review article principles, applications ... - son systems lab

Review ArticlePrinciples Applications and Challenges of Synchronization inNature for Future Mobile Communication Systems

Hyun-Ho Choi1 and Jung-Ryun Lee2

1Department of Electrical Electronic and Control Engineering and the Institute for Information Technology ConvergenceHankyong National University Anseong 17579 Republic of Korea2School of the Electrical Engineering Chung-Ang University Seoul 06974 Republic of Korea

Correspondence should be addressed to Jung-Ryun Lee jrleecauackr

Received 18 November 2016 Revised 13 December 2016 Accepted 22 December 2016 Published 22 January 2017

Academic Editor Yujin Lim

Copyright copy 2017 Hyun-Ho Choi and Jung-Ryun Lee This is an open access article distributed under the Creative CommonsAttribution License which permits unrestricted use distribution and reproduction in any medium provided the original work isproperly cited

In dynamic and complex natural environments a number of individuals such as fireflies flocks of birds schools of fish andpacemaker cells show various synchronization phenomena with the purpose of achieving their certain goals in an efficient anddistributed manner So far there has been considerable research attention on the working principles behind synchronizationphenomena in nature and as a result various models and theoretical investigation have been developed to apply synchronizationprinciples to various mobile communication systems In this article we present an exploration of synchronization phenomenain nature Some representative models on synchronization are investigated and its working principles are analyzed In additionwe survey some key applications inspired by synchronization principles for the future mobile communication systems Thecharacteristics and limitations of the applications inspired by synchronization in nature are evaluated in the context of the useof nature-inspired technologies Finally we provide the discussion of further research challenges for developing the advancedapplication of natural synchronization phenomena in the future mobile communication systems

1 Introduction

The inclination of living entities ranging from animalsto humans to synchronize with each other is the mostcommon tendency in the universe [1] Thousands of firefliessynchronously illuminate while geese fly at the same speed information Applause at concert hallsmerges to produce a har-monized sound in time and themenstrual periods of womenwho closely interact for a long time also synchronize Thou-sands of cardiac pacemaker cells in the heart fire in synchro-nization to sustain life Inanimate objects such as particlesand planets synchronize as well Laser beams are createdwhen trillions of atoms oscillating in sync emit photons ofthe same phase and frequency Moreover only one side ofthe moon can be viewed because the orbital and rotationalperiods of the moon are synchronized by the gravitationalpull between the earth and moon

As shown in the above examples synchronization occurswith both conscious living organisms and inanimate objects

without consciousness A key aspect of this phenomenon isthat it does not involve a leader who directs the behaviornor does it require the obtaining of clues from the sur-rounding environment Rather the entities synchronize in acertain rhythm In other words this order of synchronizationemergesmdashit naturally occurs out of ldquonothingrdquomdashin all cases[2]

Fireflies flocks of birds schools of fish and pacemakercells are all modeled as populations of oscillators [3] Anoscillator is an entity that continues to repeat itself in a regulartime interval If more than one oscillator influences anotheroscillator through a physical or chemical process they areconsidered to be connected Fireflies communicate throughlight birds in a flock identify their respective position usingtheir sight and pacemaker cells transmit electric currents toeach other Nature utilizes all available communication chan-nels for the connecting of oscillators This kind of communi-cation often results in synchronization

Hindawi Publishing CorporationMobile Information SystemsVolume 2017 Article ID 8932631 13 pageshttpsdoiorg10115520178932631

2 Mobile Information Systems

Models ofsynchronization in nature

Flocking

Raynoldsflocking

behavior model

Cuckerndashsmaleflockingmodel

Pacemaker cell Firefly

Kuramotofireflymodel

Consensus

Olfati-saberconsensus

model

Pulse-coupled oscillator

Mirollo and Strogatzpulse-coupled oscillator

model

Peskinpacemakercell model

Figure 1 A taxonomy of models of synchronization in nature

Synchronization in nature has been attracting consider-able research attention and has been applied to theoreticalinvestigations and a variety of mobile communication andnetworking systems [4ndash6] Some examples of its applicationsinclude time synchronization and scheduling in large-scaledistributed networks distributed fusion in sensor networksrapid consensus in various network topologies distributedformation control of multiple vehicles and solutions ofnonlinear optimization problems Due to the advancementof communication technologies in the last few years thenumber of nodes available for networking is dramaticallyincreasing Consequently applications that are intended toachieve collective goals through cooperation among nodesare increasing However the conventional approach of con-trolling numerous nodes in a centralized system has lim-itations in terms of cost complexity resource availabilityand performanceTherefore approaches inspired by the phe-nomena of natural orders such as synchronizationmdashwhichefficiently achieve certain goals in complex environmentswith various constraintsmdashare being investigated to addresscomplex engineering problems

In this article we examine synchronization phenomenain nature elucidate representative mathematical models con-cerning synchronization and investigate the principles ofsynchronization We then introduce some applications thatapply synchronization principles to communication networksystems We thereby analyze the attributes and limitations ofnature-inspired synchronizationmethods Finally we discussresearch challenges for developing the advanced applicationsbased on synchronization phenomena in the future mobilecommunication systems

The remainder of this article is organized as followsSection 2 overviews representative theoretical studies onsynchronization and the principles of synchronization areinvestigated Section 3 examines major applications basedon these principles Section 4 describes challenges involvedin applying synchronization phenomena Section 5 discussesthe similarities between synchronization in nature and cer-tain engineering problems while highlighting the use ofnature-inspired technologies Section 6 presents our conclu-sions

2 Principles of Synchronization

In this section representative theoretical synchronizationmodels are outlined to elucidate synchronization principlesand related conditions The taxonomy of synchronization

model is provided in Figure 1 which is classified by the objectsin nature that inspire to make each synchronization model

21 Reynolds Flocking Behavior Model Reynolds created thefirst behavioralmodel of animal groups such as those of birdsand fish and he demonstrated their behaviors through com-puter simulation [7] In this ldquoflockingrdquo model animal groupbehavior aligns with three rulesmdashseparation alignment andcohesionmdashas shown in Figure 2 Each animal in a groupindependently controls its own position and speed based onthese rules According to the separation rule each animaltends to maintain a certain distance from its neighboringgroup members In the alignment rule each animal tends todetermine its heading direction in accordance with the aver-age direction of its neighboring group members Accordingto the cohesion rule each animal tends to steer itself towardthe average position of its neighboring group members toavoid being separated from them An animalrsquos neighbors inthis context are determined according to the physical distancebetween the respective animal and the surrounding animalsas well as their respective visual fields The other groupmembers that are not recognized in this way are excludedfrom the calculation Accordingly each animal in the groupadjusts its position and speed based on the average positionand speed of its neighbors in accordance with the three rulesAfter a period of time the animals move at the same speed inthe same distanceThat is synchronization occurs in terms ofdistance and speed

22 Peskin Pacemaker Cell Model Peskin modeled cardiacpacemaker cells as a parallel circuit with capacitance andresistance [8] This electrical circuit is charged along agradually increasing curve and fires when the voltage reachesa certain threshold as shown in Figure 3 Thereafter thevoltage resets and a new cycle beginsThat is this behaves likean oscillator The pacemaker cell model mimics the periodicfiring of cardiac cells and decreasing of voltage to the bottomof a curve

Then Peskin modeled the cardiac cellular phenomenonas a network of electrical oscillators Among119873 oscillators thevoltage of the 119894th oscillator 119909119894 is modeled as

119889119909119894119889119905

= 1198780 minus 120574119909119894 0 le 119909119894 le 1 119894 = 1 2 119873 (1)

where 1198780 is the initial value and 120574 indicates the rate of dis-sipation The 119894th oscillator fires when 119909119894 = 1 and it regresses

Mobile Information Systems 3

(a) Separation (b) Alignment (c) Cohesion

Figure 2 Three rules of the flocking behavior model

Fire Fire Threshold

Time

Pacemakercellrsquos

voltage

Figure 3 Cardiac pacemaker cell model

to 119909119894 = 0 Oscillators influence each other through firing Ifan oscillator fires either the voltages of the other oscillatorsthat detect it increase by coupling strength 120598 or they fire whentheir voltages increase beyond the given threshold This ruleis expressed as

119909119894 (119905) = 1 997888rarr 119909119895 (119905+) = min (1 119909119895 (119905) + 120598) forall119895 = 119894 (2)

Assuming that all oscillators are identical (ie they have thesame charge curve) and are coupled by the same strengthPeskin proved that two oscillators always simultaneouslyconverge and fire even when starting activity in randominitial conditions

