design and analysis of self-healing tree-based hybrid...

Research ArticleDesign and Analysis of Self-Healing Tree-Based Hybrid SpectralAmplitude Coding OCDMA System

Waqas A Imtiaz1 Affaq Qamar1 Atiq Ur Rehman2 Haider Ali1

Adnan Rashid Chaudhry2 and Javed Iqbal3

1Department of Electrical Engineering Abasyn University Peshawar Pakistan2Faculty of Computing and Information Technology Northern Border University Rafha Saudi Arabia3Sarhad University of Science and Information Technology Peshawar Pakistan

Correspondence should be addressed to Waqas A Imtiaz waqasahmed15gmailcom

Received 6 February 2017 Accepted 15 May 2017 Published 20 June 2017

Academic Editor Vincenzo Conti

Copyright copy 2017 Waqas A Imtiaz et alThis is an open access article distributed under theCreativeCommonsAttribution Licensewhich permits unrestricted use distribution and reproduction in any medium provided the original work is properly cited

This paper presents an efficient tree-based hybrid spectral amplitude coding optical code division multiple access (SAC-OCDMA)system that is able to provide high capacity transmission along with fault detection and restoration throughout the passive opticalnetwork (PON) Enhanced multidiagonal (EMD) code is adapted to elevate systemrsquos performance which negates multiple accessinterference and associated phase induced intensity noise through efficient two-matrix structure Moreover system connectionavailability is enhanced through an efficient protection architecture with tree and star-ring topology at the feeder and distributionlevel respectively The proposed hybrid architecture aims to provide seamless transmission of information at minimum costMathematical model based on Gaussian approximation is developed to analyze performance of the proposed setup followed bysimulation analysis for validation It is observed that the proposed system supports 64 subscribers operating at the data rates of25 Gbps and above Moreover survivability and cost analysis in comparison with existing schemes show that the proposed tree-based hybrid SAC-OCDMA system provides the required redundancy at minimum cost of infrastructure and operation

1 Introduction

Adaptation of passive optical network (PON) requires anefficient multiple access technology that can support maxi-mum number of subscribers at high data rates and extendedreach Time division multiplexed (TDM) PON is deployed toprovide the necessary broadband services in access domainHowever the time-sharing property of TDM PON limitsthe amount of data per subscriber Consequently wave-length divisionmultiplexed (WDM) technology is adapted toenhance network capacity in terms of data and the numberof subscribers Nevertheless deployment of WDM-PON islimited by the use of relatively expensive components atboth the transmitter and receiver modules Moreover thescarcity of available wavelengths confines the use of WDM-PON for large number of subscribers in next-generationPONs [1] Consequently optical code division multipleaccess (OCDMA) systems are expected to solve the last-mile

bottleneck between high speed coremetro networks andaccess domain [1ndash4]

Among all OCDMA systems for the access domain spec-tral amplitude coding (SAC) family has gained significantattention because of the desirable correlation properties flex-ibility of implementation and support for high transmissioncapacity [5] SAC-OCDMA coding schemes are a family of119870binary [01015840119904 11015840119904] sequences characterized by auto- and cross-correlation (120582119888) properties along with desirable length (119873)and weight (119908) [4ndash6] An optimal SAC-OCDMA codingscheme is one that provides desirable BER le10minus9 by reduc-ing multiaccess interference (MAI) and associated phaseinduced intensity noise (PIIN) through suitable correlationproperties minimum length large weights flexibility ofimplementation and reduced hardware complexity Severalcoding algorithms have been designed in SAC family toprovide high capacity along with extended reach however

HindawiSecurity and Communication NetworksVolume 2017 Article ID 2524513 12 pageshttpsdoiorg10115520172524513

2 Security and Communication Networks

FF

PIN

PIN

PIN

BBS

BBS

BBS

DFsDecoder1Encoder1

Decoder2Encoder2

Decoder3Encoder3

1N

OCP

N1

OCP

Figure 1 Conventional SAC-OCDMA system

the use of suitable binary sequence is still an open issue[3 5 7ndash10]

