outage = 119875 120588 (119889 119891) lt 2(119877119861)

minus 1 (30)

119875outage = 11987520 log( (119905)

2120587119867) minus 119860 (119889 119891) minus 119873 (119891)

lt 2(119877119861)

minus 1

(31)

119875outage = 119875log( (119905)

2120587119867)

lt2(119877119861)

+ 119860 (119889 119891) + 119873 (119891) minus 1

20

(32)

119875outage = 119875 (119905) lt 2120587119867 anti

minus log(2(119877119861)

+ 119860 (119889 119891) + 119873 (119891) minus 1

20)

(33)

Equation (33) clearly indicates that the outage probabilityat any instant is totally dependent on the depth of the oceanand the attenuation and noise factors occurring in oceancurrents In terms of SNR and ignoring the (minus1) term (33)can be expressed as

119875outage

= 119875 (119905) lt 2120587119867 exp(2(119877119861)

+ SL + 120588 (119889 119891)

20)

(34)

8 Reliability in UWA Channel

With the passage of time and advancement in technologythere are lots of methods that are used to avoid the loss ofdata when the channel is in outages like employing ARQprotocols obtaining information from the transmitter sidechannel or coding for a longer period [28] However thisstudy concentrates and focuses on the reliability of the linkto be obtained through the use of routing by isolating thediversity obtaining issue and the results can be associatedwith other diversity techniques Different links combineto form hops and then these multiple-hop paths make asequential combination of nodes which pass the informationto one another and ultimately lead to the destination119863 froma source 119878 That event will be deemed as a successful end-to-end transmission in which all the packets or transmissionsare successful and the probability of occurrence of an event isdefined as End-to-End Reliability denoted byR [29] HenceR can be written asR = 1 minus 119875outage (35)

R

= 1

minus (119905) lt 2120587119867 exp(2(119877119861)

+ SL + 120588 (119889 119891)

20)

(36)

According to this expression R is a monotonically decreas-ing function and establishes the result for a point-to-pointlink It is dependent on the depth of the water channel stateand the distance between two nodes The net reliability forthe entire end-to-end path can be computed from (36) asR = 1 minus sum

119899

119894=1(2120587119867119894exp((2(119877119861) + SL + 120588(119889 119891)

119894)20))

Themaximum reliability route is the route thatminimizesthis sum and the maximum amount of power that can bespent in relaying the information from 119878 to 119863 is limited tothe summation of SNR for individual nodes as given in (14)

9 Performance Evaluation of SPARCO

Major metrics of performance for all compared protocols aredefined as follows

91 Performance Metrics

911 Stability Period The total time interval between theinitiation of network till the death of first node is calledstability period

912 Residual Energy It is defined as the difference of theinitial applied energy and that of the utilized energy by thenodes during network operation

913 Network Lifetime The total time taken by the networkoperation is called network lifetime

914 Throughput Throughput is the term which shows thetotal number of successfully received packets at the sink


915 Delay Spread The measure of density or abundanceof the multipath of a communication channel is known asdelay spread It is calculated by taking the difference betweenthe latest multipath component and the earliest multipathcomponent of the received signal in network operation

916 Path-Loss Measured in decibels (dB) is the differenceof the transmitted and received powers of transmitting nodesand receiving nodes respectively

92 Results andDiscussions Existing schemes CoDBR Cog-Coop iAMCTD and Coop (Re and dth) are used asbenchmark to evaluate and analyze SPARCO performanceNodes are randomly deployed for every simulated techniqueUtilizing multiple-sink model of conventional methods hav-ing 10 sinks deployed on the water surface 225 nodes aredeployed randomly in a network field of 500m times 500mwith 4 courier nodes acting as relays The alive nodesin the network transmit threshold-based data to anotherneighbor node which in turn forwards it to another neighborThere is a proper coordination between the nodes that allshare proper physical parameters notably weight and depththreshold with the neighbor node to keep informed withthe fluctuating circumstances of the network Nodes calculatetheir distances from their neighbors after fixed intervalsSensors forward their information to the upper layer usingcooperation of neighboring sensors till that informationreaches the sinkThe sink supervises the depth thresholds andadaptivemobility of cooperating sensorsThe introduction ofdepth thresholds and cooperation makes SPARCO scheme afeasible application for data-critical situations

Figure 7 shows a comparison of the stability period ofSPARCO CoDBR iAMCTD Coop (Re and dth) and Cog-Coop with respect to network lifetime It is obvious fromthe figure that due to the reason of maintaining lower path-loss and neglecting the unnecessary data forwarding resultsin higher stability than the other four schemes In the timeinterval after which the nodes start dying in the simulationsof 10000 seconds the initial or first node in case of SPARCOdies at 4290th second which is more extended than theother four This increases the stability duration such thatin CoDBR first node dies after 961st second in iAMCTDit dies after 3185th second in Cog-Coop it dies after 1857thsecond and in Coop (Re and dth) it dies after 961st secondIn other words we can say that the instability period startsafter around 4300 seconds after which the end-to-end delaydecreases very slowly Cooperative nodes play an importantrole in SPARCO Coop (Re and dth) CoDBR and Cog-Coopby the introduction of cooperation scheme because thesenodes distribute and share the load of data forwarding whichresults in achievement of load balancing hence increasing thestability period Network lifetime of iAMCTD gets increasedcompared to CoDBR as there is a slow raise in energyconsumption As the network becomes sparse slowly numberof neighbor nodes falls suddenly in CoDBR which causesthe instability in the network iAMCTD considers two for-warding attributes depth and residual energy which causesa trade-off between network stability and path-loss This

0 2000 4000 6000 8000 100000

50

100

150

200

250

Time (seconds)

Num

ber o

f aliv

e nod

es

SPARCOCog-CoopCoop (Re and dth)

CoDBRiAMCTD

Figure 7 Alive nodes versus network lifetime

does not make it suitable for reactive applications Networkstability of iAMCTD is raised in comparison to CoDBR dueto lower throughput by responsive network

Cooperation between nodes causes load balancing inboth SAPRCO and Coop (Re and dth) In our proposedprotocol utility of Thorps energy model identifies the netchannel losses which are quite needful for selection ofappropriate data forwarding The consideration of channellosses is absent in the Co-DBR scheme Raise in stabilityperiod of SPARCO also confirms lowering of redundanttransmissionsThe stability period of iAMCTD is better thanCoDBR andCoop (Re and dth) however the network energyconsumption is greater inCoDBR In iAMCTDandCoop (Reand dth) the source nodes are transmitting the date to next-hop neighbor nodes only whereas CoDBR is utilizing sourcenode along with two relay nodes to transmit data to the nexthop In CoDBR the stability period finishes too suddenly dueto prioritization of residual energy in selection of reasonableneighbor nodes which causes inefficient instability periodTable 1 indicates a numerical comparison of all the fourcompared protocols in terms of alive nodes after equalintervals The table shows that as the stability periods ofCo-DBR and Coop (Re and dth) are the least hence if wekeep them as reference then the percentage improvementsin other schemes are shown numerically with regard to theseprotocols

Figure 8 shows the graphical representation of the differ-ence between the energy consumption of iAMCTD (existingnoncooperative schemes) CoDBR and Cog-Coop (existingcooperative schemes) and SPARCO As the graph showsthe energy is more efficiently utilized in SPARCO than theexisting schemes due to efficient forwading of data withthe help of neighbor nodes and load balancing is ultimatelyachieved Another reason for the efficiency is that energyconsumption is improved with the help of effective weightimplementation As SPARCOrsquos main concern is with those


Table 1 Alive nodes available after specified intervals in seconds

Protocol First node dies at Efficiency in percentage 1000 2000 4000 6000 8000 10000CoDBR 961 100 225 207 176 158 146 136iAMCTD 3185 3314 225 225 220 177 120 86Cog-Coop 1857 1932 225 225 187 126 80 52Coop (Re and dth) 961 100 225 213 193 162 138 131SPARCO 4290 4464 225 225 225 200 155 138

0 2000 4000 6000 8000 100000

5

10

15

20

25

30

35

Time (seconds)

Tota

l ene

rgy

cons

umpt

ion

(J)

SPARCOCog-CoopCop (Re and dth)

iAMCTDCoDBR

Figure 8 Total energy consumption versus network lifetime

applications which are time bound therefore it has to beefficient and for the purpose of which it utilizes cooperationand depth difference between data forwarders in the networkto solve the issue of energy consumption

In iAMCTD the larger distance between the neighboringnodes makes them consume high energy while in otherexisting schemes like CoDBR the frequent switching toselect high energy nodesmakes them consume higher energythan other existing techniques making it least efficient Thestability period of Cog-Coop is extended compred to CoDBRand iAMCTD however the network energy consumption isgreater in Cog-Coop The cooperative transmission schemeCog-Coop is consuming approximately three times moreenergy than CoDBR and iAMCTD in stable region Asthe nodes start to die after 3000 seconds the total energyconsumption tends to increase more sharply In Cog-Coopand iAMCTD the source nodes are transmitting the date tonext hop neighbor nodes only whereas CoDBR is utilizingsource node along with two relay nodes to transmit data tothe next hop In Coop (Re and dth) there is a sudden increasein network energy consumption during the initial rounds asall nodes become active and perform the routing processLater on energy consumption decreases because nodes failto find relay nodes due to reduction in network densityHence chances of cooperative routing being performed byany source node are reduced which in turn reduces energy

consumption Table 2 illustrates a comparison of residualenergy left in percentage of all the five compared protocolsafter equal intervals which clearly shows that in SPARCO thevalue of residual energy drop is the least among the fourThe table shows that the maximum residual energy drop isin Cog-Coop and if we consider its drop to be maximumthen the other schemes show improvements in reference tothis schemeThemaximum efficiency in percentage is shownby SPARCO whose averaged drop is 652 in comparison toCog-Coop which is assumed to be 100

Packet delivery ratio (PDR) is the ratio of data packetsreceived at receiver end to those generated by the sourcewhich is the throughput in other words The comparativeanalysis of all the four schemes under study based on PDRcomparison is illustrated in Figure 9 As we can see the dropin the PDR in SPARCO is much less as compared to otherschemes Network lifetime of Coop (Re and dth) gets loweredbut the delivery ratios of iAMCTD and CoDBR show similarplots The delivery ratio in Cog-Coop is improved comparedto that of SPARCO although both schemes are utlizingcooperation This is due to the fact that Cog-Coop utilizesonly two relay nodes whereas SPARCO utilizes multiple relaynodes and considers SNR in the links Traffic is flooded fromthe source nodes in case of iAMCTD and CoDBR whenthe total time between the source and destination in nodesis small which ultimately results in lower PDR due to theincrease in packet collision

In SPACRO scheme instead of single path a number ofdifferent multiple paths are used to forward the data andthen they are combined at the receiver node Due to thisthe data which is sent in packets has higher chance of beingsuccessfully transmitted A larger number of cooperatingnodes are available for data forwarding whereas higherreliability can be achieved The sudden drop in the deliveryratio of SPARCO is due to the fact that the relay nodes usingcooperation consume more energy than normal nodes andhence die early which creates instability in the network Onthe other hand the other two schemes are lagging behindin the reliability of forwarding packets because in CoDBRthe distant propagation as well as multiple forwarding usedresults in lower PDR In iAMCTD the propagations remainstable due to the fact that both residual energy and depthare considered in the weight function computation whichmakes it better than CoDBR in terms of transmission lossCoop (Re and dth) scheme shows a similar type of rise-fall behavior in case of PDR because the scheme does notconsider the channel conditions as well as the SNR of the linkand the throughput decreases due to quick fall in networkdensity Table 3 shows a numerical comparison of all the five