23 Strogatz Pulse-Coupled Oscillator Model Mirollo andStrogatz modeled the synchronization phenomenon of enti-ties into a pulse-coupled oscillator (PCO) [9] PCO has anintegrated clock It emits pulses along a certain defined cycleand receives the pulses of other oscillators through the mediaexisting between them When receiving these pulses eachoscillator adjusts its internal clock according to appropriateclock adjustment rulesThe synchronization then occurs aftera period of time

Figure 4 illustrates the PCO phase synchronization pro-cess In the PCOmodel each node acts as an oscillator with afixed time cycle 119879 The oscillator has an internal time phase119905 and increases at a certain rate from 119905 = 0 to 119905 = 119879 When119905 = 119879 the nodes fire and the phase is reset to 119905 = 0 Atthat point the surrounding nodes that detect the firing signal

f(t)Fire Fire

Threshold

120598

0 T 2Tt

Figure 4 PCO phase synchronization process

adjust their phases of 119905 to the upper point to shorten the timeuntil firing The new phase value 119905update is determined as

119905update = 119891minus1 (119891 (119905) + 120598) (3)

where firing function 119891 should be an asymptotically increas-ing concave function and the increased amount 120598 is a realnumber smaller than 1 When the value of 119905update is greaterthan119879 the node immediately fires and 119905update is reset to 0Thenodes initiate activity with different 119905 values nonetheless inaccordance with the same rules their phases of 119905 respectivelyconverge to the same value in the course of time

Strogatz generalized the coupled oscillator model ofPeskin into 119873 oscillators According to Strogatz modelunder the assumption that all oscillators are identical andconnected it is proved that119873 oscillators with random initialvalues always synchronize However it was found that PCOmodel synchronization is not realized in actuality when noiseor a pulse transmission delay exists [10]

24 Kuramoto Firefly Model Assuming that identical oscil-lators are weakly coupled and the interactions between theoscillators are dependent on the sine function of the phasedifference Kuramotomodeled the synchronization of firefliesas

119889120579119894119889119905

= 120596119894 minus119870119873

119873

sum119895=1

sin (120579119894 minus 120579119895) 119894 = 1 2 119873 (4)

where 120579119894 is the phase of each individual 120596119894 is the frequencyof each individual and 119870 is the coupling strength [11ndash13]


minus1 minus05 0 05 1minus1

minus05

0

05

1

(a) Initial state0

minus1

minus05

0

05

1

05 1minus1 minus05(b) After convergence

Figure 5 Phase synchronization phenomenon according to the Kuramoto model

In the initial state before synchronization each individualoscillates at its own frequency and its phase differs from thoseof the others However the phase of each individual changesin time through interactions by sin(120579119894 minus 120579119895) Ultimatelysynchronization occurs when the phases of all the individualsconverge as shown in Figure 5

Unlike the Strogatz model the Kuramoto model con-siders oscillators that have different natural frequencies andassumes that the interactions of the oscillators are limitedto the sine function Therefore the Kuramoto model canbe exactly solved in mathematical terms even though itis a nonlinear model This model achieves synchronizationwith a random initial value on the condition that 119870 value issufficiently large in a fully connected network [14] Howeversynchronization is not always guaranteed when the differ-ences of the natural frequencies of the oscillators transcenda specific range or when a transmission delay exists [15]

25 CuckerndashSmale Flocking Model Cucker and Smale math-ematically modeled the flocking behavior of birds in flight as

V119894 (119905 + 1) minus V119894 (119905)

= 120582119873

119873

sum119895=1

Ψ(10038161003816100381610038161003816119909119895 (119905) minus 119909119894 (119905)10038161003816100381610038161003816) (V119895 (119905) minus V119894 (119905))

(5)

where 119909119894(119905) and V119894(119905) are the position and velocity of the119894th bird at time 119905 respectively [16] 119873 is the total numberof birds 120582 is the coupling strength among the birds andΨ is the communication range function with input of thedistance between two birds For example if each bird in aflock has only information of the adjacent birds within acertain distance 119903 then Ψ = 1 when |119909119895 minus 119909119894| le 119903 and Ψ =0 otherwise As expressed in (5) the CuckerndashSmale flock-ing model shows that each bird adjusts its velocity by aver-aging the velocity of its perceived neighbors Therefore thismodel can be a mathematical representation of the Reynoldsflocking behavior model

Cucker and Smale considered various Ψ functions andproved the synchronization conditions for the general Ψfunctions in which path loss is proportional to distance [1617] They proved that the convergence value by (5) becomesthe average of the initial velocities of the birds Additionallythey showed that the position of each bird does not divergerather it remains within a certain range of space whenconverging The convergence characteristics of the CuckerndashSmale model are expressed as

V119888 fl lim119905rarrinfin

V119894 (119905) =1119873

119873

sum119894=1

V119894 (0)

for forall119894

sup0le119905ltinfin

10038161003816100381610038161003816119909119894 (119905) minus 119909119895 (119905)10038161003816100381610038161003816 lt infin for forall119894 = 119895

(6)

Figure 6 shows the simulated synchronization phenomenonof bird flocking according to this model

26 Olfati-Saber ConsensusModel In dynamic systems suchasmultiagent networks the term ldquoconsensusrdquo is used tomeanldquosynchronizationrdquo [18] Consensus connotes the reaching ofan agreement regarding a certain quantity of interest thatdepends on the state of all agents The consensus algorithmrefers to an interaction rule that specifies the informationexchange between an agent and all of its neighbors in thenetwork in order to reach a consensus

In accordance with continuous time and discrete timeintervals in a linear system a generalized consensus algo-rithm is respectively expressed as

1199091015840119894 (119905) = sum119895isin119873119894

119886119894119895 (119909119895 (119905) minus 119909119894 (119905))

119909119894 (119896 + 1) = 119909119894 (119896) + 120598 sum119895isin119873119894

119886119894119895 (119909119895 (119896) minus 119909119894 (119896)) (7)

where 119909119894(119905) is the status value of node 119894 at time 119905119873119894 is the setof neighboring nodes of node 119894 119886119894119895 is the coupling strength


20minus10100

yx 010minus1020

0

2

4

6

8

10

z

(a) Initial state

150100

yx50

0 050

minus50minus100

minus50

0

200

400

600

z

(b) After convergence

Figure 6 Synchronization phenomenon of bird flocking in three-dimensional space according to the CuckerndashSmale flocking model

Number of nodesInitial values

Network topology

DelayNoise

Path loss

Input

Convergence existence

Convergence value

Convergence rate

Coupling strength

Output

Nature-inspiredsynchronization

algorithm(local average

algorithm)

Figure 7 System model of synchronization

between the two nodes 119894 and 119895 and 120598 gt 0 is the step sizeAccording to (7) the change of the status value of each nodeis determined by the sumof the differences of the neighboringnodesrsquo state values Based on these algorithms the statusvalue of each node converges in time to the average of the ini-tial values of all the nodes It is given as

120572 = 1119873

119873

sum119894=1

119909119894 (0) (8)

where 119873 is the total number of nodes The expressions andconvergence result of this kind of consensus algorithm cor-respond to those of the CuckerndashSmale model

Various theoretical studies have been conducted on theconsensus model and their convergence rate has been ana-lyzed Equations (7) can be simply converted into the matrixforms as follows

x1015840 = minusLx

x (119896 + 1) = Px (119896) (9)

where L is a Laplacian matrix and P is a Perron matrixThesematrices are determined according to the given networktopology The convergence rate of each consensus algorithmis determined depending on the second smallest eigenvalueof Laplacian L and the second largest eigenvalue of Perron Prespectively [18]

27 Synchronization System Model In synchronization stud-ies to date the action of each individual is mathemat-ically modeled Then the existence of convergence theconvergence value and the convergence rate are theoret-ically analyzed based on the various conditions such asthe number of participating nodes initial values networktopology coupling strength between nodes delay noiseand path loss The principle of synchronization algorithmsis that each node repeats the process of updating its owninformation using only the information of the neighboringnodes This update approach can thus be represented as thelocal average algorithm by which the convergence value isdecided solely based on the average of the initial values Thissynchronization algorithm can bemodeled into a generalizedsystem with inputs and outputs as depicted in Figure 7

3 Applications of Synchronization

In this section we introduce some major applications basedon synchronization principles focusing on communicationand networking systems The taxonomy of synchronizationapplications is provided in Figure 8 which is classified by themajor applications used in communication and networkingsystems

31 Distributed Time Synchronization In an environmentin which many nodes exist and the network dynamicallychanges according to the movements entering and exiting