Another problem that limits the deployment of SAC-OCDMA system at access domain is the lack of a suitableprotection mechanism [13 14] Conventional SAC-OCDMAsystems are built with tree architecture that utilize a longfeeder fiber (FF) to connect optical line terminal (OLT) atCO with remote node (RN) at the subscriber premises asshown in Figure 1 End-face of the passive RN connectsmultiple optical network units (ONUs) through short spandedicated distribution fibers (DFs) Figure 1 shows the basicarchitecture of SAC-OCDMA system where no protection isprovided between CO andONUs [3 4]This significantly ele-vates the loss of data in case of failures at optical components(including fibers) Therefore it is imperative to design anefficient protection mechanism that facilitates the detectionand recovery of cutsfailures throughout the network inorder to ensure the success of SAC-OCDMA technology atthe access domain

This paper proposes an efficient SAC-OCDMA sys-tem with high transmission capacity through enhancedmultidiagonal (EMD) code Moreover a robust tree-basedhybrid protection architecture is proposed for survivabilityenhancement along with self-healing capacity at both feederand distribution fibers The proposed solution includes anew star-ring protection architecture that provides pre-planned redundancy without duplicating the entire fibersetup at distribution network Performance of the proposedtree-based hybrid system is analyzed using comprehen-sive analytical and simulation analysis by referring to biterror rate (BER) and eye patterns Redundancy operationof the proposed system is also observed for failures atboth feeder and distribution fibers Furthermore capitalexpenditure (CAPEX) and operation expenditure (OPEX)of the proposed setup are also computed to determine thefeasibility of implementation in comparison with existingsolutions

Rest of the paper is organized as follows Section 2 dis-cusses the construction of EMD code followed by a detaileddesign and implementation of the proposed architecture inSection 3 Analytic and simulation analysis for transmissioncapacity of the proposed setup is performed in Section 4Section 5 covers the survivability analysis of the proposedtree-based SAC-OCDMA system for failures at the feederand distribution levels Detailed cost analysis is presented inSection 6 that is followed by the final conclusion in Section 7

2 Enhanced Multidiagonal Code

EMD code is used in the proposed system to achieve therequired transmission capacity in terms of data reach andnumber of subscribers EMD code is designed with two-matrix structure including data matrix (119863119899) and code matrix(119862119899) Chips in 119863119899 are used to encode the data bits afterspectral translation whereas 119862119899 chips refer to the code bitsthat are embedded to elevate the performance of EMD basedSAC-OCDMA system [15] Moreover EMD coding algo-rithm is designed with three basic performance parametersincluding119873119908 and120582119888 between adjacent codesThe followingsection outlines the properties of119863119899 and 119862119899 followed by theimplementation of EMDmatrix

EMD = [119863119899 | 119862119888]119870times119873 (1)

21 Data Matrix 119863119899 consists of a 119870 times 119870 diagonal matrixwhere119870 rows and119870 columns of thematrix represent numberof subscribers and length of 119863119899 respectively Furthermore119863119899 is designed with multiple rows of length 119870 119908 = 1 and119888 = 0 Therefore the basic 2 times 2 119863119899 can be written as

119863119899 = [1 00 1] (2)

22 Code Matrix 119862119899 consists of a 119870 times 119869 matrix where 119870rows and 119869 columns represent the number of subscribersand length of the matrix respectively 119862119899 for EMD code isdesigned with multiple rows of 119869 = 119870 + 1 119908 = 2 and 119888 = 1between the codes in adjacent rowsThe basic 2times3 119862119899 can bewritten as

119862119899 = [0 1 11 1 0] (3)

Consequently the proposed EMD code while using (1)and the properties of119863119899 119862119899 is given by

EMD2 = [1 0 | 0 1 10 1 | 1 1 0]2times5

(4)

Equation (4) shows that 10158401 2 11015840 combination is main-tained between adjacent EMD codes The number of sub-scribers in EMD code can be increased through simplediagonal mapping while maintaining the respective prop-erties of 119863119899 and 119862119899 This increases the cardinality of SAC-OCDMA systemwithminimum impact on119873 for EMD codeConsequently EMD code matrix for 4 subscribers is given by