Table 2 Residual energy drop in percent after equal intervals in seconds

Protocol Averaged efficiency in percentage 2000 4000 6000 8000 10000CoDBR 956 506 114 24 367 3987iAMCTD 881 365 158 227 2912 365Cog-Coop 100 265 178 2552 3635 40Coop (Re and dth) 725 32 115 174 23 336SPARCO 652 8 92 144 198 2837

Table 3 Packet delivery ratio after equal intervals in seconds


0 2000 4000 6000 8000 1000030

40

50

60

70

80

90

100

110

Time (seconds)

Pack

et d

eliv

ery

ratio


CoDBRiAMCTD

Figure 9 Packet delivery ratio versus network lifetime

compared protocols in terms of PDR after equal number ofintervals The table also highlights the averaged efficiency ofall the compared schemes in terms of PDR and SPARCOshows an averaged efficiency of 94

Figure 10 further illustrates that the path-loss of thenetwork inCoDBRCoop (Re and dth) Cog-Coop and iAM-CTD is much higher than SPARCO because of prioritizingSNR in its modeling The plots very clearly indicate thatthe effective use of relay nodes in SPARCO and iAMCTDoutperforms the other three schemes Distant transmissionsare preferred in both Cog-Coop and CoDBR while onthe contrary to the existing schemes Uricks model andThorps attenuation model for UWA are utilized in SPARCOand iAMCTD to trace out and measure the total loss intransmission during data transfer between the source andthe destination It takes into account transmission frequency

0 2000 4000 6000 8000 100000

50

100

150

200

250

Time (seconds)

Path

loss

(dB)


CoDBRiAMCTD

Figure 10 Path-loss of network versus network lifetime

bandwidth efficiency and noise effects which scrutinize thesignal quality during data forwarding CoDBR and Coop (Reand dth) are cooperative schemes but do not consider anynoise factors and transmission losses as in SPARCO that iswhy their performance is much less compared to SPARCO

Cog-Coop due to the high network density showsless losses in starting phases of node deployment As thenetwork scatters its losses increase which haunts the networkperformance ultimately resulting in high packet loss duringits transmission In iAMCTD channel loss conditions arebetter than CoDBR and Coop (Re and dth) as its costcomputation considers depth as well as residual energy offorwarding nodes hence the propagations remain stableSPARCO again gives good results compared to the otherschemes as its SNR calculation takes into account both depthand residual energy of the relay nodes as well as the utilization


Table 4 Path loss (dB) after equal intervals in seconds


of cooperation scheme which ultimately results in morestable propagations In SPARCO we have utilized Uricksmodel and Thorps attenuation model for UW environmentto calculate the transmission loss in packet forwardingThesemodels determine the effect of path-loss on transmissionfrequency bandwidth efficiency and noise density duringdata transmission Table 4 indicates a numerical comparisonof all the four compared protocols in terms of path-loss afterequal intervals traversed The table shows that the efficiencyof our scheme SPARCO is maximum in comparison to otherschemes in terms of path-loss and the schemeCo-DBR showsthe minimum efficiency that is approximately 21

Figure 11 shows a comparison of dead nodes in case ofSPARCO with those of CoDBR Coop (Re and dth) Cog-Coop and iAMCTDThe figure shows the time interval afterwhich the nodes start dying In the simulations of 10000seconds the initial or first node in SPARCO dies beingin operation after 4300th second which is greater than theother four schemes This increased the stability durationsuch that in CoDBR nodes start dying after 970th secondin iAMCTD they start dying after 1700th second in Coop(Re and dth) they start dying after 1000 seconds and inCog-Coop the nodes start dying after around 1900 secondsCooperative nodes play an important role by the introductionof cooperation scheme in both SPARCO and Coop (Re anddth) because relay nodes distribute and share the load of dataforwarding which results in achievement of load balancinghence increasing the stability period

Due to the fact that in iAMCTD there is a gradual increasein network energy consumption its stability period is greaterthan CoDBR and Coop (Re and dth) The primary reasonwhich causes network instability in CoDBR is that as thenetwork becomes sparse the number of neighbor nodesdecreases quickly As discussed there are two forwardingattributes in iAMCTD which are depth and residual energyThis results in the trade-off between path-loss of packetsand network lifetime which is not appropriate for reactiveapplications During the instability period of iAMCTD andCoop (Re and dth) network slowly gets sparse creating loadon high energy nodes while the number of neighbors ishandled by variations in depth thresholdThe stability periodof Cog-Coop is less than SPARCO because the nodes areconsuming three times more energy than SPARCO Lifetimeof SPARCO is increased due to lower throughput by reactivenetwork In the proposed schemes Thorps energy modelshelp to analyze specifically and in detail the channel losseswhich is very useful in reactive networks for data forwardingThe redundant transmissions are also reduced due to theincrease in stability period Table 5 illustrates the comparative

0 2000 4000 6000 8000 100000

20

40

60

80

100

120

140

160

180

Time (seconds)

Num

ber o

f dea

d no

des


CoDBRiAMCTD

Figure 11 Dead nodes versus network lifetime

analysis of all the five schemes under study in terms of thetotal dead nodes after equal intervals of time The differencein the table clearly shows that SPARCO has least number ofdead nodes compared to its counterparts

End-to-end delay with respect to five schemes is depictedin Figure 12 Their comparisons show that the delay in caseof SPARCO is lower than the previous schemes due to leasttransmission distances between the sensors in both sparseand dense situations Coop (Re and dth) has the highestdelay compared to the other four schemes because it isutilizing relay nodes in every data transfer and also does notconsider any losses present in the UW medium In CoDBRdelay is much higher in the beginning due to distant datatransmission It slowly gets lowered with the sparseness ofthe network and the network causes data transferring atleast distance In iAMCTD the end-to-end delay is betterthan the previous schemes due to the load balancing of boththreshold variations and weight functions But in SPARCOthere is a minimum time lag because of the considerationof SNR difference in depths of sender and receiver sensorsand introduction of cooperation iAMCTD and CoDBRtransfer data with least hops however the attenuated pathraises the data loss at receiver and packets need to bereforwarded This intensifies the packet delay While allthe four protocols are focusing on channel estimation datapackets are forwarded with better reliability resulting in


Table 5 Dead nodes available after specified intervals in seconds

Protocol First node dies at Efficiency in percentage 1000 2000 4000 6000 8000 10000CoDBR 961 100 2 18 42 55 68 85iAMCTD 3185 3314 Zero 3 28 78 127 147Cog-Coop 1857 1932 Zero 1 38 99 145 173Coop (Re and dth) 961 100 2 13 32 63 87 94SPARCO 4290 4464 Zero Zero Zero 25 70 87

Table 6 End-to-end delay after equal intervals in seconds


0 2000 4000 6000 8000 100000

005

01

015

02

025

03

035

Time (seconds)

End-

to-e

nd d

elay

(sec

onds

)


CoDBRiAMCTD

Figure 12 End-to-end delay versus network lifetime

lower redundancy especially in cooperative schemes Coop(Re and dth) CoDBR and Cog-Coop The packets hencereach the destination with much reduced delay Table 6indicates a numerical comparison of all the five comparedschemes in terms of delay It also highlights the percentageof efficiency of all the compared schemes with the maximumdelay achieved by Coop (Re and dth) which is assumed tobe 100 and all other improvements or drops in delays areexpressed with reference to it

Figure 13 shows the outage probability of proposedSPARCO scheme with a number of cooperative nodes actingas relays These plots help in system design because outageprobability is an important QoS parameter and they allowus to find the optimal number of cooperative nodes For thissimulation we set the transmission rate equal to 1 as outageprobability is dependent on the transmission rate The blueline in Figure 12 shows the variation in outage probability

0 5 10 15 20SNR (dB)

Out

age p

roba

bilit

y

BER for direct communicationBER for relay communication

10minus3

10minus2

10minus1

100

Figure 13 Outage probability versus SNR Outage probabilityversus SNR for direct and relay communication

for direct communication between a transmitting node andreceiving node and no relay is utilized On the other handvariation of outage probability with that of SNR for a fullycooperative environment is shown by the green line

The plots in Figure 13 clearly indicate that for lowervalues of SNR till 6 dB there is no major difference in thetwo plots However beyond this value we find a remarkableimprovement in outage probability for cooperative or relaycommunication in contrast to direct communication This isdue to the fact that with the increase in SNR value morerelay nodes take part in data transfers rather than directcommunication in order to combat the varying underwaterchannel losses Hence we deduce that AF scheme can providecomplete diversity such that the diversity order which utilizesoutage probability as a performance measure is the netnumber of cooperating nodes in the networkThis shows thatthe best number of cooperating sensors is in fact a complex


Table 7 Outage probability for direct versus relay communication after equal intervals of SNR

Type of datatransfer

BER at SNR0 dB BER at SNR 5 dB BER at SNR

10 dBBER at SNR

15 dBBER at SNR

20 dBDirect 0125 0067 005 003 028Relay 0125 005 002 001 0004

function of SNR and the cooperative diversity method inuse Table 7 shows the comparison of the direct data transferversus the relay utilized data transfer after equal intervalsof SNR The table clearly shows that the outage probabilityreduces significantly with SNR increase in case of relaytransfer comparative to direct transfer

93 Performance with Trade-Offs In SPARCO improvementin delay is targeted at the cost of time lag This end-to-enddelay in SPARCO is improved as compared to iAMCTD andCoDBR but at the cost of possible time lag due to consider-ation of SNR and cooperation mechanism In CoDBR delayis improved at the cost of repeated transmissions The delayin CoDBR is much higher in initial stages due to distantdata forwarding but at the cost of redundant transmissionsbecause the attenuated underwater channel increases packetloss at the receiving end In AMCTD delay is improved atthe cost of energy depletion The delay in iAMCTD is betterthanCoDBR as variations in depth threshold aswell as weightfunctions perform load balancing however at the cost ofabrupt energy depletion of the nodes End-to-end delay inCoop (Re and dth) is improved but at the cost of energyutilization and transmission loss

In SPARCO the stability period is improved at thecost of more forwarding nodes and energy consumptionThe protocol enhances the network lifetime by avoidingthe transmission of useless data and maintaining lowertransmission loss however at the cost of utilization of relaynodes and proper selection of relay forwarding nodes InSPARCO the instability period starts after 4300 secondsafter which the PDR remains even however total energyconsumption decreases slowly In iAMCTD the stabilityperiod is achieved at the cost of transmission loss In thisprotocol depth and residual energy are the only two forward-ing selecting variables used in the schemeThis considerationcompromises between the transmission loss of packets andthe network lifetime or is a kind of trade-off between themwhich is not appropriate for reactive application at all Duringthe instability period of iAMCTD network slowly becomessparse creating load on high energy nodes In CoDBR thestability period is improved at the cost of greater energyconsumption In Cog-Coop the stability period improves butat the cost of energy consumption as it utilizes two relay nodesin the data forwarding and also the delay increases as the datahas to flow through the relay nodes till the sink is the next hopIn Coop (Re and dth) the stability period is improved at thecost of delay and transmission loss