Applications of nature-inspired synchronization

Timesynchronization

Transmissiontimesynchronization

Frequencysynchronization

Timedesynchronization

desynchronization

Fairscheduling

Multihopdesynchronization

Data fusion

Kalmanfilterdesign

Least-squaresestimatordesign

Consensusfilterdesign

Formation control

Multivehicleformationcontrol

Rendezvouscontrol

Collisionavoidancecontrol

Network design forfast synchronization

Parameterdesign

Topologydesign

Optimization

Meta-heuristicOptimization

Particle swarmoptimization

Ant colonyoptimization

Bee colonyoptimization

Maxndashmin objectiveproblem withsolidarity property

Maximizemultihoprate

Minimizemulticelloutage

Figure 8 A taxonomy of applications of nature-inspired synchronization

Fire

Initialstate

Brsquos localview

AB

C

D

E

Fire

AB

C

FireA

B

C

Fire

Desynchronizedstate

A

B

CD

E

adjustmentBrsquos B998400

C998400

Figure 9 DESYNC algorithm

of nodesmdashsuch as in mobile ad hoc wireless mesh andsensor networksmdasha nature-inspired distributed synchroniza-tion method is more suitable than the conventional cen-tralized synchronization methods Various studies on thispremise have been conducted A synchronization methodfor node transmission time was suggested by applying thesynchronization principle of fireflies to a large-scale network[19 20] Therein the Strogatz PCO model was applied toa large-scale sensor network Synchronization performancewas then analyzed according to the number of nodes andoptimized according to the tradeoff of the energy efficiencyand convergence rate [21] A technique for synchronizingthe timing of separate networks was suggested by usingthe positions and mobility of the nodes in mobile ad hocnetworks [22] A distributed frequency synchronization tech-nique using a bioinspired algorithm was suggested to solvethe problem of multiple frequency offsets in mesh networkswherein multiple nodes have different carrier frequencies[23] Lastly a synchronization method was suggested forrealistic situations in which a time delay occurs across nodesand not all of them are connected [15 24]

32 Desynchronization Unlike synchronization by whichoscillators converge to the same time phase desynchroniza-tion (DESYNC) separates the oscillators as far as possible

from each other As a result the phase differences of all thenodes become the same In DESYNC the synchronizationtarget is not the phase value itself instead it is the phasedifference between the nodes A distributed DESYNC algo-rithm based on the PCOmodel was suggested for the desyn-chronizing of time event [25 26] As shown in Figure 9 eachoscillator detects the firing of another oscillator immediatelybefore and immediately after its own firing It adjusts its ownphase to the medium position of the firing time of its twoneighbors This process is repeated and DESYNC is com-pleted

A round-robin scheduling method that fairly allocatesresources in a distributed manner without access colli-sion was suggested by applying the DESYNC algorithm totime-division multiple access (TDMA) scheduling [25 26]DESYNC-based TDMA scheduling enables high throughputand low control overhead even in a heavy load becauseit does not induce signaling messages for scheduling Fur-thermore a distributed scheduling method was proposedby applying DESYNC to multihop networks This approachflexibly controls the transmission intervals between nodes ina situation wherein the nodes randomly enter or exit [27]Lastly proportional fair scheduling was achieved by enablingeach node to operate two oscillators through the expansionof the DESYNC algorithm [28]


33 Distributed Fusion in Sensor Networks In a sensor net-work it is necessary to reduce the amount of transmittedinformation through the fusion of sensed information inintermediate nodes rather than to concentrate all informa-tion sensed from each node to a center and simultane-ously compute it In particular as the number of nodesincreases the amount of sensed information to transmitrapidly increases Thus the fusion of sensed information ismore necessary in various parts of the network The fusionof sensed information is a process in which multiple sensornodes cooperatively make a decision This process is used toextract a useful piece of information by reaching a consensusamong the nodes

Various methods that compute the average in a dis-tributedmanner in sensor networks were suggested to realizethe Kalman filter in this context [29 30] Additionally thelinear least-squares estimator (LSE) method was proposed tocompute an average based on the consensus among networknodes [31] Lastly to more efficiently compute the average ofsensed information low and high pass consensus filters weresuggested [32]

34 Network Design for Fast Synchronization Nature-in-spired synchronization algorithms require iterations and thusrequire time to obtain the desired convergence value To over-come this constraint the development of a network designthat achieves fast synchronization is an active research topicIn [33] for example a study on shortening the convergencetime of a synchronization algorithm was presented Theconvergence timewas reduced by varying theweight which ismultiplied for each operation when the average is calculated

In another study the weight was fixed and the networktopology was varied to stimulate faster synchronization [3435] When a small-world network was created by rewiringrandom nodes in a given network topology it was confirmedthat the synchronization speed is dramatically faster Fig-ure 10 shows the difference in convergence times betweena regular network in which 100 nodes are wired withinthree hops and a small-world network wherein an additionalconnection is made between random 300 nodes on theregular network

35 Distributed Formation Control The multivehicle systemis an important networking system for commercial andmilitary applications Controlling the number of vehicles tobe automatically driven in a certain formation requires coop-eration between nodes It is thus necessary for the vehicles tosynchronize to achieve their collective mission It was proventhat distributed multivehicle formation control is a matterof synchronization and a related theoretical framework wasdeveloped [36]

In addition the matter of a space rendezvous is regardedas a synchronization problem concerning the position ofvarious individuals that interact [37 38] In the rendezvousproblem the occurrence of convergence according to thenetwork topology is important however it is irrelevant tothe convergence value Furthermore a theoretical frameworkwas developed for the design and analysis of a flocking algo-rithm that avoids collision with obstacles [39] For flocking

the relation with obstacles should be considered as well as theinteraction between individuals In this study the synchro-nization algorithm is used to enable an individual to avoidcolliding with obstacles and other individuals based only onthe information of the surroundings while synchronizing thevelocities of all individuals

36 Solution for Optimization Problems Variousmetaheuris-tic optimization methods based on swarm intelligencewere suggested some examples of which include particleswarm optimization ant colony optimization and bee colonyoptimization [40] In this type of optimization techniquenumerous particles repeatedly explore a multidimensionalsolution space by exchanging information with each other todetermine themost optimal valueThis approach is similar tothe synchronization phenomenon in the sense that variousindividuals interact in a distributed manner and convergetoward a single optimal point However it does not alignwith the basic synchronization algorithm which obtains theaverage of the local information of interest

For a certain optimization problem it was proven thatthe optimal solution is to synchronize specific variables[41ndash43] When the value sought by each individual has aldquosolidarityrdquo property (ie the increasedecrease of the valueof an individual causes the decreaseincrease of the valuesof the others) and the objective problem is to maximize theminimum of these values (ie the maxndashmin objective prob-lem) it was proven that the optimal solution is to make thesevalues equal that is to synchronize them

Based on the above proposition [41] suggested a dis-tributed transmit-power control algorithm that maximizesthe end-to-end throughput in wireless multihop networksand [43] proposed a distributed transmit-power controlalgorithm that minimizes the outage probability in wirelesscellular networksThe rate of each link in thewireless networkhas a solidarity property on account of mutual interferenceMoreover the end-to-end throughput in multihop networksis decided by the minimum value of the rates of links con-sisting of a multihop path Thus the maximization problemof multihop end-to-end throughput has the following maxndashmin objective problem

maxP

1198771198902119890 = maxP

min 11987712 11987723 119877119899(119899+1) (10)

where vector P indicates the transmit power set of 119899 trans-mitting nodes and 119877119894119895 denotes the rate of each link onthe multihop path Therefore in this problem the optimalsolution is tomake the transmission rate of each link equal Tothis end a synchronization algorithm of each link rate basedon the CuckerndashSmale flocking model [16] was proposed as

119877119894119895 (119905 + 1) minus 119877119894119895 (119905)

= 1119899

sumforall119896119897=119896+1

Ψ (1003816100381610038161003816119909119896 minus 1199091198941003816100381610038161003816) (119877119896119897 (119905) minus 119877119894119895 (119905))

(11)

To support the link rate determined by (11) each transmittingnode performs the transmit power control in each time slot

In both the maximization problem of multihop end-to-end throughput in [41] and the minimization problem of



minus05

0

05

1

(a) Topology of regular network

Time (sec)0 50 100 150

0

20

40

60

80

100

Stat

e val

ue(b) State evolution of regular network


minus05

0

05

1

(c) Topology of small-world network

0 10Time (sec)

0

20

40

60

80

100St

ate v

alue

8642

(d) State evolution of small-world network

Figure 10 Comparison of convergence times between a regular network and a small-world network

outage in [43] the synchronization phenomenon of eachlink rate and the transmission power variation in each nodeover time are depicted in Figure 11 The suggested algo-rithm basically follows the CuckerndashSmale flocking modelTherefore the convergence is guaranteed The results alsoshow that when the link transmission rates converge thenode with the best initial link rate uses the lowest transmitpower while the node with the worst initial link rate usesthe maximum transmit power Because the higher transmitpower induces more interference the optimal strategy is toenable the transmitting node with the best link to use thelowest transmit power thereby causing the least interferenceand increasing the transmission rates of the other links