EMD4 =[[[[[[

1 0 0 0 | 0 0 0 1 10 1 0 0 | 0 0 1 1 00 0 1 0 | 0 1 1 0 00 0 0 1 | 1 1 0 0 0

]]]]]]4times9

(5)

Security and Communication Networks 3

OCP

OCP

PINLPFDM

PINLPFDM

OCP

MZM

MZM

MZM

MZM

PIN LPF DM

LED

Data

MZM

1

OS

DFsPF

ONU

os3

21

2

2

Encoder

Encoder

Encoder

Encoder

Encoder

EDFA

OCP

1

BPF

BPF

BPF

ONU1

LEDM

LED1

ONU2

ONU3

ONU4

ONU5

ONU6

ONUMminus1

ONUM

OLTTx

OLTRx

FFp

FFs

2N

OCP

Figure 2 Proposed tree-based hybrid SAC-OCDMA system

23 Properties of EMD Code

(i) EMD code can be designed with 119908 ge 3(ii) Length119873 of each code is equal to119870+ [119870(119870minus 2) + 1](iii) 120582119888 = 1 between adjacent codes only of the EMD

matrix(iv) Auto- and cross-correlation properties of EMD code

when119861119896(119894) denotes the 119894th element of119870th EMD codeare given by

119873sum119894=1

119861119896 (119894) 119861119897 (119894) =

119908 119896 = 1198971 119896 = 119897 for adjacent code0 119896 = 119897 otherwise

(6)

3 Proposed SAC-OCDMA System

The proposed tree-based hybrid SAC-OCDMA system con-sists of the following components

31 Central Office Central office is located at the serviceprovider premises and houses multiple OLT ports whichgenerates required services towards the PON distributionnetwork OLT consists of the following

311 Optical Source Optical source is crucial in designingan efficient and reliable SAC-OCDMA system because of itssignificant effect on the cost and performance of PON [3 16]OCDMA systems can be classified into two types accordingto the choice of optical sources coherent and in-coherentOCDMA Prior system requires highly coherent ultra-shortlight pulses (MLL supercontinuum laser etc) in order tomaintain nominal BER in high capacity systems Such sourcesare difficult to tune and involve precise laser positioning to

assure efficient performance This significantly increases thecost and complexity of the OCDMA system

Moreover localized power distribution in coherent opti-cal sources requires complex equipment to translate binary01015840119904 and 11015840119904 into spectral representation Most of the encodersrequire complex tuning and fast electronic circuits thatfurther elevates the cost and complexity of implementationfor coherent SAC-OCDMA systems at the last-mile sectionof the network [17]

Broadband sources (BBSs) with relatively same powerdistribution throughout their bandwidth facilitate therequired encoding at the expense of simple setup MoreoverBBSs like light emitting diodes (LEDs) have shown efficientperformance with simple in-coherent decoding techniquesincluding balanced and direct detection [16] FurthermoreSAC-OCDMA systems are targeted to support economicalaccess networks therefore optical source with low costsimple drive circuits flexibility of encoding immunityto beat noise and stable performance are suitable forimplementation of the proposed tree-based hybrid SAC-OCDMA system for PON

Thus OLT for the proposed system contains an array ofLEDs which are tuned at different wavelengths to supportdesirable number of subscribers Output from the LED isfed into 1 3 optical coupler (OCP) in order to provide therequired spectrum for code chips at input of the encoder asshown in Figure 2

312 Encoder Design of a suitable encoder is observedas of critical concern for the implementation of SAC-OCDMA system in terms of functioning and ability toperform the required encoding Various devices have beenproposed for encoding operation in SAC-OCDMA systemhowever the use of suitable device is still an open issue[3 4 18 19]