In SPARCO the path-loss reduces at the cost of stabilityperiod and the energy consumption as redundant trans-missions between sender sensors and sink is raised due to

the use of relay nodes Figure 9 illustrates how SPARCO isbetter than other techniques in terms of medium path-lossbecause of giving preference to SNR in its computations InCoDBR multiple transmissions increase path-loss betweensender node and the sink We utilize Thorps attenuationmodel for UWA to formulate the transmission loss in packettransferring between a source and destination node throughrelay It considers transmission frequency bandwidth andnoise factors which scrutinize the signal quality during datatransmission Higher throughput in iAMCTD is achieved atthe cost of redundant transmissions between nodes and sinkIn iAMCTD channel losses are better than CoDBR and Cog-Coop as the cost function considers both residual energyand depth of transferring sensors hence the propagationsremain stable With the passage of time in later stages thequalified forward nodes decrease in number due to which thenetwork experiences the delay in the transmission as far as thepacket loss in iAMCTD In Coop (Re and dth) the path-lossis improved at cost of high delay

In SPARCO throughput is improved at the cost of timelag The drop in PDR in SPARCO is lower than that ofother schemes Higher traffic is sent from a source as theinterarrival time of data packets gets lowered This raises thechances of packet collision resulting in a less PDR SPARCOprotocol enhances the probability of successful packet recep-tion by forwarding data on various links and then aggregatingat the receiver node This improvement is achieved at thecost of higher energy consumption of the network as morenodes are involved in the data forwarding mechanism InCoDBR transmission loss improved at the cost of low PDRIn this protocol higher transmission loss is achieved than theother two schemes as it utilizes distant propagations as wellas multiple forwarding iAMCTD achieves improvement inthroughput at the cost of packet loss and delay In iAMCTDchannel loss conditions are better than CoDBR as the weightfunction considers depth and residual energy of active nodestherefore the propagations remain stable However in laterstages the performance of iAMCTD slowly gets loweredwith the decrement in qualified forwarders therefore boththe packet loss and delay increase but at the cost of dropin its throughput Table 8 indicates the various performanceparameters which are enhanced on the price which they haveto pay for the five compared protocols

10 Conclusion

In this research we have proposed SPARCO routing schemeto enhance the stability period and reduce the energy con-sumption of underwater networks Introduction of coop-eration and SNR improves the network lifetime improves


Table 8 Performance parameters with their trade-offs

Protocol Advances achieved Reference Price to pay Reference

SPARCO

Stability period extends Figures 7 and 11 More forwarding nodes and energy consumption Figure 8Transmission loss declines Figure 10 Shorter stability period and energy consumption Figures 8 and 11Throughput increases Figure 9 Time lag and energy consumption Figure 8

End-to-end delay improves Figure 12 Transmission loss Figure 10

CoDBR

Stability period extends Figures 7 and 11 Packet delivery ratio Fig 9Transmission loss declines Figure 10 Redundant transmissions and lesser stability period Figures 7 and 11Throughput increases Figure 9 Transmission loss and delay Figures 10 and 12

End-to-end delay improves Figure 12 Sharp energy depletion Figure 8

iAMCTD

Stability period extends Figures 7 and 11 Extra forwarding nodes and energy consumption Figure 8Transmission loss declines Figure 10 Packet delivery ratio and delay Figures 9 and 12Throughput increases Figure 9 Transmission loss due to distant propagations Figure 10

End-to-end delay improves Figure 12 Transmission loss due to courier nodes Figure 10

Cog-Coop

Stability period extends Figures 7 and 11 Energy consumption due to two nodes relaying and delay Figure 8Transmission loss declines Figure 10 Redundant transmissions and lesser stability period Figures 7 and 11Throughput increases Figure 9 Transmission loss and delay Figures 9 and 11


Coop (Reand dth)

Stability period extends Figures 7 and 11 Redundant transmissions and packet delivery ratio Figure 9Transmission loss declines Figure 10 End-to-end delay Figure 12Throughput increases Figure 9 Transmission loss and greater energy consumption Figures 8 and 10

End-to-end delay improves Figure 12 Transmission loss due to lag of SNR Figure 10

the PDR and reduces the overall network energy con-sumption particularly for delay-sensitive applications andalso in sparse conditions The data forwarding protocolswithout cooperation are focusing on channel conditions thatenhance the quality of the received packet at destinationhowever transmissions along a single path are influencedas the channel quality varies The relay selection criteriatake into account instantaneous path conditions and distanceamong neighboring nodes to reliably forward packets to areceiver node in a limited environment of UWSN Features ofmultihop and single-hop communicationmethods have beenconsidered to lower the path-loss and improve the networkstability Optimal weight calculation and act of cooperationprovide load balancing in the network and give considerableimprovement in the network lifetime

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

References

[1] A Umar M Akbar Z Iqbal Z A Khan U Qasim and NJavaid ldquoCooperative partner nodes selection criteria for coop-erative routing in underwater WSNsrdquo in Proceedings of the 5thNational Symposium on Information Technology Towards NewSmart World Wireless Systems and Networks Dhahran SaudiArabia January 2015

[2] M Najimi A Ebrahimzadeh S M H Andargoli and AFallahi ldquoA novel sensing nodes and decision node selection

method for energy efficiency of cooperative spectrum sensingin cognitive sensor networksrdquo IEEE Sensors Journal vol 13 no5 pp 1610ndash1621 2013

[3] H Nasir N Javaid H Ashraf et al ldquoCoDBR cooperative depthbased routing for underwater wireless sensor networksrdquo inProceedings of the 9th IEEE International Conference on Broad-band andWireless Computing Communication andApplications(BWCCA rsquo14) pp 52ndash57 IEEE Guangdong China November2014

[4] N JavaidM R Jafri Z A Khan U Qasim T A Alghamdi andMAli ldquoiAMCTD improved adaptivemobility of courier nodesin threshold-optimized DBR protocol for underwater wirelesssensor networksrdquo International Journal of Distributed SensorNetworks vol 2014 Article ID 213012 12 pages 2014

[5] Z Li N Yao and Q Gao ldquoRelative distance based forwardingprotocol for underwater wireless networksrdquo International Jour-nal of Distributed Sensor Networks vol 2014 Article ID 17308911 pages 2014

[6] N Javaid M R Jafri Z A Khan N Alrajeh M Imranand A Vasilakos ldquoChain-based communication in cylindricalunderwater wireless sensor networksrdquo Sensors vol 15 no 2 pp3625ndash3649 2015

[7] X Wang M Xu H Wang Y Wu and H Shi ldquoCombinationof interacting multiple models with the particle filter for three-dimensional target tracking in underwater wireless sensornetworksrdquo Mathematical Problems in Engineering vol 2012Article ID 829451 16 pages 2012

[8] M Hammoudeh and R Newman ldquoAdaptive routing in wirelesssensor networks QoS optimisation for enhanced applicationperformancerdquo Information Fusion vol 22 pp 3ndash15 2015

[9] H Wu X Chen C Shi Y Xiao and M Xu ldquoAn ACOA-AFSA fusion routing algorithm for underwater wireless sensor


networkrdquo International Journal of Distributed Sensor Networksvol 2012 Article ID 920505 9 pages 2012

[10] A Bereketli and S Bilgen ldquoRemotely powered underwateracoustic sensor networksrdquo IEEE Sensors Journal vol 12 no 12pp 3467ndash3472 2012

[11] Y-S Chen and Y-W Lin ldquoMobicast routing protocol forunderwater sensor networksrdquo IEEE Sensors Journal vol 13 no2 pp 737ndash749 2013

[12] S Zhang D Li and J Chen ldquoA link-state based adaptivefeedback routing for underwater acoustic sensor networksrdquoIEEE Sensors Journal vol 13 no 11 pp 4402ndash4412 2013

[13] P K Chong and D Kim ldquoSurface-level path loss modeling forsensor networks in flat and irregular terrainrdquoACMTransactionson Sensor Networks vol 9 no 2 article 15 2013

[14] J Poncela M C Aguayo and P Otero ldquoWireless underwatercommunicationsrdquo Wireless Personal Communications vol 64no 3 pp 547ndash560 2012

[15] G Han J Jiang N Bao L Wan and M Guizani ldquoRoutingprotocols for underwater wireless sensor networksrdquo IEEE Com-munications Magazine vol 53 no 11 pp 72ndash78 2015

[16] S Wang L Chen H Hu Z Xue and W Pan ldquoUnderwaterlocalization and environment mapping using wireless robotsrdquoWireless Personal Communications vol 70 no 3 pp 1147ndash11702013

[17] C-M Chao and M-W Lu ldquoEnergy-efficient transmissions forbursty traffic in underwater sensor networksrdquo InternationalJournal of Ad Hoc and Ubiquitous Computing vol 13 no 1 pp1ndash9 2013

[18] A Sanchez S Blanc P Yuste A Perles and J J Serrano ldquoAnultra-low power and flexible acoustic modem design to developenergy-efficient underwater sensor networksrdquo Sensors vol 12no 6 pp 6837ndash6856 2012

[19] Y Wang D Wang Q Lu D Luo and W Fang ldquoAquatic debrisdetection using embedded camera sensorsrdquo Sensors vol 15 no2 pp 3116ndash3137 2015

[20] Z Liu H GaoWWang S Chang and J Chen ldquoColor filteringlocalization for three dimensional underwater acoustic sensornetworksrdquo Sensors vol 15 no 3 pp 6009ndash6032 2015

[21] M Xu G Liu and H Wu ldquoAn energy-efficient routingalgorithm for underwater wireless sensor networks inspiredby ultrasonic frogsrdquo International Journal of Distributed SensorNetworks vol 2014 Article ID 351520 12 pages 2014

[22] J Cao J Dou and S Dong ldquoBalance transmission mechanismin underwater acoustic sensor networksrdquo International Journalof Distributed Sensor Networks vol 2015 Article ID 429340 12pages 2015

[23] H Chen Y Li J L Rebelatto B F Uchoa-Filho and B VuceticldquoHarvest-then-cooperate wireless-powered cooperative com-municationsrdquo IEEE Transactions on Signal Processing vol 63no 7 pp 1700ndash1711 2015

[24] S Wang H Liu and K Liu ldquoAn improved clustering cooper-ative spectrum sensing algorithm based on modified double-threshold energy detection and its optimization in cognitivewireless sensor networksrdquo International Journal of DistributedSensor Networks vol 2015 Article ID 136948 7 pages 2015

[25] M E Tipping and C M Bishop ldquoProbabilistic principalcomponent analysisrdquo Journal of the Royal Statistical SocietySeries B (Statistical Methodology) vol 61 no 3 pp 611ndash6221999

[26] R J Urick Principles of Underwater Sound for Engineers TataMcGraw-Hill Education 1967

[27] L Berkhovskikh and Y Lysanov Fundamentals of Ocean Acous-tics Springer New York NY USA 1982

[28] W Zhang and U Mitra ldquoA delay-reliability analysis for multi-hop underwater acoustic communicationrdquo in Proceedings of the2nd Workshop on Underwater Networks (WuWNet rsquo07) pp 57ndash64 ACM Montreal Canada September 2007

[29] A Khandani J Abounadi E Modiano and L Zheng ldquoReliabil-ity and route diversity in wireless networksrdquo IEEE TransactionsonWireless Communications vol 7 no 12 pp 4772ndash4776 2008

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

RoboticsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014


Active and Passive Electronic Components

Control Scienceand Engineering

Journal of



RotatingMachinery


Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Submit your manuscripts athttpwwwhindawicom