The above reasoning is supported by the notion thatthe strongest bird positions itself at the head of the flockwhere the air resistance is the greatest in order to supportthe weaker birds that follow when migrating as a flock Thevelocity vectors of birds have a solidarity property owing

to the air resistance Additionally they have a maxndashminobjective because their flying speed is limited by the speedof the slowest bird Accordingly we can understand that thesynchronization of velocity in the flights of flocking birds isthe optimal solution for achieving their objective [42 43]

37 Comparisons between Nature-Inspired and ConventionalSynchronizations In this section the nature-inspired andconventional approaches are compared in terms of timesynchronization The characteristics advantages and disad-vantages of nature-inspired time synchronization methodsare evaluated compared with the conventional centralizedand distributed time synchronization methods

Centralized time synchronization methods use globaltime orientations such as global positioning system (GPS)time [44] and network time protocol (NTP) [45] All nodeslisten for the reference time provided from the central clockand adjust their own time in order to synchronize This


Node 1 Node 2 Node 3 Node 4

Node 5 Node 6 Node 7

0

05

1

15

2

25

3

35

4Li

nk ra

te (b

sH

z)

300 10 20 40Time

(a)



3010 20 400Time

25

5

10

15

20

Tran

smiss

ion

pow

er (d

Bm)

(b)

Figure 11 (a) Link rate and (b) transmission power over time for the maximization problem of end-to-end throughput in wireless multihopnetworks

process enables accurate synchronization with high reliabil-ity however it involves a high cost in terms of hardwareimplementation and energy consumptionMoreover the syn-chronization performance is related to the respective distanceto the central clock which limits the network scalability

The conventional reference broadcast synchronization(RBS) method [46] is a typical distributed time synchro-nization method that is not nature-inspired Without using acentral global clock RBS engenders mutual synchronizationamong the network nodes Thus precise synchronization ispossible without a delay from the central clock Howeverthe time information of each node should be shared withall the other nodes so that an immense amount of data canbe exchanged between the nodes As the number of nodesincreases multihop RBS is required and the data exchangeoverhead likewise increases resulting in the decrease of per-formance

Unlike these methods nature-inspired time synchro-nization methods [19ndash24] can achieve synchronization in adistributed and efficient manner regardless of the number ofnodes in a large-scale network Table 1 summarizes the char-acteristics advantages and disadvantages of the conventionalsynchronization methods and those of the nature-inspiredsynchronization methods

4 Research Challenges

In this section we examine further research challenges thatshould be considered in utilizing natural synchronizationphenomena for the future communication and networking

systems The considered research challenges are classified asin Figure 12

41 Challenges for Convergence Performance The synchro-nization algorithm conducts iterations in a dynamic environ-ment Thus the performance related to convergence is veryimportant According to the number of nodes initial valuenetwork topology and coupling strength we must theoreti-cally analyze the existence value and rate of convergence orwe must empirically recognize the tendency of convergenceTo increase the convergence rate we must design a networktopology that can stimulate rapid synchronization based onthe results of many experiments in a variety of network envi-ronments These include random small-world and large-scale networks as well as locally connected regular networks

According to studies conducted thus far a networkwherein all nodes are fully connected is known to guaranteethe convergence regardless of the number of nodes and theinitial value [9] Therefore in order to guarantee the conver-gence it is necessary to arbitrarily make the network fullyconnected Alternatively a hierarchical method can be con-sidered in which a network is clustered as a fully connectednetwork at a local level Then synchronization is engaged atthe upper level with the locally converged information

The convergence rate has a tradeoff relationship with theaccuracy of the convergence value If the convergence valueis set at a discrete level instead of a real value from a practicalperspective the convergence rate can be controlled accordingto the number of levels For example the convergence valuein the synchronization of transmission rate is mapped to


Table 1 Comparison of conventional and nature-inspired synchronization methods

Item Conventional synchronization method Nature-inspired synchronization method

Reliability High reliability and accuracy shortconvergence time

Guaranteed reliability accurate synchronization inlarge-scale networks

Scalability Limited by the numbers of nodes andhops

Scalability maintained regardless of the number of nodesand changes

Robustness Vulnerable to network topology changessuch as cluster head failures

Robust to poor communication environments and networktopology changes

Cost Very high message exchange of upperlayers is required

Low amount of data exchange can be simply operatedthrough pulse exchange at the physical layer without upperlayer protocols

Complexity Very high hardware (HW)software (SW)complexity

Low HWSW complexity (not necessary to save timeinformation of other nodes in the memory)

Influential factors forperformance

Synchronization accuracy depends on thedistance to root nodes

Affected by node density coupling strength delay noisepath loss and modulation method

Advantage

High reliability effective with anappropriate number of nodes notdependent on the global clock in the caseof RBS

High scalability and robustness low cost and complexityeffective with a large number of nodes

Disadvantage

Poor performance of outer nodes in thetopology in cases of centralizedsynchronization a large amount ofexchanged data is required in distributedsynchronization

Synchronization is impossible when the transmission andreception delays take a long time synchronization speeddecreases with weak coupling strength and a small numberof nodes

Performanceimprovement

Environmentalconstraints

Convergenceperformance

Research challenges

Convenienceanalysis

Networktopologydesign

Convergencerate control

Wirelesschannelimpairment

Communicationdelays

Deafnessproblem

Heterogeneousoscillators

Signalingoverhead

Parameteroptimization

Security Multipleoscillators

Leaderselection

Figure 12 Classification of research challenges

the modulation and coding scheme (MCS) level in realityThus the convergence rate can be controlled according tothe number of MCS levels On the contrary the numberof MCS levels can be determined according to the requiredconvergence rate

42 Challenges of Environmental Constraints Path lossnoise delay and packet loss occurring in real environ-ments influence synchronizationThe weakening of couplingstrength due to path loss and packet loss reduces the con-vergence rate while delay and noise are major factors thatinterfere with synchronization [10] When applying a syn-chronization algorithm to a real environment it is thereforenecessary to design the algorithm with consideration of allthese interfering factors

A receiving node may be unable to immediately respondon account of a communication delay Consequently the syn-chronization may not be completed Specifically the com-munication delay can be divided into a transmission delay

decoding delay and propagation delay It is necessary toredesign a synchronization algorithm considering the statis-tical ranges of these delays Moreover a wireless communi-cation node cannot receive data while simultaneously trans-mitting Thus it cannot recognize the firing of other nodeswhile it is firing Therefore the firing period should be set tocross one another to avoid such ldquodeafnessrdquo problems in wire-less communication environments

To date research has been conducted only on homoge-nous oscillators that follow the same rule However inreality heterogeneous oscillators may exist that adhere todifferent synchronization rules in the same situation onaccount of the network heterogeneity Therefore researchshould be conducted on synchronization algorithms whenthe heterogeneous oscillators are intended to cooperativelyachieve a certain collective purpose

For the real operation of the synchronization algorithmsdescribed in Section 2 periodic data exchange among neigh-boring nodes is required Consequently signaling overhead


occurs Considering this kind of overhead cost the pointat which it is more effective to apply nature-inspired syn-chronization methods over conventional synchronizationmethods should be determined Depending on the situationa hybrid synchronization algorithm can be considered thatadaptively applies both types of methods

43 Challenges to Performance Improvement Depending onnetwork topology changes in the wireless environment syn-chronization performance can be improved by optimizing theoperation parameters of the synchronization algorithm suchas the coupling strength weight and step size The mappingrelations between operation parameters and environmentvariables should be configured through various experiments

In terms of network security a method is required toenhance the tolerance of systems againstmalicious nodes thatintentionally interferewith the synchronization For examplea synchronization algorithm can be designed to ignore cer-tain information with the maximum deviation from the aver-age of the information received from neighboring nodes Asynchronization algorithm that considers unnecessary infor-mation should be devised and it should be verified whetherit is possible to achieve a desired convergence perform-ance

When there are two or more oscillators in a single node amethod that can utilize multiple oscillators should be consid-ered In [28] proportional fair scheduling was conducted in adistributed manner using two oscillatorsTherefore complexsynchronization phenomena with several oscillators need tobe studied for developing a new application that utilizes them

It is possible to converge the values of all nodes into adesired value by selecting a leader and setting only its valueFor example in aircraft engineering by controlling the singleleader in flock-like flight formation this method can enableall the rest of the aircrafts to automatically follow that specificaircraft In addition the transmission rate of all nodes can beequalized by controlling only the rate of the leader Howeverthis kind of approach that controls the convergence value onlythrough a leader may be unable to guarantee the convergencedepending on the environment Thus a theoretical study toaddress that issue should be conducted