Figure 3 Spectral representation of encoded [0 0 1 0 1 1 0 0 0] bits in EMD code

This paper utilizes optical multiplexer (MUX) for trans-lating the proposed binary sequence into spectral repre-sentation at OLT [7 8] Each input of the optical MUXcontains a bandpass filter which is tuned to select the desiredspectrum in accordance with EMD chips Figure 3 representsthe implementation of EMD code [0 0 1 0 1 1 0 0 0]through a 3 1 MUX with 04 nm bandpass Bessel filters atthe input It is observed that the proposed arrangement canefficiently perform the required binary to optical conversionwith maximum precision and simple architecture

End-face of the optical encoder is connected with Mach-Zehnder Modulator (MZM) which combines optical signalwith data from individual subscriber Output of the modulat-ing arrangement is applied to N 1 OCP with 119873 input and 1output port as shown in Figure 2 OCP is further connectedwith a three-port optical circulator (OCco) that is fed intoswitching arrangement (SA) through an erbium doped fiberamplifier (EDFA)

313 Switching Arrangement Switching arrangement (SAco)consists of a 1 2 optical switch (OSsa) that connects OLTwith both primary and secondary feeder fibers (FFp FFs) asshown in Figure 2 OSsa is used to perform the necessaryswitching between FFp and FFs when a cutfailure occurs atthe feeder fiber Under normal mode of operation OSsa is atposition 1 and all traffic is supplied through FFp Positionof the proposed OSsa is controlled by OLT medium accesscontrol (MAC) layer which is responsible for fault detectionand restoration at the feeder level Consequently no extraarrangement is required for changing OSsa positions forseamless transmission between OLT and ONUmodules

32 Remote Nodes The proposed tree-based hybrid SAC-OCDMA system utilizes a single remote node to facilitatecommunication between CO and subscriber premises RNhouses a 2 119873 OCP that connects each ONU with both FFpand FFs at the input and short span dedicated DFs at theend-face as shown in Figure 2 RN facilitates ubiquitoustransmission of down- and upstream traffic by simultane-ously connecting each ONU with both FFp FFs fibers Thisallows the proposed system to immediately restore the flowof affected traffic in case of failure across the network

33 Optical Network Unit Optical network unit ONU119898119898 = 1 2 3 119872 consists of a 1 2 OCP with 3 ports

as shown in Figure 2 Ports 1 and 2 are connected todedicated DF119898 and a 2 1 optical switch OSonu119898 respec-tively Port 3 extends the ring based protection fiber (PFr)towards its neighboring ONUs Ring based redundancyarrangement provides the required preplanned protectionwithout duplicating entire fiber at the distribution networkMoreover use of star-ring topology eliminates the significantpower loss that is encountered in single-ring based PON[13 20ndash22]

OSonu119898 is at position 1 under normal mode of operationand all traffic is being sent and received though dedicatedDFsWhile in case of failure at theDF optical switch ismovedtowards position 2 which connects the effected ONU withredundant PFr Fault detection and switching operation atdistribution level are controlled through ONU119898 MAC layerin order to avoid extra arrangement for network survivabilityat ONU modules

End-face of OSonu119898 is fed into a three-port opticalcirculator (OConu) which is further connected with ONU119898receivingmodule as shown in Figure 2ONU receivermoduleconsists of a bandpass filter (for SDD technique) whichrecovers the nonoverlapping spectral chip 119863119899 only Outputof decoding arrangement is connected with PIN photodiodewhich is fed into Bessel low pass filter (LPF) followed bydecision-making (DM) circuit

331 Spectral Direct Detection Technique SDD techniqueis readily employed for coding schemes with ideal cross-correlation properties owing to its significant advantages likesimple implementation minimum expenditure and removalof MAI and associated PIIN at the receiving photodiodeThe principle of SDD technique is based on the recoveryof nonoverlapping spectral chip only rendering maximumautocorrelationwith the intended signal andminimumcross-correlation with interfererrsquos [7 23]