VLSI Design



Shock and Vibration


Civil EngineeringAdvances in

Acoustics and VibrationAdvances in



Electrical and Computer Engineering

Journal of

Advances inOptoElectronics


Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

SensorsJournal of


Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014


Chemical EngineeringInternational Journal of Antennas and

Propagation




Navigation and Observation



DistributedSensor Networks



have to be taken into account for the design of UWAwirelesssystems

Having now sufficient technological advancements madein the field of radio communications researches are tryingto enhance the working of UW systems using modern tech-niques adopted from radio communications A promisingtechnique is cooperative communication already being usedin terrestrial WSNs It is a potential approach for distributedUWSNs to upgrade the quality of link connecting sensors aswell as the reliability in both point-to-point and multipointenvironments having multiple relays doing cooperationWireless network designs take into account the diversityto improve the overall successful transmissions by allowingduplicate signals at the receiver In contrast to this approachMultiple-Input Multiple-Output (MIMO) technique alsouses a promising mechanism to improve Signal-to-NoiseRatio (SNR) by enhancing diversity gain But the techniqueneeds extra equipment cost for each sensor with muchcomplexity A different approach for gaining diversity is toutilize several sensors cooperatingwith each other to upgradethe quality of communication channel In variation to anindividual sensor having an antenna array duplicate data isforwarded by an array of distributed antennas comprised ofseveral sensors to reach the end-point but introducing somedelay Spatial diversity concept has directed regular efforts toits use in wireless networks particularly in WSNs If variouspaths existing between two end devices are not dependent oneach other and have adequate working channel efficiency canbe improved by forwarding multiple duplicates of data alongthese links and merging them at the destination The totalerror probability decreases since the paths are independentwhich makes the channel and system performance increaseHowever in general case of WSNs nodes are very small tothat needed for the support of such distributed antennasHence to combat such issues the idea of cooperative routingis proposed Cooperation is defined as a group of entitiesworking together to achieve a common goal while sharingeach otherrsquos resources In these systems transmitter forwardsone copy of data packets to a node acting as relay The relaythen decodes or amplifies each data packet as the schemesuggests and reforwards it to the final receiver Relay nodeuses a path which is generally different from the direct pathThe destinationmerges or utilizes both of the received signalsto extract the forwarded information

Cooperative diversity an alternative to combat fadingin wireless channels allows distributed users to help relayinformation of each other to explore inherent spatial diver-sity present in channels Various cooperation-based pro-tocols have been proposed in literature like fixed relayingprotocol adaptive relaying protocol user cooperation pro-tocol and coded cooperation schemes In fixed relayingschemes such as Amplify-and-Forward (AF) and Decode-and-Forward (DF) relays provide assistance to providethe source information In AF scheme the relays amplifyand forward the information whereas in DF scheme therelays decode the received information and then transmitit to the receiver Although in DF the relay forwardsthe decoded information to the receiver however thisscheme has adequate degradation in performance if the

relay does not decode the transmitters information prop-erly

2 Related Work

Earlier efforts to investigate the underwater behavior werejust depending on the technology available for terrestrialsensor networks However UWSNs show several structuraldesign differences in comparison to terrestrial networksespecially due to the water used as the transmission mediumand the signal required for data transmission Design of asuitable structure for UWSNs is also complex due to thebehavior of communication system

An efficient scheme for UWSN Coop (Re and dth) [1]employs cooperative routing which involves data transmis-sion via partner noderelay towards sink In this paper twodifferent partner node selection criteria are implementedand compared The authors have considered source nodedepth-threshold (dth) potential relays depth and residualenergy (Re) as one criterion and SNR of the link connectingsource node with relay or destination as another criterionfor selection parameters Cog-Coop [2] is an efficient schemefor maximization of network lifetime using residual energyof the nodes It improves the spectrum sensing performancealong with having better energy consumption of the sensorsOptimal conditions are attained based on the standard opti-mizationmethods to find the priority of sensors for spectrumsensingThey showed that cooperation among cognitive sen-sors is necessary for lowering the fading as well as shadowingeffects and hence correct sensing Cooperative Depth-BasedRouting (CoDBR) for UWSNs [3] is a cooperation-basedrouting protocol proposed to enhance network performanceEfficient nodes for relaying are selected on the basis ofinformation of depth Source node transmits data to thesink node by relay sensors using cooperation In [4] authorspropose forwarding function based routing scheme iAMCTDfor UWSNs The reactive protocol enhances the underwaternetwork lifetime by the best possible mobility design of sink

A communication path based routing scheme calledRelative Distance Based Forwarding (RDBF) is presented in[5] The authors have used a mathematical factor to judgeand measure the level of suitability for a sensor to relaythe data packets Few nodes are involved in forwardingmechanism helping in reduction of end-to-end delay aswell as energy consumption The protocol also controls theforwarding time of the multiple transmitters to minimizethe duplicate data transmission In [6] the authors havepresented routing protocols for cylindrical networks anddeveloped mathematical models to achieve best selection ofpath for information forwarding Their results showed thatthe suggested 4-chain based protocol performs efficiently interms of lifetime of the network path-loss end-to-end delayand packet sending rate

The problem of tracking UW moving targets is tackledin [7] For 3-dimensional UW maneuvering target track-ing the cooperating model technique is added with theparticle filter to handle with qualms Simulation outcomesexplain that the suggested scheme is a good alternatefor customary sensor-based or imaging-based approaches








3 Motivation
















between 119878119899and 119877

119898 denoted by 119862(119878

119899 119877119898) is the minimum


119878119899= 1199041 1199042 119904

119899 and 119877

119898= 1199031 1199032 119903

119898



S

D

R3












1199041 and 119877



119899119898and phase delay 120601

119899119898as shown in



119909 (119905) = 120572119890119895120601

119891 (119905) + 119873 (119905) (1)


119873 Net




to-point 119862119901minus119901

(1199041 1199031) is given by [24]

119862119901minus119901

(1199041 1199031) = 119875119879=SNRmin 119875119873

1205722

119875119879prop

1

1205722=

1

1198892

(2)

s1 r112057211

12060111


12057211

12057212

12057213

s1

1205721m

r1

r2

r3

rm


that is119862119901minus119901


1and 1199031


119899= 1199041 and 119877

119898= 1199031 1199032 119903

119898 as shown in


1is received


119879 Minimum


(1199041 119877119898) expressed as

119862119901minus119898

(1199041 119877119898) = maximize 119862

11(1199041 1199031) 11986212(1199041 1199032)

1198621119898

(1199041 119903119898)

(3)





119895are expressed






of the vector 119909119905= 119882119879

(V119905minus120583) where119882119879 = (119908

1 1199082 119908

119902)119879


matrix 119864[119909 sdot 119909119879



119899= 1199041 1199042 119904

119899 and 119877

119898= 1199031 as shown in



1205721n

r1

sn

12057211

12057212

12057213

s1

s2

s3





119897 = 119910 (119909 120579) + 120598 (4)



119897 = 119882119909 + 120583 + 120598 (5)





1198991and




2


119901 (119897119909) = (21205871205902

)minus1199052

exp minus1

21205902

1003817100381710038171003817119897 minus 119882119909 minus 1205831003817100381710038171003817

2

(6)


119901 (119909) = (2120587)minus1199022 exp minus1

2sdot 119909119879

sdot 119909 (7)


119901 (119909) = (2120587)minus12 exp minus1

2sdot 119909119879

sdot 119909 (8)


119901 (119897) = int119901 (119897119909) sdot 119901 (119909) sdot 119889119909 (9)

119901 (119909)

= (2120587)minus1199022


2sdot (119897 minus 120583)

119879

sdot 119862minus1

(119897 minus 120583)

(10)


119862 = 1205902

119868 + 119882119882119879

(11)


119894=1|119891119894|2 and the received




minimize119899

sum

119894=1

10038161003816100381610038161198911198941003816100381610038161003816

2

subject to1003816100381610038161003816sum119899

119894=11198911198941205721198941

1003816100381610038161003816

2

119875119873

ge SNRmin

1003816100381610038161003816100381610038161003816100381610038161003816

119899

sum

119894=1

1198911198941205721198941

1003816100381610038161003816100381610038161003816100381610038161003816

2

ge SNRmin 119875119873

10038161003816100381610038161198911198941003816100381610038161003816

2

1003816100381610038161003816100381610038161003816100381610038161003816

119899

sum

119894=1

1205721198941

1003816100381610038161003816100381610038161003816100381610038161003816

2

ge SNRmin 119875119873

(12)


10038161003816100381610038161198911198941003816100381610038161003816 =

1205721198941

1003816100381610038161003816sum119899

119894=11205721198941

1003816100381610038161003816

2

radicSNRmin 119875119873 (13)


(119878 1199031)


119862119898minus119901

(119878 1199031) = 119875119879=

1

sum119899

119894=1(1205722

1198941 (SNRmin 119875119873))

(14)


119899= 1199041 1199042 119904

119899 and 119877

119898= 1199031 1199032 119903

119898




r1

r2

r3

sn

s1

s2

s3

rm






119860 (119889 119891) = 1198600119889119896

119886 (119891)119889

(15)



10 log 119886 (119891) =011119891

2

1 + 1198912+

441198912

4200 + 119891+275119891

2

104+ 0003

(for 119891 gt 04) [dBkm]

10 log 119886 (119891) = 0002 +011119891

1 + 119891+ 0011119891

(for 119891 lt 04) [dBkm]

(16)



119904)



119873(119891) = 119873119905(119891) + 119873

119904(119891) + 119873

119908(119891) + 119873th (119891) (17)

where

10 log119873119905(119891) = 17 minus 30 log119891 (18)

10 log119873119904(119891) = 40 + 20 (119904 minus 05) + 26 log119891

minus 60 log (119891 + 003)

(19)

10 log119873119908(119891) = 50 + 75radic119908 + 20 log119891

minus 40 log (119891 + 04)

(20)

10 log119873th (119891) = minus15 + 20 log119891 (21)







119868 = (119905)

2120587119867 (23)




119879119889= 10 log

10119889 + 10

minus3

119886 (119891) 119889 (24)






Outage formulation

SNR computation

Follow relay path

No

Direct transfer

Yes


Satisfies the cost


Point-to-



Cooperation phase


and initialization


No

function criteria

Routing phase

Yes

SNR lt SNRmin


link

SNR ge SNRmin



119862 (119889 120588) = 119861 log2(1 + 120588 (119889 119891)) (25)








119862 (119889 120588) ge 119877 (26)


Outage = 119862 (119889 120588) lt 119877 (27)


119875outage = 119875 119862 (119889 120588) lt 119877 (28)

119875outage = 119875 119861 log2(1 + 120588 (119889 119891)) lt 119877 (29)

119875outage = 119875 120588 (119889 119891) lt 2(119877119861)

minus 1 (30)

119875outage = 11987520 log( (119905)

2120587119867) minus 119860 (119889 119891) minus 119873 (119891)

lt 2(119877119861)

minus 1

(31)

119875outage = 119875log( (119905)

2120587119867)

lt2(119877119861)

+ 119860 (119889 119891) + 119873 (119891) minus 1

20

(32)

119875outage = 119875 (119905) lt 2120587119867 anti

minus log(2(119877119861)

+ 119860 (119889 119891) + 119873 (119891) minus 1

20)

(33)


119875outage

= 119875 (119905) lt 2120587119867 exp(2(119877119861)

+ SL + 120588 (119889 119891)

20)

(34)



R

= 1

minus (119905) lt 2120587119867 exp(2(119877119861)

+ SL + 120588 (119889 119891)

20)

(36)


119899

119894=1(2120587119867119894exp((2(119877119861) + SL + 120588(119889 119891)

119894)20))














0 2000 4000 6000 8000 100000

50

100

150

200

250

Time (seconds)

Num

ber o

f aliv

e nod

es


CoDBRiAMCTD








0 2000 4000 6000 8000 100000

5

10

15

20

25

30

35

Time (seconds)