5 Discussion and Prospect

The reason why we focus on synchronization phenomenacan be explained as follows First of all the natural envi-ronment in which synchronization occurs is similar to theenvironments in which many engineering problems occurSynchronization occurs among a number of individuals Italso occurs in a completely distributed system in whicheach individual decides its own action with consideration ofonly the surrounding environment or situation In additionindividuals perfectly synchronize in dynamic environmentswhere obstacles or adversariesmay appear andnodes enter orexit These kinds of environments in which synchronizationoccurs inspire many engineering problems in large-scalenetworking distributed and dynamic environments

Secondly the purpose of synchronization in nature canbe understood as the same as engineering objectives Insects

birds fish and other animals that form groups are known tosynchronize for migration mating protection from preda-tors and efficient searching for food [3] These creaturesexhibit solidarity and thus engage in synchronization for col-lective goals These goals are typically for survival purposesSuch animal populations use synchronization as an optimalstrategy for surviving while evolving through time Likewisea number of nodes in systems such as sensor networksmultivehicle systems and multihop network must achievea collective objective with solidarity Their objectives are torapidly obtain useful information quickly move togetherover long distances and maximize the end-to-end through-put respectively These goals are not for maximizing the per-formance of individual nodes rather they serve to maximizethe performance of the entire systemTherefore the purposesof engineering systems correspond to those demonstratedby collectives of creatures It is thus reasonable to employsynchronization as the optimal solution to problems that existin engineering systems

Lastly synchronization in nature is more efficient thanany centralized or distributed system available For examplethousands of fireflies engender the synchronization phe-nomenon in the fastest and most efficient way with limitedresources They use a very simple method and achievesynchronization within a short period of time Thereforemimicking these natural approaches can enable engineeringsystems to achieve performances at a similar level of lowcomplexity and cost in situations with limited resources

The point at which nature-inspired technologies willbecomemore effective than existing approaches will be whenthe number of nodes comprising a system increases beyond acertain level Owing to the advancement of communicationnetwork technologies the number of nodes connected tonetworks is drastically increasing each year The increaseof nodes accelerates the shortage of resources and incursthe high cost of centralized management Conventionalcentralized technologies developed in consideration of alimited number of nodes can neither operate with a largenumber of nodes nor guarantee the required performanceThus naturally a shift to distributed systems is requiredHowever distributed system technologies developed underthe centralized paradigm still bear the limitation concerningthe increase in the number of nodes Furthermore they arevulnerable to environmental changes due to the incomplete-ness of the distribution [46] Therefore a paradigm shift tonew distributed systems is necessary and it is imperative tofocus on the distributed algorithms of collective organismsin nature as an effective methodology to be applied

6 Conclusion

In this article we identified the principles of various naturalsynchronization phenomena and studied how they can beapplied in practical communication and networking systemsIn terms of time synchronization methods in networksthe strengths and weaknesses of conventional approachesand nature-inspired approaches were compared Furtherresearch challenges that are necessary for realizing nature-inspired synchronization technologies were presented The


rapid increase of the number of networking nodes willaccelerate the occurrence of complex and distributed networkenvironments which are getting similar to the environmentof nature Therefore it would be a great help for us to mimicand apply the fundamental principles of biocommunities inwhich each entity has shown collective behaviors to sustaintheir lives in the midst of complex and chaotic environmentsIn the future various studies considering practical issuesare required to effectively apply these principles to mobilecommunication and networking systems

Competing Interests

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

Acknowledgments

This research was supported by Basic Science ResearchProgram through theNational Research Foundation of Korea(NRF) funded by the Ministry of Science ICT amp FuturePlanning (NRF-2016R1C1B1016261) by the Basic ScienceResearch Program through the National Research Founda-tion of Korea (NRF) funded by the Ministry of EducationScience and Technology (NRF-2015R1D1A1A01060207) andby the Human Resources Development of the Korea Instituteof Energy Technology Evaluation and Planning (KETEP)grant funded by the Korea government Ministry of TradeIndustry and Energy (no 20154030200860)

References

[1] A Pikovsky M Rosenblum and J Kurths Synchronization AUniversal Concept in Nonlinear Sciences Cambridge UniversityPress 2001

[2] H-H Choi and J-R Lee ldquoCommunication and networkingtechnologies based on bio-inspired algorithmsrdquo KICS Informa-tion amp CommunicationsMagazine vol 29 no 4 pp 62ndash71 2012(Korean)

[3] S H Strogatz Sync How Order Emerges from Chaos in theUniverse Nature andDaily Life Hyperion NewYork NYUSA2004

[4] F Dressler and O B Akan ldquoA survey on bio-inspired network-ingrdquo Computer Networks vol 54 no 6 pp 881ndash900 2010

[5] Z Zhang K Long J Wang and F Dressler ldquoOn swarm intel-ligence inspired self-organized networking its bionic mecha-nisms designing principles and optimization approachesrdquo IEEECommunications Surveys amp Tutorials vol 16 no 1 pp 513ndash5372014

[6] C Zheng and D C Sicker ldquoA survey on biologically inspiredalgorithms for computer networkingrdquo IEEE CommunicationsSurveys amp Tutorials vol 15 no 3 pp 1160ndash1191 2013

[7] C W Reynolds ldquoFlocks herds and schools a distributedbehavioral modelrdquo ACM Computer Graphics vol 21 no 4 pp25ndash34 1987

[8] C S Peskin ldquoMathematical aspects of heart physiologyrdquo TechRep Courant Institute of Mathematical Sciences New YorkUniversity 1975

[9] R E Mirollo and S H Strogatz ldquoSynchronization of pulse-coupled biological oscillatorsrdquo SIAM Journal on Applied Math-ematics vol 50 no 6 pp 1645ndash1662 1990

[10] U Ernst K Pawelzik and T Geisel ldquoSynchronization inducedby temporal delays in pulse-coupled oscillatorsrdquo PhysicalReview Letters vol 74 no 9 pp 1570ndash1573 1995

[11] Y Kuramoto and H Araki Eds Lecture Notes in Physics Inter-national Symposium on Mathematical Problems in TheoreticalPhysics Springer New York NY USA 1975

[12] S H Strogatz ldquoFrom Kuramoto to Crawford exploring theonset of synchronization in populations of coupled oscillatorsrdquoPhysica D Nonlinear Phenomena vol 143 no 1-4 pp 1ndash202000

[13] J A Acebron L L Bonilla C J Perez-Vicente F Ritort and RSpigler ldquoThe Kuramoto model a simple paradigm for synchro-nization phenomenardquo Reviews of Modern Physics vol 77 no 1pp 137ndash185 2005

[14] R Sepulchre D Paley and N Leonard ldquoCollective motion andoscillator synchronizationrdquo in Proceedings of the Block IslandWorkshop Cooperative Control Block Island RI USA June2003

[15] A Papachristodoulou and A Jadbabaie ldquoSynchronization inoscillator networks switching topologies and non-homoge-neous delaysrdquo in Proceedings of the 44th IEEE Conference onDecision and Control and the European Control Conference(CDC-ECC rsquo05) pp 5692ndash5697 December 2005

[16] F Cucker and S Smale ldquoEmergent behavior in flocksrdquo IEEETransactions on Automatic Control vol 52 no 5 pp 852ndash8622007

[17] S-Y Ha and J-G Liu ldquoA simple proof of the Cucker-Smaleflocking dynamics and mean-field limitrdquo Communications inMathematical Sciences vol 7 no 2 pp 297ndash325 2009

[18] R Olfati-Saber J A Fax and R M Murray ldquoConsensus andcooperation in networked multi-agent systemsrdquo Proceedings ofthe IEEE vol 95 no 1 pp 215ndash233 2007

[19] A Tyrrell G Auer and C Bettstetter ldquoFireflies as role modelsfor synchronization in ad hoc networksrdquo in Proceedings of the1st Bio-Inspired Models of Network Information and ComputingSystems pp 1ndash7 Madonna di Campiglio Italy December 2006

[20] A Tyrrell andG Auer ldquoImposing a reference timing onto fireflysynchronization in wireless networksrdquo in Proceedings of theIEEE 65th Vehicular Technology Conference (VTC rsquo07)-Springpp 222ndash226 April 2007

[21] Y-W Hong and A Scaglione ldquoA scalable synchronizationprotocol for large scale sensor networks and its applicationsrdquoIEEE Journal on Selected Areas in Communications vol 23 no5 pp 1085ndash1099 2005

[22] GA Puerta E A Aguirre andMAAlzate ldquoEffect of topologyand mobility in bio-inspired synchronization of mobile ad hocnetworksrdquo in Proceedings of the IEEE Latin-American Confer-ence on Communications (LATINCOM rsquo10) pp 1ndash6 September2010