This technique is implemented through a simple band-pass filter which is centered to recover the nonoverlappingspectral chip only However it is necessary to send theentire coded spectrum in order to maintain integrity ofthe full address along with efficient power at the receivingphotodiode Figure 4 shows the implementation of SDDtechnique at the receiving ONU in the proposed setupDecoding arrangement consists of a single bandpass filterwhich is centered to recover 119863119899 chip only This operation


Table 1 Description and specification of noise variance parameters in (8)

Description Parameters ValuesElectron charge 119890 1602 times 10minus19Receiver noise equivalent electrical bandwidth 119861 500MHzBoltzmannrsquos constant 119870119887 138 times 10minus23Receiver noise absolute temperature 119879119899 300KReceiver load resistance 119877119871 1030ΩQuantum efficiency 120578 06Plankrsquos constant ℎ 66260 times 10minus34Optical source bandwidth Δ120592 375 times 1012Optical wavelength 120582 1550 nmEffective power of BBS at the receiver 119875sr minus10 dBm

MZM

Data

FFPINEncoder Filter

1N

OCP

N1

OCP

Figure 4 Spectral direct detection technique

eradicates MAI and associated PIIN since the overlappingspectrum at the transmitted 119862119899 is removed before the signalis converted to electrical domain [4 7]

4 Performance Analysis

Conversion of the received optical signal into electricaldomain is accompanied with shot PIIN and thermal noisesgenerated at the PIN photodiode Presence of these noisesources significantly reduces quality of the received sig-nal which elevates BER at the receiving end Thereforeit is imperative to determine performance of the pro-posed tree-based hybrid SAC-OCDMA system in the pres-ence of shot PIIN and thermal noises at the receivingphotodiode

41 Analytic Analysis Performance of the proposed tree-based hybrid SAC-OCDMA system is analyzed with refer-ence to signal to noise ratio (SNR) and BER of EMD code atthe encoder and SDDat the decoder Gaussian approximationis used to evaluate the performance while considering noisepowers at the receiving end [4] If 119868 represents the incidentcurrent 1198682 represents the power spectral density (PSD) and1205902 gives noise variance at the receiving photodiodeThen thetotal SNR is given by [8 15]

SNR = [ 11986821205902] (7)

where 1205902 is the sum of shot ⟨1198942shot⟩ PIIN ⟨1198942PIIN⟩ and thermal⟨1198942thermal⟩ noise powers which can also be written as

1205902 = 2119890119861119868 + 1198682119861120591119888 + 4119870119887119879119899119861119877119871 (8)

Table 1 shows the specifications and description of 1205902parameters in (8) Moreover the coherent time 120591119888 of intensitynoise terms in the photodiode output is given by

120591119888 = intinfin119900119866 (120592)2 119889120592

[intinfin119900119866 (120592) 119889120592]2 (9)

119866(120592) in (9) shows the PSD at the receiving photodiodeIn order to analyze performance of the proposed tree-based hybrid SAC-OCDMA system while using Gaussianapproximation it is assumed that each optical source isideally polarized with smooth spectrum throughout its band-width Moreover PIN photodiodes receive identical spectralwidth and power (119875sr) from each optical source [4 7 8]

If 119861119896(119894) denotes the 119894th element of 119870th EMD codethen the correlation properties at SDD based decoder if therequired information is recoverable from 119863119899 chip only aregiven by

119873sum119894=1

119861119896 (119894) 119861119897 (119894) = 119908 forall119896 = 1198970 forall119896 = 119897 (10)

For the incident current

119868 = Rintinfin119900

119866 (120592) 119889120592 (11)

PSD under the effect of various noise sources is given by

119866 (120592) = 119875srΔ120592119870sum119896=1

119889119896119873sum119894=1

119861119896 (119894) 119861119897 (119894) [119906 (Δ120592119873 )] (12)

119875sr Δ120592 and 119906 in (12) show the power at input of thePIN photodiode line-width of LED and unit step function


respectively For total 119868 at the receiving photodiode (12) isintegrated from 0 toinfin

intinfin0

119866 (120592) 119889120592

= intinfin0

119875srΔ120592119870sum119896=1

119889119896119873sum119894=1

119861119896 (119894) 119861119897 (119894) 119906 [Δ120592119873 ]119889120592(13)

sum119870119896=1 119889119896 in (13) represent the transmitted bit either 0 or1 by a single subscriber Equation (13) can be simplified byusing the correlation properties in (10) which leads to

intinfin0

119866 (120592) 119889120592 = 119875sr119908119873 (14)