Tota

l ene

rgy

cons

umpt

ion

(J)


iAMCTDCoDBR












0 2000 4000 6000 8000 1000030

40

50

60

70

80

90

100

110

Time (seconds)

Pack

et d

eliv

ery

ratio


CoDBRiAMCTD




0 2000 4000 6000 8000 100000

50

100

150

200

250

Time (seconds)

Path

loss

(dB)


CoDBRiAMCTD










0 2000 4000 6000 8000 100000

20

40

60

80

100

120

140

160

180

Time (seconds)

Num

ber o

f dea

d no

des


CoDBRiAMCTD









0 2000 4000 6000 8000 100000

005

01

015

02

025

03

035

Time (seconds)

End-

to-e

nd d

elay

(sec

onds

)


CoDBRiAMCTD




0 5 10 15 20SNR (dB)

Out

age p

roba

bilit

y


10minus3

10minus2

10minus1

100








10 dBBER at SNR

15 dBBER at SNR

20 dBDirect 0125 0067 005 003 028Relay 0125 005 002 001 0004







10 Conclusion





SPARCO



CoDBR



iAMCTD



Cog-Coop



Coop (Reand dth)






References



































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation
















3 Motivation
















between 119878119899and 119877

119898 denoted by 119862(119878

119899 119877119898) is the minimum


119878119899= 1199041 1199042 119904

119899 and 119877

119898= 1199031 1199032 119903

119898



S

D

R3












1199041 and 119877



119899119898and phase delay 120601

119899119898as shown in



119909 (119905) = 120572119890119895120601

119891 (119905) + 119873 (119905) (1)


119873 Net




to-point 119862119901minus119901

(1199041 1199031) is given by [24]

119862119901minus119901

(1199041 1199031) = 119875119879=SNRmin 119875119873

1205722

119875119879prop

1

1205722=

1

1198892

(2)

s1 r112057211

12060111


12057211

12057212

12057213

s1

1205721m

r1

r2

r3

rm


that is119862119901minus119901


1and 1199031


119899= 1199041 and 119877

119898= 1199031 1199032 119903

119898 as shown in


1is received


119879 Minimum


(1199041 119877119898) expressed as

119862119901minus119898

(1199041 119877119898) = maximize 119862

11(1199041 1199031) 11986212(1199041 1199032)

1198621119898

(1199041 119903119898)

(3)





119895are expressed






of the vector 119909119905= 119882119879

(V119905minus120583) where119882119879 = (119908

1 1199082 119908

119902)119879


matrix 119864[119909 sdot 119909119879



119899= 1199041 1199042 119904

119899 and 119877

119898= 1199031 as shown in



1205721n

r1

sn

12057211

12057212

12057213

s1

s2

s3





119897 = 119910 (119909 120579) + 120598 (4)



119897 = 119882119909 + 120583 + 120598 (5)





1198991and




2


119901 (119897119909) = (21205871205902

)minus1199052

exp minus1

21205902

1003817100381710038171003817119897 minus 119882119909 minus 1205831003817100381710038171003817

2

(6)


119901 (119909) = (2120587)minus1199022 exp minus1

2sdot 119909119879

sdot 119909 (7)


119901 (119909) = (2120587)minus12 exp minus1

2sdot 119909119879

sdot 119909 (8)


119901 (119897) = int119901 (119897119909) sdot 119901 (119909) sdot 119889119909 (9)

119901 (119909)

= (2120587)minus1199022


2sdot (119897 minus 120583)

119879

sdot 119862minus1

(119897 minus 120583)

(10)


119862 = 1205902

119868 + 119882119882119879

(11)


119894=1|119891119894|2 and the received




minimize119899

sum

119894=1

10038161003816100381610038161198911198941003816100381610038161003816

2

subject to1003816100381610038161003816sum119899

119894=11198911198941205721198941

1003816100381610038161003816

2

119875119873

ge SNRmin

1003816100381610038161003816100381610038161003816100381610038161003816

119899

sum

119894=1

1198911198941205721198941

1003816100381610038161003816100381610038161003816100381610038161003816

2

ge SNRmin 119875119873

10038161003816100381610038161198911198941003816100381610038161003816

2

1003816100381610038161003816100381610038161003816100381610038161003816

119899

sum

119894=1

1205721198941

1003816100381610038161003816100381610038161003816100381610038161003816

2

ge SNRmin 119875119873

(12)


10038161003816100381610038161198911198941003816100381610038161003816 =

1205721198941

1003816100381610038161003816sum119899

119894=11205721198941

1003816100381610038161003816

2

radicSNRmin 119875119873 (13)


(119878 1199031)


119862119898minus119901

(119878 1199031) = 119875119879=

1

sum119899

119894=1(1205722

1198941 (SNRmin 119875119873))

(14)


119899= 1199041 1199042 119904

119899 and 119877

119898= 1199031 1199032 119903

119898




r1

r2

r3

sn

s1

s2

s3

rm






119860 (119889 119891) = 1198600119889119896

119886 (119891)119889

(15)



10 log 119886 (119891) =011119891

2

1 + 1198912+

441198912

4200 + 119891+275119891

2

104+ 0003

(for 119891 gt 04) [dBkm]

10 log 119886 (119891) = 0002 +011119891

1 + 119891+ 0011119891

(for 119891 lt 04) [dBkm]

(16)



119904)



119873(119891) = 119873119905(119891) + 119873

119904(119891) + 119873

119908(119891) + 119873th (119891) (17)

where

10 log119873119905(119891) = 17 minus 30 log119891 (18)

10 log119873119904(119891) = 40 + 20 (119904 minus 05) + 26 log119891

minus 60 log (119891 + 003)

(19)

10 log119873119908(119891) = 50 + 75radic119908 + 20 log119891

minus 40 log (119891 + 04)

(20)

10 log119873th (119891) = minus15 + 20 log119891 (21)







119868 = (119905)

2120587119867 (23)




119879119889= 10 log

10119889 + 10

minus3

119886 (119891) 119889 (24)






Outage formulation

SNR computation

Follow relay path

No

Direct transfer

Yes


Satisfies the cost


Point-to-



Cooperation phase


and initialization


No

function criteria

Routing phase

Yes

SNR lt SNRmin


link

SNR ge SNRmin



119862 (119889 120588) = 119861 log2(1 + 120588 (119889 119891)) (25)








119862 (119889 120588) ge 119877 (26)


Outage = 119862 (119889 120588) lt 119877 (27)


119875outage = 119875 119862 (119889 120588) lt 119877 (28)

119875outage = 119875 119861 log2(1 + 120588 (119889 119891)) lt 119877 (29)

119875outage = 119875 120588 (119889 119891) lt 2(119877119861)

minus 1 (30)

119875outage = 11987520 log( (119905)

2120587119867) minus 119860 (119889 119891) minus 119873 (119891)

lt 2(119877119861)

minus 1

(31)

119875outage = 119875log( (119905)

2120587119867)

lt2(119877119861)

+ 119860 (119889 119891) + 119873 (119891) minus 1

20

(32)

119875outage = 119875 (119905) lt 2120587119867 anti

minus log(2(119877119861)

+ 119860 (119889 119891) + 119873 (119891) minus 1

20)

(33)


119875outage

= 119875 (119905) lt 2120587119867 exp(2(119877119861)

+ SL + 120588 (119889 119891)

20)

(34)



R

= 1

minus (119905) lt 2120587119867 exp(2(119877119861)

+ SL + 120588 (119889 119891)

20)

(36)


119899

119894=1(2120587119867119894exp((2(119877119861) + SL + 120588(119889 119891)

119894)20))














0 2000 4000 6000 8000 100000

50

100

150

200

250

Time (seconds)

Num

ber o

f aliv

e nod

es


CoDBRiAMCTD








0 2000 4000 6000 8000 100000

5

10

15

20

25

30

35

Time (seconds)

Tota

l ene

rgy

cons

umpt

ion

(J)


iAMCTDCoDBR












0 2000 4000 6000 8000 1000030

40

50

60

70

80

90

100

110

Time (seconds)

Pack

et d

eliv

ery

ratio


CoDBRiAMCTD




0 2000 4000 6000 8000 100000

50

100

150

200

250

Time (seconds)

Path

loss

(dB)


CoDBRiAMCTD










0 2000 4000 6000 8000 100000

20

40

60

80

100

120

140

160

180

Time (seconds)

Num

ber o

f dea

d no

des


CoDBRiAMCTD









0 2000 4000 6000 8000 100000

005

01

015

02

025

03

035

Time (seconds)

End-

to-e

nd d

elay

(sec

onds

)


CoDBRiAMCTD




0 5 10 15 20SNR (dB)

Out

age p

roba

bilit

y


10minus3

10minus2

10minus1

100








10 dBBER at SNR

15 dBBER at SNR

20 dBDirect 0125 0067 005 003 028Relay 0125 005 002 001 0004







10 Conclusion





SPARCO



CoDBR



iAMCTD



Cog-Coop



Coop (Reand dth)






References



































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation























between 119878119899and 119877

119898 denoted by 119862(119878

119899 119877119898) is the minimum


119878119899= 1199041 1199042 119904

119899 and 119877

119898= 1199031 1199032 119903

119898



S

D

R3












1199041 and 119877



119899119898and phase delay 120601

119899119898as shown in



119909 (119905) = 120572119890119895120601

119891 (119905) + 119873 (119905) (1)


119873 Net




to-point 119862119901minus119901

(1199041 1199031) is given by [24]

119862119901minus119901

(1199041 1199031) = 119875119879=SNRmin 119875119873

1205722

119875119879prop

1

1205722=

1

1198892

(2)

s1 r112057211

12060111


12057211

12057212

12057213

s1

1205721m

r1

r2

r3

rm


that is119862119901minus119901


1and 1199031


119899= 1199041 and 119877

119898= 1199031 1199032 119903

119898 as shown in


1is received


119879 Minimum


(1199041 119877119898) expressed as

119862119901minus119898

(1199041 119877119898) = maximize 119862

11(1199041 1199031) 11986212(1199041 1199032)

1198621119898

(1199041 119903119898)

(3)





119895are expressed






of the vector 119909119905= 119882119879

(V119905minus120583) where119882119879 = (119908

1 1199082 119908

119902)119879


matrix 119864[119909 sdot 119909119879



119899= 1199041 1199042 119904

119899 and 119877

119898= 1199031 as shown in



1205721n

r1

sn

12057211

12057212

12057213

s1

s2

s3





119897 = 119910 (119909 120579) + 120598 (4)



119897 = 119882119909 + 120583 + 120598 (5)





1198991and




2


119901 (119897119909) = (21205871205902

)minus1199052

exp minus1

21205902

1003817100381710038171003817119897 minus 119882119909 minus 1205831003817100381710038171003817

2

(6)


119901 (119909) = (2120587)minus1199022 exp minus1

2sdot 119909119879

sdot 119909 (7)


119901 (119909) = (2120587)minus12 exp minus1

2sdot 119909119879

sdot 119909 (8)


119901 (119897) = int119901 (119897119909) sdot 119901 (119909) sdot 119889119909 (9)

119901 (119909)

= (2120587)minus1199022


2sdot (119897 minus 120583)

119879

sdot 119862minus1

(119897 minus 120583)

(10)


119862 = 1205902

119868 + 119882119882119879

(11)