[23] H Yoo M Lee and Y Cho ldquoA distributed frequency synchro-nization technique for OFDMA-based mesh networks usingbio-inspired algorithmrdquo The Journal of Korea Information andCommunications Society vol 37B no 11 pp 1022ndash1032 2012

[24] G Werner-Allen G Tewari A Patel M Welsh and R NagpalldquoFirefly-inspired sensor network synchronicity with realisticradio effectsrdquo in Proceedings of the 3rd ACM InternationalConference on Embedded Networked Sensor Systems (SenSysrsquo05) pp 142ndash153 ACM SanDiego Calif USA November 2005

[25] J Degesys I Rose A Patel and R Nagpal ldquoDESYNC self-organizing desynchronization and TDMA on wireless sensornetworksrdquo in Proceedings of the 6th International Symposium on


Information Processing in Sensor Networks (IPSN rsquo07) pp 11ndash20April 2007

[26] A Patel J Degesys and R Nagpal ldquoDesynchronization thetheory of self-organizing algorithms for round-robin schedul-ingrdquo in Proceedings of the 1st International Conference on Self-Adaptive and Self-Organizing Systems (SASO rsquo07) pp 87ndash96Boston Mass USA July 2007

[27] J Degesys andRNagpal ldquoTowards desynchronization ofmulti-hop topologiesrdquo in Proceedings of the 2nd IEEE InternationalConference on Self-Adaptive and Self-Organizing Systems (SASOrsquo08) pp 129ndash138 Venice Italy October 2008

[28] R Pagliari Y-W P Hong and A Scaglione ldquoBio-inspiredalgorithms for decentralized round-robin and proportional fairschedulingrdquo IEEE Journal on Selected Areas in Communicationsvol 28 no 4 pp 564ndash575 2010

[29] R Olfati-Saber and J S Shamma ldquoConsensus filters for sensornetworks and distributed sensor fusionrdquo in Proceedings ofthe 44th IEEE Conference on Decision and Control and theEuropean Control Conference (CDC-ECC rsquo05) pp 6698ndash6703December 2005

[30] D P Spanos R Olfati-Saber and R M Murray ldquoApproximatedistributed kalman filtering in sensor networks with quantifi-able performancerdquo in Proceedings of the 4th International Sym-posium on Information Processing in SensorNetworks (IPSN rsquo05)pp 133ndash139 Los Angeles Calif USA April 2005

[31] L Xiao S Boyd and S Lall ldquoA Scheme for robust distributedsensor fusion based on average consensusrdquo in Proceedings of the4th International Symposium on Information Processing in Sen-sor Networks (IPSN rsquo05) pp 63ndash70 April 2005

[32] D Spanos R Olfati-Saber and R M Murray ldquoDynamic con-sensus on mobile networksrdquo in Proceedings of the 16th Interna-tional Federation of Automatic Control World Congress (IFACrsquo05) Prague Czech Republic 2005

[33] L Xiao and S Boyd ldquoFast linear iterations for distributedaveragingrdquo Systems and Control Letters vol 53 no 1 pp 65ndash782004

[34] D J Watts and S H Strogatz ldquoCollective dynamics of rsquosmall-worldrsquo networksrdquoNature vol 393 no 6684 pp 440ndash442 1998

[35] R Olfati-Saber ldquoUltrafast consensus in small-world networksrdquoin Proceedings of the American Control Conference (ACC rsquo05)pp 2371ndash2378 June 2005

[36] J A Fax and R M Murray ldquoInformation flow and cooperativecontrol of vehicle formationsrdquo Institute of Electrical and Elec-tronics Engineers Transactions onAutomatic Control vol 49 no9 pp 1465ndash1476 2004

[37] J Lin A S Morse and B D O Anderson ldquoThe multi-agentrendezvous problemrdquo in Proceedings of the 42nd IEEE Confer-ence onDecision andControl pp 1508ndash1513MauiHawaii USADecember 2003

[38] J Cortes S Martinez and F Bullo ldquoRobust rendezvous formobile autonomous agents via proximity graphs in arbitrarydimensionsrdquo Institute of Electrical and Electronics EngineersTransactions on Automatic Control vol 51 no 8 pp 1289ndash12982006

[39] R Olfati-Saber ldquoFlocking for multi-agent dynamic systemsalgorithms and theoryrdquo Institute of Electrical and ElectronicsEngineers Transactions on Automatic Control vol 51 no 3 pp401ndash420 2006

[40] X-S Yang Z Cui R Xiao A H Gandomi and M Kara-manoglu Swarm Intelligence and Bio-Inspired ComputationTheory and Applications Elsevier 1st edition 2013

[41] H-H Choi and J-R Lee ldquoDistributed transmit power controlfor maximizing end-to-end throughput in wireless multi-hopnetworksrdquoWireless Personal Communications vol 74 no 3 pp1033ndash1044 2014

[42] H-H Choi and J-R Lee ldquoA bio-inspired transmit powercontrol algorithm for linear multi-hop wireless networksrdquo inProceedings of the IARIA International Conference on Networks(ICN rsquo14) February 2014

[43] H Choi and J Lee ldquoA biologically-inspired power control algo-rithm for energy-efficient cellular networksrdquoEnergies vol 9 no3 article no 161 2016

[44] J Mannermaa K Kalliomaki T Mansten and S TurunenldquoTiming performance of various GPS receiversrdquo in Proceedingsof the IEEE International Frequency Control Symposium 1999Proceedings of the 1999 Joint Meeting of the European pp 287ndash290 April 1999

[45] D L Mills ldquoInternet time synchronization the network timeprotocolrdquo IEEE Transactions on Communications vol 39 no 10pp 1482ndash1493 1991

[46] J Elson L Girod and D Estrin ldquoFine-grained network timesynchronization using reference broadcastsrdquo SIGOPSOperatingSystems Review vol 36 pp 147ndash163 2002

Submit your manuscripts athttpswwwhindawicom

Computer Games Technology

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014


Distributed Sensor Networks


Advances in

FuzzySystems

Hindawi Publishing Corporationhttpwwwhindawicom

Volume 2014


ReconfigurableComputing

Hindawi Publishing Corporation httpwwwhindawicom Volume 2014


Applied Computational Intelligence and Soft Computing

thinspAdvancesthinspinthinsp

Artificial Intelligence

HindawithinspPublishingthinspCorporationhttpwwwhindawicom Volumethinsp2014

Advances inSoftware EngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014


Electrical and Computer Engineering

Journal of

Journal of

Computer Networks and Communications


Hindawi Publishing Corporation

httpwwwhindawicom Volume 2014

Advances in

Multimedia


Biomedical Imaging


ArtificialNeural Systems

Advances in


RoboticsJournal of



Computational Intelligence and Neuroscience

Industrial EngineeringJournal of


Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014


Human-ComputerInteraction

Advances in

Computer EngineeringAdvances in



Models ofsynchronization in nature

Flocking

Raynoldsflocking

behavior model

Cuckerndashsmaleflockingmodel

Pacemaker cell Firefly

Kuramotofireflymodel

Consensus

Olfati-saberconsensus

model

Pulse-coupled oscillator

Mirollo and Strogatzpulse-coupled oscillator

model

Peskinpacemakercell model

Figure 1 A taxonomy of models of synchronization in nature

Synchronization in nature has been attracting consider-able research attention and has been applied to theoreticalinvestigations and a variety of mobile communication andnetworking systems [4ndash6] Some examples of its applicationsinclude time synchronization and scheduling in large-scaledistributed networks distributed fusion in sensor networksrapid consensus in various network topologies distributedformation control of multiple vehicles and solutions ofnonlinear optimization problems Due to the advancementof communication technologies in the last few years thenumber of nodes available for networking is dramaticallyincreasing Consequently applications that are intended toachieve collective goals through cooperation among nodesare increasing However the conventional approach of con-trolling numerous nodes in a centralized system has lim-itations in terms of cost complexity resource availabilityand performanceTherefore approaches inspired by the phe-nomena of natural orders such as synchronizationmdashwhichefficiently achieve certain goals in complex environmentswith various constraintsmdashare being investigated to addresscomplex engineering problems

In this article we examine synchronization phenomenain nature elucidate representative mathematical models con-cerning synchronization and investigate the principles ofsynchronization We then introduce some applications thatapply synchronization principles to communication networksystems We thereby analyze the attributes and limitations ofnature-inspired synchronizationmethods Finally we discussresearch challenges for developing the advanced applicationsbased on synchronization phenomena in the future mobilecommunication systems

The remainder of this article is organized as followsSection 2 overviews representative theoretical studies onsynchronization and the principles of synchronization areinvestigated Section 3 examines major applications basedon these principles Section 4 describes challenges involvedin applying synchronization phenomena Section 5 discussesthe similarities between synchronization in nature and cer-tain engineering problems while highlighting the use ofnature-inspired technologies Section 6 presents our conclu-sions