Now the total incident current from (14) can be written as

119868 = R119875sr119908119873 (15)

R represents responsitivity of the receiving photodiodegiven by R = 120578119890ℎ120592119888 Table 1 shows the description andspecification of R parameters Thus total shot noise powerdue to a single LED after using (15) and (8) becomes

⟨1198942shot⟩ = 2119890119861R119875sr119908119873 (16)

It is assumed that the required information is recoverablefrom the nonoverlapping 119863119899 chip only which is directlydetected by the receiving photodiode as in normal intensitymodulationThus PIINnoise power is set equal to zero due tothe large difference between the auto- and cross-correlationproperty amongst the intended subscriber and interfererrsquos[4 7 23]

Consequently the total SNR based on (7) (8) and (16)when the probability of each user transmitting bit 1 at anyinstant is 12 can be written as

SNR = [R119875sr119908119873]2119890119861R119875sr119908119873 + 4119870119887119879119899119861119877119871 (17)

The estimated BER while using Gaussian approximationfor the proposed tree-based hybrid SAC-OCDMA systemcan be calculated as BER = (12)erfcradicSNR8 Figure 5shows transmission capacity of the proposed system byvarying number of subscribers at performance parameters inTable 1 It is shown that the proposed systemprovides efficientperformance in terms of SNR and BER as we increase thenumber of subscribers Fixed cross-correlation property ofEMD code reduces interference from adjacent subscribers [815] Moreover SDD technique receives the nonoverlappingspectral chip only which results in maximum differencebetween auto- and cross-correlation functions at the receiv-ing photodiode [4 7] Therefore SNR and related BER areoptimized since bothMAI and associated PIIN are eliminatedbefore the signal is converted to electrical domain

Figure 6 shows systemrsquos capacity at different data rates formultiple subscribers while using system parameters in Table 1

SNRBER

0

80

160

240

320

400

480

560

640

720

800

Sign

al to

noi

se ra

tio (S

NR)

80 8856 64 7240 48 9632 112 120 128104

Number of subscribers

1E minus 33

1E minus 30

1E minus 27

1E minus 24

1E minus 21

1E minus 18

1E minus 15

1E minus 12

1E minus 9

1E minus 6

1E minus 3

Bit e

rror

rate

(BER

)

Figure 5 BER and SNR versus number of subscribers at differentvalues of 119875sr

36 40 44 48 52 56 60 6432


1E minus 40

1E minus 36

1E minus 32

1E minus 28

1E minus 24

1E minus 20

1E minus 16

1E minus 12

1E minus 8

1E minus 4

Bit e

rror

rate

(BER

)

15 Gbps2Gbps25 Gbps

3Gbps4Gbps5Gbps

Figure 6 BER versus data rate at different number of subscribers

[15] It is shown that quality of received signal deteriorates aswe increase the amount of data at the transmitter Pulse widthof the coded signal is inversely proportional to the number oftransmitted bitsTherefore increase in data reduces the pulsewidth whichmakes the systemmore vulnerable to the effectsof dispersion across the medium However it is observedthat the proposed system is able to handle data rates beyond25Gbps while maintaining efficient performance parameterthat is BER of 10minus9 owing to its efficient correlation proper-ties in EMD code and recovery of119863119899 chip only through SDDtechnique


Table 2 Simulation parameters

System parameters ValuesFiber attenuation 025 dBkmDispersion 18 psnmkmFiber length 25 kmPIN responsitivity 075AWPIN cutoff frequency 075 times Bitrate119875sr minus10 dBm

BER (analytical)BER (sim B2B)BER (sim fiber)

125 200 225100 150 250 275175

Data rate (Gbps)