119894=1|119891119894|2 and the received




minimize119899

sum

119894=1

10038161003816100381610038161198911198941003816100381610038161003816

2

subject to1003816100381610038161003816sum119899

119894=11198911198941205721198941

1003816100381610038161003816

2

119875119873

ge SNRmin

1003816100381610038161003816100381610038161003816100381610038161003816

119899

sum

119894=1

1198911198941205721198941

1003816100381610038161003816100381610038161003816100381610038161003816

2

ge SNRmin 119875119873

10038161003816100381610038161198911198941003816100381610038161003816

2

1003816100381610038161003816100381610038161003816100381610038161003816

119899

sum

119894=1

1205721198941

1003816100381610038161003816100381610038161003816100381610038161003816

2

ge SNRmin 119875119873

(12)


10038161003816100381610038161198911198941003816100381610038161003816 =

1205721198941

1003816100381610038161003816sum119899

119894=11205721198941

1003816100381610038161003816

2

radicSNRmin 119875119873 (13)


(119878 1199031)


119862119898minus119901

(119878 1199031) = 119875119879=

1

sum119899

119894=1(1205722

1198941 (SNRmin 119875119873))

(14)


119899= 1199041 1199042 119904

119899 and 119877

119898= 1199031 1199032 119903

119898




r1

r2

r3

sn

s1

s2

s3

rm






119860 (119889 119891) = 1198600119889119896

119886 (119891)119889

(15)



10 log 119886 (119891) =011119891

2

1 + 1198912+

441198912

4200 + 119891+275119891

2

104+ 0003

(for 119891 gt 04) [dBkm]

10 log 119886 (119891) = 0002 +011119891

1 + 119891+ 0011119891

(for 119891 lt 04) [dBkm]

(16)



119904)



119873(119891) = 119873119905(119891) + 119873

119904(119891) + 119873

119908(119891) + 119873th (119891) (17)

where

10 log119873119905(119891) = 17 minus 30 log119891 (18)

10 log119873119904(119891) = 40 + 20 (119904 minus 05) + 26 log119891

minus 60 log (119891 + 003)

(19)

10 log119873119908(119891) = 50 + 75radic119908 + 20 log119891

minus 40 log (119891 + 04)

(20)

10 log119873th (119891) = minus15 + 20 log119891 (21)







119868 = (119905)

2120587119867 (23)




119879119889= 10 log

10119889 + 10

minus3

119886 (119891) 119889 (24)






Outage formulation

SNR computation

Follow relay path

No

Direct transfer

Yes


Satisfies the cost


Point-to-



Cooperation phase


and initialization


No

function criteria

Routing phase

Yes

SNR lt SNRmin


link

SNR ge SNRmin



119862 (119889 120588) = 119861 log2(1 + 120588 (119889 119891)) (25)








119862 (119889 120588) ge 119877 (26)


Outage = 119862 (119889 120588) lt 119877 (27)


119875outage = 119875 119862 (119889 120588) lt 119877 (28)

119875outage = 119875 119861 log2(1 + 120588 (119889 119891)) lt 119877 (29)

119875outage = 119875 120588 (119889 119891) lt 2(119877119861)

minus 1 (30)

119875outage = 11987520 log( (119905)

2120587119867) minus 119860 (119889 119891) minus 119873 (119891)

lt 2(119877119861)

minus 1

(31)

119875outage = 119875log( (119905)

2120587119867)

lt2(119877119861)

+ 119860 (119889 119891) + 119873 (119891) minus 1

20

(32)

119875outage = 119875 (119905) lt 2120587119867 anti

minus log(2(119877119861)

+ 119860 (119889 119891) + 119873 (119891) minus 1

20)

(33)


119875outage

= 119875 (119905) lt 2120587119867 exp(2(119877119861)

+ SL + 120588 (119889 119891)

20)

(34)



R

= 1

minus (119905) lt 2120587119867 exp(2(119877119861)

+ SL + 120588 (119889 119891)

20)

(36)


119899

119894=1(2120587119867119894exp((2(119877119861) + SL + 120588(119889 119891)

119894)20))














0 2000 4000 6000 8000 100000

50

100

150

200

250

Time (seconds)

Num

ber o

f aliv

e nod

es


CoDBRiAMCTD








0 2000 4000 6000 8000 100000

5

10

15

20

25

30

35

Time (seconds)

Tota

l ene

rgy

cons

umpt

ion

(J)


iAMCTDCoDBR












0 2000 4000 6000 8000 1000030

40

50

60

70

80

90

100

110

Time (seconds)

Pack

et d

eliv

ery

ratio


CoDBRiAMCTD




0 2000 4000 6000 8000 100000

50

100

150

200

250

Time (seconds)

Path

loss

(dB)


CoDBRiAMCTD










0 2000 4000 6000 8000 100000

20

40

60

80

100

120

140

160

180

Time (seconds)

Num

ber o

f dea

d no

des


CoDBRiAMCTD









0 2000 4000 6000 8000 100000

005

01

015

02

025

03

035

Time (seconds)

End-

to-e

nd d

elay

(sec

onds

)


CoDBRiAMCTD




0 5 10 15 20SNR (dB)

Out

age p

roba

bilit

y


10minus3

10minus2

10minus1

100








10 dBBER at SNR

15 dBBER at SNR

20 dBDirect 0125 0067 005 003 028Relay 0125 005 002 001 0004







10 Conclusion





SPARCO



CoDBR



iAMCTD



Cog-Coop



Coop (Reand dth)






References



































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation










S

D

R3












1199041 and 119877



119899119898and phase delay 120601

119899119898as shown in



119909 (119905) = 120572119890119895120601

119891 (119905) + 119873 (119905) (1)


119873 Net




to-point 119862119901minus119901

(1199041 1199031) is given by [24]

119862119901minus119901

(1199041 1199031) = 119875119879=SNRmin 119875119873

1205722

119875119879prop

1

1205722=

1

1198892

(2)

s1 r112057211

12060111


12057211

12057212

12057213

s1

1205721m

r1

r2

r3

rm


that is119862119901minus119901


1and 1199031


119899= 1199041 and 119877

119898= 1199031 1199032 119903

119898 as shown in


1is received


119879 Minimum


(1199041 119877119898) expressed as

119862119901minus119898

(1199041 119877119898) = maximize 119862

11(1199041 1199031) 11986212(1199041 1199032)

1198621119898

(1199041 119903119898)

(3)





119895are expressed






of the vector 119909119905= 119882119879

(V119905minus120583) where119882119879 = (119908

1 1199082 119908

119902)119879


matrix 119864[119909 sdot 119909119879



119899= 1199041 1199042 119904

119899 and 119877

119898= 1199031 as shown in



1205721n

r1

sn

12057211

12057212

12057213

s1

s2

s3





119897 = 119910 (119909 120579) + 120598 (4)



119897 = 119882119909 + 120583 + 120598 (5)





1198991and




2


119901 (119897119909) = (21205871205902

)minus1199052

exp minus1

21205902

1003817100381710038171003817119897 minus 119882119909 minus 1205831003817100381710038171003817

2

(6)


119901 (119909) = (2120587)minus1199022 exp minus1

2sdot 119909119879

sdot 119909 (7)


119901 (119909) = (2120587)minus12 exp minus1

2sdot 119909119879

sdot 119909 (8)


119901 (119897) = int119901 (119897119909) sdot 119901 (119909) sdot 119889119909 (9)

119901 (119909)

= (2120587)minus1199022


2sdot (119897 minus 120583)

119879

sdot 119862minus1

(119897 minus 120583)

(10)


119862 = 1205902

119868 + 119882119882119879

(11)


119894=1|119891119894|2 and the received




minimize119899

sum

119894=1

10038161003816100381610038161198911198941003816100381610038161003816

2

subject to1003816100381610038161003816sum119899

119894=11198911198941205721198941

1003816100381610038161003816

2

119875119873

ge SNRmin

1003816100381610038161003816100381610038161003816100381610038161003816

119899

sum

119894=1

1198911198941205721198941

1003816100381610038161003816100381610038161003816100381610038161003816

2

ge SNRmin 119875119873

10038161003816100381610038161198911198941003816100381610038161003816

2

1003816100381610038161003816100381610038161003816100381610038161003816

119899

sum

119894=1

1205721198941

1003816100381610038161003816100381610038161003816100381610038161003816

2

ge SNRmin 119875119873

(12)


10038161003816100381610038161198911198941003816100381610038161003816 =

1205721198941

1003816100381610038161003816sum119899

119894=11205721198941

1003816100381610038161003816

2

radicSNRmin 119875119873 (13)


(119878 1199031)


119862119898minus119901

(119878 1199031) = 119875119879=

1

sum119899

119894=1(1205722

1198941 (SNRmin 119875119873))

(14)


119899= 1199041 1199042 119904

119899 and 119877

119898= 1199031 1199032 119903

119898




r1

r2

r3

sn

s1

s2

s3

rm






119860 (119889 119891) = 1198600119889119896

119886 (119891)119889

(15)



10 log 119886 (119891) =011119891

2

1 + 1198912+

441198912

4200 + 119891+275119891

2

104+ 0003

(for 119891 gt 04) [dBkm]

10 log 119886 (119891) = 0002 +011119891

1 + 119891+ 0011119891

(for 119891 lt 04) [dBkm]

(16)



119904)



119873(119891) = 119873119905(119891) + 119873

119904(119891) + 119873

119908(119891) + 119873th (119891) (17)

where

10 log119873119905(119891) = 17 minus 30 log119891 (18)

10 log119873119904(119891) = 40 + 20 (119904 minus 05) + 26 log119891

minus 60 log (119891 + 003)

(19)

10 log119873119908(119891) = 50 + 75radic119908 + 20 log119891

minus 40 log (119891 + 04)

(20)

10 log119873th (119891) = minus15 + 20 log119891 (21)







119868 = (119905)

2120587119867 (23)




119879119889= 10 log

10119889 + 10

minus3

119886 (119891) 119889 (24)






Outage formulation

SNR computation

Follow relay path

No

Direct transfer

Yes


Satisfies the cost


Point-to-



Cooperation phase


and initialization


No

function criteria

Routing phase

Yes

SNR lt SNRmin


link

SNR ge SNRmin



119862 (119889 120588) = 119861 log2(1 + 120588 (119889 119891)) (25)








119862 (119889 120588) ge 119877 (26)


Outage = 119862 (119889 120588) lt 119877 (27)


119875outage = 119875 119862 (119889 120588) lt 119877 (28)

119875outage = 119875 119861 log2(1 + 120588 (119889 119891)) lt 119877 (29)

119875outage = 119875 120588 (119889 119891) lt 2(119877119861)

minus 1 (30)

119875outage = 11987520 log( (119905)

2120587119867) minus 119860 (119889 119891) minus 119873 (119891)

lt 2(119877119861)

minus 1

(31)

119875outage = 119875log( (119905)

2120587119867)

lt2(119877119861)

+ 119860 (119889 119891) + 119873 (119891) minus 1

20

(32)

119875outage = 119875 (119905) lt 2120587119867 anti

minus log(2(119877119861)

+ 119860 (119889 119891) + 119873 (119891) minus 1

20)

(33)


119875outage

= 119875 (119905) lt 2120587119867 exp(2(119877119861)

+ SL + 120588 (119889 119891)

20)

(34)



R

= 1

minus (119905) lt 2120587119867 exp(2(119877119861)

+ SL + 120588 (119889 119891)

20)

(36)


119899

119894=1(2120587119867119894exp((2(119877119861) + SL + 120588(119889 119891)

119894)20))














0 2000 4000 6000 8000 100000

50

100

150

200

250

Time (seconds)