2 Principles of Synchronization

In this section representative theoretical synchronizationmodels are outlined to elucidate synchronization principlesand related conditions The taxonomy of synchronization

model is provided in Figure 1 which is classified by the objectsin nature that inspire to make each synchronization model

21 Reynolds Flocking Behavior Model Reynolds created thefirst behavioralmodel of animal groups such as those of birdsand fish and he demonstrated their behaviors through com-puter simulation [7] In this ldquoflockingrdquo model animal groupbehavior aligns with three rulesmdashseparation alignment andcohesionmdashas shown in Figure 2 Each animal in a groupindependently controls its own position and speed based onthese rules According to the separation rule each animaltends to maintain a certain distance from its neighboringgroup members In the alignment rule each animal tends todetermine its heading direction in accordance with the aver-age direction of its neighboring group members Accordingto the cohesion rule each animal tends to steer itself towardthe average position of its neighboring group members toavoid being separated from them An animalrsquos neighbors inthis context are determined according to the physical distancebetween the respective animal and the surrounding animalsas well as their respective visual fields The other groupmembers that are not recognized in this way are excludedfrom the calculation Accordingly each animal in the groupadjusts its position and speed based on the average positionand speed of its neighbors in accordance with the three rulesAfter a period of time the animals move at the same speed inthe same distanceThat is synchronization occurs in terms ofdistance and speed

22 Peskin Pacemaker Cell Model Peskin modeled cardiacpacemaker cells as a parallel circuit with capacitance andresistance [8] This electrical circuit is charged along agradually increasing curve and fires when the voltage reachesa certain threshold as shown in Figure 3 Thereafter thevoltage resets and a new cycle beginsThat is this behaves likean oscillator The pacemaker cell model mimics the periodicfiring of cardiac cells and decreasing of voltage to the bottomof a curve

Then Peskin modeled the cardiac cellular phenomenonas a network of electrical oscillators Among119873 oscillators thevoltage of the 119894th oscillator 119909119894 is modeled as

119889119909119894119889119905

= 1198780 minus 120574119909119894 0 le 119909119894 le 1 119894 = 1 2 119873 (1)

where 1198780 is the initial value and 120574 indicates the rate of dis-sipation The 119894th oscillator fires when 119909119894 = 1 and it regresses




Fire Fire Threshold

Time

Pacemakercellrsquos

voltage



119909119894 (119905) = 1 997888rarr 119909119895 (119905+) = min (1 119909119895 (119905) + 120598) forall119895 = 119894 (2)




f(t)Fire Fire

Threshold

120598

0 T 2Tt



119905update = 119891minus1 (119891 (119905) + 120598) (3)




119889120579119894119889119905

= 120596119894 minus119870119873

119873

sum119895=1

sin (120579119894 minus 120579119895) 119894 = 1 2 119873 (4)




minus05

0

05

1

(a) Initial state0

minus1

minus05

0

05

1






V119894 (119905 + 1) minus V119894 (119905)

= 120582119873

119873

sum119895=1

Ψ(10038161003816100381610038161003816119909119895 (119905) minus 119909119894 (119905)10038161003816100381610038161003816) (V119895 (119905) minus V119894 (119905))

(5)




V119894 (119905) =1119873

119873

sum119894=1

V119894 (0)

for forall119894

sup0le119905ltinfin


(6)




1199091015840119894 (119905) = sum119895isin119873119894

119886119894119895 (119909119895 (119905) minus 119909119894 (119905))

119909119894 (119896 + 1) = 119909119894 (119896) + 120598 sum119895isin119873119894

119886119894119895 (119909119895 (119896) minus 119909119894 (119896)) (7)



20minus10100

yx 010minus1020

0

2

4

6

8

10

z

(a) Initial state

150100

yx50

0 050

minus50minus100

minus50

0

200

400

600

z




Network topology

DelayNoise

Path loss

Input


Convergence value

Convergence rate

Coupling strength

Output



algorithm)



120572 = 1119873

119873

sum119894=1

119909119894 (0) (8)



x1015840 = minusLx

x (119896 + 1) = Px (119896) (9)








Timesynchronization




desynchronization

Fairscheduling


Data fusion

Kalmanfilterdesign



Formation control


Rendezvouscontrol



Parameterdesign

Topologydesign

Optimization









Fire

Initialstate

Brsquos localview

AB

C

D

E

Fire

AB

C

FireA

B

C

Fire

Desynchronizedstate

A

B

CD

E


C998400

















maxP

1198771198902119890 = maxP

min 11987712 11987723 119877119899(119899+1) (10)


119877119894119895 (119905 + 1) minus 119877119894119895 (119905)

= 1119899

sumforall119896119897=119896+1

Ψ (1003816100381610038161003816119909119896 minus 1199091198941003816100381610038161003816) (119877119896119897 (119905) minus 119877119894119895 (119905))

(11)





minus05

0

05

1


Time (sec)0 50 100 150

0

20

40

60

80

100

Stat

e val



minus05

0

05

1


0 10Time (sec)

0

20

40

60

80

100St

ate v

alue

8642











0

05

1

15

2

25

3

35

4Li

nk ra

te (b

sH

z)

300 10 20 40Time

(a)



3010 20 400Time

25

5

10

15

20

Tran

smiss

ion

pow

er (d

Bm)

(b)



























Advantage



Disadvantage






Research challenges

Convenienceanalysis




Communicationdelays

Deafnessproblem


Signalingoverhead



Leaderselection




















6 Conclusion




Competing Interests


Acknowledgments


References
























































Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in






Fire Fire Threshold

Time

Pacemakercellrsquos

voltage



119909119894 (119905) = 1 997888rarr 119909119895 (119905+) = min (1 119909119895 (119905) + 120598) forall119895 = 119894 (2)




f(t)Fire Fire

Threshold

120598

0 T 2Tt



119905update = 119891minus1 (119891 (119905) + 120598) (3)




119889120579119894119889119905

= 120596119894 minus119870119873

119873

sum119895=1

sin (120579119894 minus 120579119895) 119894 = 1 2 119873 (4)




minus05

0

05

1

(a) Initial state0

minus1

minus05

0

05

1






V119894 (119905 + 1) minus V119894 (119905)

= 120582119873

119873

sum119895=1

Ψ(10038161003816100381610038161003816119909119895 (119905) minus 119909119894 (119905)10038161003816100381610038161003816) (V119895 (119905) minus V119894 (119905))

(5)




V119894 (119905) =1119873

119873

sum119894=1

V119894 (0)

for forall119894

sup0le119905ltinfin


(6)




1199091015840119894 (119905) = sum119895isin119873119894

119886119894119895 (119909119895 (119905) minus 119909119894 (119905))

119909119894 (119896 + 1) = 119909119894 (119896) + 120598 sum119895isin119873119894

119886119894119895 (119909119895 (119896) minus 119909119894 (119896)) (7)



20minus10100

yx 010minus1020

0

2

4

6

8

10

z

(a) Initial state

150100

yx50

0 050

minus50minus100

minus50

0

200

400

600

z




Network topology

DelayNoise

Path loss

Input


Convergence value

Convergence rate

Coupling strength

Output



algorithm)



120572 = 1119873

119873

sum119894=1

119909119894 (0) (8)



x1015840 = minusLx

x (119896 + 1) = Px (119896) (9)








Timesynchronization




desynchronization

Fairscheduling


Data fusion

Kalmanfilterdesign



Formation control


Rendezvouscontrol



Parameterdesign

Topologydesign

Optimization









Fire

Initialstate

Brsquos localview

AB

C

D

E

Fire

AB

C

FireA

B

C

Fire

Desynchronizedstate

A

B

CD

E


C998400

















maxP

1198771198902119890 = maxP

min 11987712 11987723 119877119899(119899+1) (10)


119877119894119895 (119905 + 1) minus 119877119894119895 (119905)

= 1119899

sumforall119896119897=119896+1

Ψ (1003816100381610038161003816119909119896 minus 1199091198941003816100381610038161003816) (119877119896119897 (119905) minus 119877119894119895 (119905))

(11)





minus05

0

05

1


Time (sec)0 50 100 150

0

20

40

60

80

100

Stat

e val



minus05

0

05

1


0 10Time (sec)

0

20

40

60

80

100St

ate v

alue

8642











0

05

1

15

2

25

3

35

4Li

nk ra

te (b

sH

z)

300 10 20 40Time

(a)



3010 20 400Time

25

5

10

15

20

Tran

smiss

ion

pow

er (d

Bm)

(b)



