Num

ber o

f aliv

e nod

es


CoDBRiAMCTD








0 2000 4000 6000 8000 100000

5

10

15

20

25

30

35

Time (seconds)

Tota

l ene

rgy

cons

umpt

ion

(J)


iAMCTDCoDBR












0 2000 4000 6000 8000 1000030

40

50

60

70

80

90

100

110

Time (seconds)

Pack

et d

eliv

ery

ratio


CoDBRiAMCTD




0 2000 4000 6000 8000 100000

50

100

150

200

250

Time (seconds)

Path

loss

(dB)


CoDBRiAMCTD










0 2000 4000 6000 8000 100000

20

40

60

80

100

120

140

160

180

Time (seconds)

Num

ber o

f dea

d no

des


CoDBRiAMCTD









0 2000 4000 6000 8000 100000

005

01

015

02

025

03

035

Time (seconds)

End-

to-e

nd d

elay

(sec

onds

)


CoDBRiAMCTD




0 5 10 15 20SNR (dB)

Out

age p

roba

bilit

y


10minus3

10minus2

10minus1

100








10 dBBER at SNR

15 dBBER at SNR

20 dBDirect 0125 0067 005 003 028Relay 0125 005 002 001 0004







10 Conclusion





SPARCO



CoDBR



iAMCTD



Cog-Coop



Coop (Reand dth)






References



































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation










1205721n

r1

sn

12057211

12057212

12057213

s1

s2

s3





119897 = 119910 (119909 120579) + 120598 (4)



119897 = 119882119909 + 120583 + 120598 (5)





1198991and




2


119901 (119897119909) = (21205871205902

)minus1199052

exp minus1

21205902

1003817100381710038171003817119897 minus 119882119909 minus 1205831003817100381710038171003817

2

(6)


119901 (119909) = (2120587)minus1199022 exp minus1

2sdot 119909119879

sdot 119909 (7)


119901 (119909) = (2120587)minus12 exp minus1

2sdot 119909119879

sdot 119909 (8)


119901 (119897) = int119901 (119897119909) sdot 119901 (119909) sdot 119889119909 (9)

119901 (119909)

= (2120587)minus1199022


2sdot (119897 minus 120583)

119879

sdot 119862minus1

(119897 minus 120583)

(10)


119862 = 1205902

119868 + 119882119882119879

(11)


119894=1|119891119894|2 and the received




minimize119899

sum

119894=1

10038161003816100381610038161198911198941003816100381610038161003816

2

subject to1003816100381610038161003816sum119899

119894=11198911198941205721198941

1003816100381610038161003816

2

119875119873

ge SNRmin

1003816100381610038161003816100381610038161003816100381610038161003816

119899

sum

119894=1

1198911198941205721198941

1003816100381610038161003816100381610038161003816100381610038161003816

2

ge SNRmin 119875119873

10038161003816100381610038161198911198941003816100381610038161003816

2

1003816100381610038161003816100381610038161003816100381610038161003816

119899

sum

119894=1

1205721198941

1003816100381610038161003816100381610038161003816100381610038161003816

2

ge SNRmin 119875119873

(12)


10038161003816100381610038161198911198941003816100381610038161003816 =

1205721198941

1003816100381610038161003816sum119899

119894=11205721198941

1003816100381610038161003816

2

radicSNRmin 119875119873 (13)


(119878 1199031)


119862119898minus119901

(119878 1199031) = 119875119879=

1

sum119899

119894=1(1205722

1198941 (SNRmin 119875119873))

(14)


119899= 1199041 1199042 119904

119899 and 119877

119898= 1199031 1199032 119903

119898




r1

r2

r3

sn

s1

s2

s3

rm






119860 (119889 119891) = 1198600119889119896

119886 (119891)119889

(15)



10 log 119886 (119891) =011119891

2

1 + 1198912+

441198912

4200 + 119891+275119891

2

104+ 0003

(for 119891 gt 04) [dBkm]

10 log 119886 (119891) = 0002 +011119891

1 + 119891+ 0011119891

(for 119891 lt 04) [dBkm]

(16)



119904)



119873(119891) = 119873119905(119891) + 119873

119904(119891) + 119873

119908(119891) + 119873th (119891) (17)

where

10 log119873119905(119891) = 17 minus 30 log119891 (18)

10 log119873119904(119891) = 40 + 20 (119904 minus 05) + 26 log119891

minus 60 log (119891 + 003)

(19)

10 log119873119908(119891) = 50 + 75radic119908 + 20 log119891

minus 40 log (119891 + 04)

(20)

10 log119873th (119891) = minus15 + 20 log119891 (21)







119868 = (119905)

2120587119867 (23)




119879119889= 10 log

10119889 + 10

minus3

119886 (119891) 119889 (24)






Outage formulation

SNR computation

Follow relay path

No

Direct transfer

Yes


Satisfies the cost


Point-to-



Cooperation phase


and initialization


No

function criteria

Routing phase

Yes

SNR lt SNRmin


link

SNR ge SNRmin



119862 (119889 120588) = 119861 log2(1 + 120588 (119889 119891)) (25)








119862 (119889 120588) ge 119877 (26)


Outage = 119862 (119889 120588) lt 119877 (27)


119875outage = 119875 119862 (119889 120588) lt 119877 (28)

119875outage = 119875 119861 log2(1 + 120588 (119889 119891)) lt 119877 (29)

119875outage = 119875 120588 (119889 119891) lt 2(119877119861)

minus 1 (30)

119875outage = 11987520 log( (119905)

2120587119867) minus 119860 (119889 119891) minus 119873 (119891)

lt 2(119877119861)

minus 1

(31)

119875outage = 119875log( (119905)

2120587119867)

lt2(119877119861)

+ 119860 (119889 119891) + 119873 (119891) minus 1

20

(32)

119875outage = 119875 (119905) lt 2120587119867 anti

minus log(2(119877119861)

+ 119860 (119889 119891) + 119873 (119891) minus 1

20)

(33)


119875outage

= 119875 (119905) lt 2120587119867 exp(2(119877119861)

+ SL + 120588 (119889 119891)

20)

(34)



R

= 1

minus (119905) lt 2120587119867 exp(2(119877119861)

+ SL + 120588 (119889 119891)

20)

(36)


119899

119894=1(2120587119867119894exp((2(119877119861) + SL + 120588(119889 119891)

119894)20))














0 2000 4000 6000 8000 100000

50

100

150

200

250

Time (seconds)

Num

ber o

f aliv

e nod

es


CoDBRiAMCTD








0 2000 4000 6000 8000 100000

5

10

15

20

25

30

35

Time (seconds)

Tota

l ene

rgy

cons

umpt

ion

(J)


iAMCTDCoDBR












0 2000 4000 6000 8000 1000030

40

50

60

70

80

90

100

110

Time (seconds)

Pack

et d

eliv

ery

ratio


CoDBRiAMCTD




0 2000 4000 6000 8000 100000

50

100

150

200

250

Time (seconds)

Path

loss

(dB)


CoDBRiAMCTD










0 2000 4000 6000 8000 100000

20

40

60

80

100

120

140

160

180

Time (seconds)

Num

ber o

f dea

d no

des


CoDBRiAMCTD









0 2000 4000 6000 8000 100000

005

01

015

02

025

03

035

Time (seconds)

End-

to-e

nd d

elay

(sec

onds

)


CoDBRiAMCTD




0 5 10 15 20SNR (dB)

Out

age p

roba

bilit

y


10minus3

10minus2

10minus1

100








10 dBBER at SNR

15 dBBER at SNR

20 dBDirect 0125 0067 005 003 028Relay 0125 005 002 001 0004







10 Conclusion





SPARCO



CoDBR



iAMCTD



Cog-Coop



Coop (Reand dth)






References



































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation










r1

r2

r3

sn

s1

s2

s3

rm






119860 (119889 119891) = 1198600119889119896

119886 (119891)119889

(15)



10 log 119886 (119891) =011119891

2

1 + 1198912+

441198912

4200 + 119891+275119891

2

104+ 0003

(for 119891 gt 04) [dBkm]

10 log 119886 (119891) = 0002 +011119891

1 + 119891+ 0011119891

(for 119891 lt 04) [dBkm]

(16)



119904)



119873(119891) = 119873119905(119891) + 119873

119904(119891) + 119873

119908(119891) + 119873th (119891) (17)

where

10 log119873119905(119891) = 17 minus 30 log119891 (18)

10 log119873119904(119891) = 40 + 20 (119904 minus 05) + 26 log119891

minus 60 log (119891 + 003)

(19)

10 log119873119908(119891) = 50 + 75radic119908 + 20 log119891

minus 40 log (119891 + 04)

(20)

10 log119873th (119891) = minus15 + 20 log119891 (21)







119868 = (119905)

2120587119867 (23)




119879119889= 10 log

10119889 + 10

minus3

119886 (119891) 119889 (24)






Outage formulation

SNR computation

Follow relay path

No

Direct transfer

Yes


Satisfies the cost


Point-to-



Cooperation phase


and initialization


No

function criteria

Routing phase

Yes

SNR lt SNRmin


link

SNR ge SNRmin



119862 (119889 120588) = 119861 log2(1 + 120588 (119889 119891)) (25)








119862 (119889 120588) ge 119877 (26)


Outage = 119862 (119889 120588) lt 119877 (27)


119875outage = 119875 119862 (119889 120588) lt 119877 (28)

119875outage = 119875 119861 log2(1 + 120588 (119889 119891)) lt 119877 (29)

119875outage = 119875 120588 (119889 119891) lt 2(119877119861)

minus 1 (30)

119875outage = 11987520 log( (119905)

2120587119867) minus 119860 (119889 119891) minus 119873 (119891)

lt 2(119877119861)

minus 1

(31)

119875outage = 119875log( (119905)

2120587119867)

lt2(119877119861)

+ 119860 (119889 119891) + 119873 (119891) minus 1

20

(32)

119875outage = 119875 (119905) lt 2120587119867 anti

minus log(2(119877119861)

+ 119860 (119889 119891) + 119873 (119891) minus 1

20)

(33)


119875outage

= 119875 (119905) lt 2120587119867 exp(2(119877119861)

+ SL + 120588 (119889 119891)

20)

(34)



R

= 1

minus (119905) lt 2120587119867 exp(2(119877119861)

+ SL + 120588 (119889 119891)

20)

(36)


119899

119894=1(2120587119867119894exp((2(119877119861) + SL + 120588(119889 119891)

119894)20))














0 2000 4000 6000 8000 100000

50

100

150

200

250

Time (seconds)

Num

ber o

f aliv

e nod

es


CoDBRiAMCTD








0 2000 4000 6000 8000 100000

5

10

15

20

25

30

35

Time (seconds)

Tota

l ene

rgy

cons

umpt

ion

(J)


iAMCTDCoDBR












0 2000 4000 6000 8000 1000030

40

50

60

70

80

90

100

110

Time (seconds)

Pack

et d

eliv

ery

ratio


CoDBRiAMCTD




0 2000 4000 6000 8000 100000

50

100

150

200

250

Time (seconds)

Path

loss

(dB)


CoDBRiAMCTD










0 2000 4000 6000 8000 100000

20

40

60

80

100

120

140

160

180

Time (seconds)

Num

ber o

f dea

d no

des


CoDBRiAMCTD









0 2000 4000 6000 8000 100000

005

01

015

02

025

03

035

Time (seconds)

End-

to-e

nd d

elay

(sec

onds

)


CoDBRiAMCTD




0 5 10 15 20SNR (dB)

Out

age p

roba

bilit

y


10minus3

10minus2

10minus1

100








10 dBBER at SNR

15 dBBER at SNR

20 dBDirect 0125 0067 005 003 028Relay 0125 005 002 001 0004







10 Conclusion





SPARCO



CoDBR



iAMCTD



Cog-Coop



Coop (Reand dth)






References



































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation











Outage formulation

SNR computation

Follow relay path

No

Direct transfer

Yes


Satisfies the cost


Point-to-



Cooperation phase


and initialization


No

function criteria

Routing phase

Yes

SNR lt SNRmin


link

SNR ge SNRmin



119862 (119889 120588) = 119861 log2(1 + 120588 (119889 119891)) (25)








119862 (119889 120588) ge 119877 (26)


Outage = 119862 (119889 120588) lt 119877 (27)


119875outage = 119875 119862 (119889 120588) lt 119877 (28)

119875outage = 119875 119861 log2(1 + 120588 (119889 119891)) lt 119877 (29)

119875outage = 119875 120588 (119889 119891) lt 2(119877119861)

minus 1 (30)

119875outage = 11987520 log( (119905)

2120587119867) minus 119860 (119889 119891) minus 119873 (119891)

lt 2(119877119861)

minus 1

(31)

119875outage = 119875log( (119905)

2120587119867)

lt2(119877119861)

+ 119860 (119889 119891) + 119873 (119891) minus 1

20

(32)

119875outage = 119875 (119905) lt 2120587119867 anti

minus log(2(119877119861)

+ 119860 (119889 119891) + 119873 (119891) minus 1

20)

(33)


119875outage

= 119875 (119905) lt 2120587119867 exp(2(119877119861)

+ SL + 120588 (119889 119891)

20)

(34)



R

= 1

minus (119905) lt 2120587119867 exp(2(119877119861)

+ SL + 120588 (119889 119891)

20)

(36)


119899

119894=1(2120587119867119894exp((2(119877119861) + SL + 120588(119889 119891)

119894)20))














0 2000 4000 6000 8000 100000

50

100

150

200

250

Time (seconds)

Num

ber o

f aliv

e nod

es


CoDBRiAMCTD








0 2000 4000 6000 8000 100000

5

10

15

20

25

30

35

Time (seconds)

Tota

l ene

rgy

cons

umpt

ion

(J)


iAMCTDCoDBR












0 2000 4000 6000 8000 1000030

40

50

60

70

80

90

100

110

Time (seconds)

Pack

et d

eliv

ery

ratio


CoDBRiAMCTD




0 2000 4000 6000 8000 100000

50

100

150

200

250

Time (seconds)

Path

loss

(dB)


CoDBRiAMCTD










0 2000 4000 6000 8000 100000

20

40

60

80

100

120

140

160

180

Time (seconds)

Num

ber o

f dea

d no

des


CoDBRiAMCTD









0 2000 4000 6000 8000 100000

005

01

015

02

025

03

035

Time (seconds)

End-

to-e

nd d

elay

(sec

onds

)


CoDBRiAMCTD




0 5 10 15 20SNR (dB)

Out

age p

roba

bilit

y


10minus3

10minus2

10minus1

100








10 dBBER at SNR

15 dBBER at SNR

20 dBDirect 0125 0067 005 003 028Relay 0125 005 002 001 0004







10 Conclusion





SPARCO



CoDBR



iAMCTD



Cog-Coop



Coop (Reand dth)






References



































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation












119862 (119889 120588) ge 119877 (26)


Outage = 119862 (119889 120588) lt 119877 (27)


119875outage = 119875 119862 (119889 120588) lt 119877 (28)

119875outage = 119875 119861 log2(1 + 120588 (119889 119891)) lt 119877 (29)

119875outage = 119875 120588 (119889 119891) lt 2(119877119861)

minus 1 (30)

119875outage = 11987520 log( (119905)

2120587119867) minus 119860 (119889 119891) minus 119873 (119891)

lt 2(119877119861)

minus 1

(31)

119875outage = 119875log( (119905)

2120587119867)

lt2(119877119861)

+ 119860 (119889 119891) + 119873 (119891) minus 1

20

(32)

119875outage = 119875 (119905) lt 2120587119867 anti

minus log(2(119877119861)

+ 119860 (119889 119891) + 119873 (119891) minus 1

20)

(33)


119875outage

= 119875 (119905) lt 2120587119867 exp(2(119877119861)

+ SL + 120588 (119889 119891)

20)

(34)



R

= 1

minus (119905) lt 2120587119867 exp(2(119877119861)

+ SL + 120588 (119889 119891)

20)

(36)


119899

119894=1(2120587119867119894exp((2(119877119861) + SL + 120588(119889 119891)

119894)20))














0 2000 4000 6000 8000 100000

50

100

150

200

250

Time (seconds)

Num

ber o

f aliv

e nod

es


CoDBRiAMCTD








0 2000 4000 6000 8000 100000

5

10

15

20

25

30

35

Time (seconds)

Tota

l ene

rgy

cons

umpt

ion

(J)


iAMCTDCoDBR












0 2000 4000 6000 8000 1000030

40

50

60

70

80

90

100

110

Time (seconds)

Pack

et d

eliv

ery

ratio


CoDBRiAMCTD




0 2000 4000 6000 8000 100000

50

100

150

200

250

Time (seconds)

Path

loss

(dB)


CoDBRiAMCTD










0 2000 4000 6000 8000 100000

20

40

60

80

100

120

140

160

180

Time (seconds)

Num

ber o

f dea

d no

des


CoDBRiAMCTD









0 2000 4000 6000 8000 100000

005

01

015

02

025

03

035

Time (seconds)

End-

to-e

nd d

elay

(sec

onds

)


CoDBRiAMCTD




0 5 10 15 20SNR (dB)

Out

age p

roba

bilit

y


10minus3

10minus2

10minus1

100








10 dBBER at SNR

15 dBBER at SNR

20 dBDirect 0125 0067 005 003 028Relay 0125 005 002 001 0004







10 Conclusion





SPARCO



CoDBR



iAMCTD



Cog-Coop



Coop (Reand dth)






References



































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation














0 2000 4000 6000 8000 100000

50

100

150

200

250

Time (seconds)

Num

ber o

f aliv

e nod

es


CoDBRiAMCTD








0 2000 4000 6000 8000 100000

5

10

15

20

25

30

35

Time (seconds)

Tota

l ene

rgy

cons

umpt

ion

(J)


iAMCTDCoDBR












0 2000 4000 6000 8000 1000030

40

50

60

70

80

90

100

110

Time (seconds)

Pack

et d

eliv

ery

ratio


CoDBRiAMCTD




0 2000 4000 6000 8000 100000

50

100

150

200

250

Time (seconds)

Path

loss

(dB)


CoDBRiAMCTD










0 2000 4000 6000 8000 100000

20

40

60

80

100

120

140

160

180

Time (seconds)

Num

ber o

f dea

d no

des


CoDBRiAMCTD









0 2000 4000 6000 8000 100000

005

01

015

02

025

03

035

Time (seconds)

End-

to-e

nd d

elay

(sec

onds

)


CoDBRiAMCTD




0 5 10 15 20SNR (dB)

Out

age p

roba

bilit

y


10minus3

10minus2

10minus1

100








10 dBBER at SNR

15 dBBER at SNR

20 dBDirect 0125 0067 005 003 028Relay 0125 005 002 001 0004







10 Conclusion





SPARCO



CoDBR



iAMCTD



Cog-Coop



Coop (Reand dth)






References



































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation












0 2000 4000 6000 8000 100000

5

10

15

20

25

30

35

Time (seconds)

Tota

l ene

rgy

cons

umpt

ion

(J)


iAMCTDCoDBR












0 2000 4000 6000 8000 1000030

40

50

60

70

80

90

100

110

Time (seconds)

Pack

et d

eliv

ery

ratio


CoDBRiAMCTD




0 2000 4000 6000 8000 100000

50

100

150

200

250

Time (seconds)

Path

loss

(dB)


CoDBRiAMCTD










0 2000 4000 6000 8000 100000

20

40

60

80

100

120

140

160

180

Time (seconds)

Num

ber o

f dea

d no

des


CoDBRiAMCTD









0 2000 4000 6000 8000 100000

005

01

015

02

025

03

035

Time (seconds)

End-

to-e

nd d

elay

(sec

onds

)


CoDBRiAMCTD




0 5 10 15 20SNR (dB)

Out

age p

roba

bilit

y


10minus3

10minus2

10minus1

100








10 dBBER at SNR

15 dBBER at SNR

20 dBDirect 0125 0067 005 003 028Relay 0125 005 002 001 0004







10 Conclusion





SPARCO



CoDBR



iAMCTD



Cog-Coop



Coop (Reand dth)






References



































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation














0 2000 4000 6000 8000 1000030

40

50

60

70

80

90

100

110

Time (seconds)

Pack

et d

eliv

ery

ratio


CoDBRiAMCTD




0 2000 4000 6000 8000 100000

50

100

150

200

250

Time (seconds)

Path

loss

(dB)


CoDBRiAMCTD










0 2000 4000 6000 8000 100000

20

40

60

80

100

120

140

160

180

Time (seconds)

Num

ber o

f dea

d no

des


CoDBRiAMCTD









0 2000 4000 6000 8000 100000

005

01

015

02

025

03

035

Time (seconds)

End-

to-e

nd d

elay

(sec

onds

)


CoDBRiAMCTD




0 5 10 15 20SNR (dB)

Out

age p

roba

bilit

y


10minus3

10minus2

10minus1

100








10 dBBER at SNR

15 dBBER at SNR

20 dBDirect 0125 0067 005 003 028Relay 0125 005 002 001 0004







10 Conclusion





SPARCO



CoDBR



iAMCTD



Cog-Coop



Coop (Reand dth)






References



































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation















0 2000 4000 6000 8000 100000

20

40

60

80

100

120

140

160

180

Time (seconds)

Num

ber o

f dea

d no

des


CoDBRiAMCTD









0 2000 4000 6000 8000 100000

005

01

015

02

025

03

035

Time (seconds)

End-

to-e

nd d

elay

(sec

onds

)


CoDBRiAMCTD




0 5 10 15 20SNR (dB)

Out

age p

roba

bilit

y


10minus3

10minus2

10minus1

100








10 dBBER at SNR

15 dBBER at SNR

20 dBDirect 0125 0067 005 003 028Relay 0125 005 002 001 0004







10 Conclusion





SPARCO



CoDBR



iAMCTD



Cog-Coop



Coop (Reand dth)






References



































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation














0 2000 4000 6000 8000 100000

005

01

015

02

025

03

035

Time (seconds)

End-

to-e

nd d

elay

(sec

onds

)


CoDBRiAMCTD




0 5 10 15 20SNR (dB)

Out

age p

roba

bilit

y


10minus3

10minus2

10minus1

100








10 dBBER at SNR

15 dBBER at SNR

20 dBDirect 0125 0067 005 003 028Relay 0125 005 002 001 0004







10 Conclusion





SPARCO



CoDBR



iAMCTD



Cog-Coop



Coop (Reand dth)






References



































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation













10 dBBER at SNR

15 dBBER at SNR

20 dBDirect 0125 0067 005 003 028Relay 0125 005 002 001 0004







10 Conclusion





SPARCO



CoDBR



iAMCTD



Cog-Coop



Coop (Reand dth)






References



































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation












SPARCO



CoDBR



iAMCTD



Cog-Coop



Coop (Reand dth)






References



































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation

































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation









research article sparco: stochastic performance analysis...

Documents