Advantage



Disadvantage






Research challenges

Convenienceanalysis




Communicationdelays

Deafnessproblem


Signalingoverhead



Leaderselection




















6 Conclusion




Competing Interests


Acknowledgments


References
























































Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in





minus05

0

05

1

(a) Initial state0

minus1

minus05

0

05

1






V119894 (119905 + 1) minus V119894 (119905)

= 120582119873

119873

sum119895=1

Ψ(10038161003816100381610038161003816119909119895 (119905) minus 119909119894 (119905)10038161003816100381610038161003816) (V119895 (119905) minus V119894 (119905))

(5)




V119894 (119905) =1119873

119873

sum119894=1

V119894 (0)

for forall119894

sup0le119905ltinfin


(6)




1199091015840119894 (119905) = sum119895isin119873119894

119886119894119895 (119909119895 (119905) minus 119909119894 (119905))

119909119894 (119896 + 1) = 119909119894 (119896) + 120598 sum119895isin119873119894

119886119894119895 (119909119895 (119896) minus 119909119894 (119896)) (7)



20minus10100

yx 010minus1020

0

2

4

6

8

10

z

(a) Initial state

150100

yx50

0 050

minus50minus100

minus50

0

200

400

600

z




Network topology

DelayNoise

Path loss

Input


Convergence value

Convergence rate

Coupling strength

Output



algorithm)



120572 = 1119873

119873

sum119894=1

119909119894 (0) (8)



x1015840 = minusLx

x (119896 + 1) = Px (119896) (9)








Timesynchronization




desynchronization

Fairscheduling


Data fusion

Kalmanfilterdesign



Formation control


Rendezvouscontrol



Parameterdesign

Topologydesign

Optimization









Fire

Initialstate

Brsquos localview

AB

C

D

E

Fire

AB

C

FireA

B

C

Fire

Desynchronizedstate

A

B

CD

E


C998400

















maxP

1198771198902119890 = maxP

min 11987712 11987723 119877119899(119899+1) (10)


119877119894119895 (119905 + 1) minus 119877119894119895 (119905)

= 1119899

sumforall119896119897=119896+1

Ψ (1003816100381610038161003816119909119896 minus 1199091198941003816100381610038161003816) (119877119896119897 (119905) minus 119877119894119895 (119905))

(11)





minus05

0

05

1


Time (sec)0 50 100 150

0

20

40

60

80

100

Stat

e val



minus05

0

05

1


0 10Time (sec)

0

20

40

60

80

100St

ate v

alue

8642











0

05

1

15

2

25

3

35

4Li

nk ra

te (b

sH

z)

300 10 20 40Time

(a)



3010 20 400Time

25

5

10

15

20

Tran

smiss

ion

pow

er (d

Bm)

(b)



























Advantage



Disadvantage






Research challenges

Convenienceanalysis




Communicationdelays

Deafnessproblem


Signalingoverhead



Leaderselection




















6 Conclusion




Competing Interests


Acknowledgments


References
























































Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in




20minus10100

yx 010minus1020

0

2

4

6

8

10

z

(a) Initial state

150100

yx50

0 050

minus50minus100

minus50

0

200

400

600

z




Network topology

DelayNoise

Path loss

Input


Convergence value

Convergence rate

Coupling strength

Output



algorithm)



120572 = 1119873

119873

sum119894=1

119909119894 (0) (8)



x1015840 = minusLx

x (119896 + 1) = Px (119896) (9)








Timesynchronization




desynchronization

Fairscheduling


Data fusion

Kalmanfilterdesign



Formation control


Rendezvouscontrol



Parameterdesign

Topologydesign

Optimization









Fire

Initialstate

Brsquos localview

AB

C

D

E

Fire

AB

C

FireA

B

C

Fire

Desynchronizedstate

A

B

CD

E


C998400

















maxP

1198771198902119890 = maxP

min 11987712 11987723 119877119899(119899+1) (10)


119877119894119895 (119905 + 1) minus 119877119894119895 (119905)

= 1119899

sumforall119896119897=119896+1

Ψ (1003816100381610038161003816119909119896 minus 1199091198941003816100381610038161003816) (119877119896119897 (119905) minus 119877119894119895 (119905))

(11)





minus05

0

05

1


Time (sec)0 50 100 150

0

20

40

60

80

100

Stat

e val



minus05

0

05

1


0 10Time (sec)

0

20

40

60

80

100St

ate v

alue

8642











0

05

1

15

2

25

3

35

4Li

nk ra

te (b

sH

z)

300 10 20 40Time

(a)



3010 20 400Time

25

5

10

15

20

Tran

smiss

ion

pow

er (d

Bm)

(b)



























Advantage



Disadvantage






Research challenges

Convenienceanalysis




Communicationdelays

Deafnessproblem


Signalingoverhead



Leaderselection




















6 Conclusion




Competing Interests


Acknowledgments


References
























































Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in





Timesynchronization




desynchronization

Fairscheduling


Data fusion

Kalmanfilterdesign



Formation control


Rendezvouscontrol



Parameterdesign

Topologydesign

Optimization









Fire

Initialstate

Brsquos localview

AB

C

D

E

Fire

AB

C

FireA

B

C

Fire

Desynchronizedstate

A

B

CD

E


C998400

















maxP

1198771198902119890 = maxP

min 11987712 11987723 119877119899(119899+1) (10)


119877119894119895 (119905 + 1) minus 119877119894119895 (119905)

= 1119899

sumforall119896119897=119896+1

Ψ (1003816100381610038161003816119909119896 minus 1199091198941003816100381610038161003816) (119877119896119897 (119905) minus 119877119894119895 (119905))

(11)





minus05

0

05

1


Time (sec)0 50 100 150

0

20

40

60

80

100

Stat

e val



minus05

0

05

1


0 10Time (sec)

0

20

40

60

80

100St

ate v

alue

8642











0

05

1

15

2

25

3

35

4Li

nk ra

te (b

sH

z)

300 10 20 40Time

(a)



3010 20 400Time

25

5

10

15

20

Tran

smiss

ion

pow

er (d

Bm)

(b)



























Advantage



Disadvantage






Research challenges

Convenienceanalysis




Communicationdelays

Deafnessproblem


Signalingoverhead



Leaderselection




















6 Conclusion




Competing Interests


Acknowledgments


References
























































Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in














maxP

1198771198902119890 = maxP

min 11987712 11987723 119877119899(119899+1) (10)


119877119894119895 (119905 + 1) minus 119877119894119895 (119905)

= 1119899

sumforall119896119897=119896+1

Ψ (1003816100381610038161003816119909119896 minus 1199091198941003816100381610038161003816) (119877119896119897 (119905) minus 119877119894119895 (119905))

(11)





minus05

0

05

1


Time (sec)0 50 100 150

0

20

40

60

80

100

Stat

e val



minus05

0

05

1


0 10Time (sec)

0

20

40

60

80

100St

ate v

alue

8642











0

05

1

15

2

25

3

35

4Li

nk ra

te (b

sH

z)

300 10 20 40Time

(a)



3010 20 400Time

25

5

10

15

20

Tran

smiss

ion

pow

er (d

Bm)

(b)



























Advantage



Disadvantage






Research challenges

Convenienceanalysis




Communicationdelays

Deafnessproblem


Signalingoverhead



Leaderselection




















6 Conclusion




Competing Interests


Acknowledgments


References
























































Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in





minus05

0

05

1


Time (sec)0 50 100 150

0

20

40

60

80

100

Stat

e val



minus05

0

05

1


0 10Time (sec)

0

20

40

60

80

100St

ate v

alue

8642











0

05

1

15

2

25

3

35

4Li

nk ra

te (b

sH

z)

300 10 20 40Time

(a)



3010 20 400Time

25

5

10

15

20

Tran

smiss

ion

pow

er (d

Bm)

(b)



























Advantage



Disadvantage






Research challenges

Convenienceanalysis




Communicationdelays

Deafnessproblem


Signalingoverhead



Leaderselection




















6 Conclusion




Competing Interests


Acknowledgments


References
























































Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in






0

05

1

15

2

25

3

35

4Li

nk ra

te (b

sH

z)

300 10 20 40Time

(a)



3010 20 400Time

25

5

10

15

20

Tran

smiss

ion

pow

er (d

Bm)

(b)



























Advantage



Disadvantage






Research challenges

Convenienceanalysis




Communicationdelays

Deafnessproblem


Signalingoverhead



Leaderselection




















6 Conclusion




Competing Interests


Acknowledgments


References
























































Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in



















Advantage



Disadvantage






Research challenges

Convenienceanalysis




Communicationdelays

Deafnessproblem


Signalingoverhead



Leaderselection




















6 Conclusion




Competing Interests


Acknowledgments


References
























































Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in















6 Conclusion




Competing Interests


Acknowledgments


References
























































Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in





Competing Interests


Acknowledgments


References
























































Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in

































Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in



review article principles, applications ... - son systems lab

Documents