1E minus 34

1E minus 32

1E minus 30

1E minus 28

1E minus 26

1E minus 24

1E minus 22

1E minus 20

1E minus 18

1E minus 16

1E minus 14

1E minus 12

1E minus 10

1E minus 8

Bit e

rror

rate

(BER

)

Figure 7 Performance analysis of the proposed tree-based hybridSAC-OCDMA system

42 SimulationAnalysis OptiSystem is used to analyze trans-mission capacity of the proposed tree-based hybrid SAC-OCDMA system in comparison with analytic model Anal-ysis is performed for 64 subscribers at back-to-back (B2B)and 25 km single mode fiber (SMF) fiber while using systemparameters shown in Table 2 Moreover EDFA gain andcomponent losses are adjusted tomaintainminus10 dBmpower atthe receiving photodiodes BER and eye pattern are observedat different data rates for performance analysis

Figure 7 shows transmission capacity of the proposedarchitecture in comparison with analytic and B2B results Itis observed that simulation and analytic analysis shows prox-imity with regard to BER for multiple subscribers simultane-ously accessing themediumwhich validates the performanceanalysis Moreover it is observed that BER increases for B2Bfollowed by fiber based simulation analysis It is evident fromthe fact that fiber introduces dispersion and attenuation to theencoded spectrum which further deteriorates performanceof the system However it is not significant enough to disruptthe normal operation Figure 7 also shows that the use of extracomponents like switches OCPs and so forth for protectionhas minuscule (almost negligible) effect on performance ofthe proposed SAC-OCDMA system

Figure 8 shows eye patterns from the simulation analysisat data rates of 1 15 and 2Gbps It is observed that eye

opening decreases as the amount of data increases howeverthe proposed system is able to provide efficient performanceby maintaining nominal eye openings at data rates of up to2Gbps

5 Survivability Analysis

This section outlines the fault detection and restorationcapacity of the proposed setup in terms of system initiationdetection and recovery at both feeder and distribution level

51 Initiation It is assumed that the proposed system is undernormalmode of operationOLTwill start and check the statusof FFp before transmitting any traffic OLT MAC layer willuse discovery gate (DG) packets [13] to check the status ofFFp If OLT receive acknowledgment (ACK) messages fromall or someONUmodules for the correspondingDGpacketsit will start its transmission using FFp However if no ACKmessages are received OLT will flip the switch towards FFsand check its status before transmission OLT will use thesame mechanism by sending DG packets to the connectedONUs for the status of FFs On the reception ofACKmessagesfrom ONUs OLT will start its operation via FFs In case ofnegative ACKs OLT will flip the switch towards its originalposition and wait for a specific time 119879s before starting theentire process again

Similarly ONUs will check for the connected DFs afterinitiating their normal operation They will also send a DGpacket towards OLT to check the working status of theircorresponding DF For positive ACK messages ONU willcontinue its normal operation through DF In case of noACKmessages ONUMAC layer will flip the switch positiontowards the redundant PF

While using the same mechanism ONU will check thestatus of the PF by sending DG packets towards OLT Forpositive ACKmessages ONU will resume its operation usingPF whereas in case of negative ACK messages ONU MAClayer will flip the switch towards its initial position and waitfor a specific time 119879s before starting the entire process again52 Feeder Fiber Failure To understand self-healing mech-anism of the proposed protection architecture it is assumedthat system is under normal mode of operation and all trafficfrom OLT is delivered through FFp During transmissionOLT will listen to the loss of signal (LOS) which representsthe absence of upstream signals from the connected ONUsIn case of no LOS OLT will continue its normal operationthrough FFp with OSsa at port 1

If failure or cut occurs at point 10158401198911015840 in FFp as shownin Figure 9 OLT will stop to receive any upstream trafficfrom the access domain At the same time OLT will notbe able to transmit any downstream traffic to the con-nected ONUs since FFp is the only transmission mediumbetween OLT and ONUs This will trigger the LOS flag atOLT

Consequently OLT MAC layer will initiate its recoverymechanism and send DG packets towards ONU modulesDG packets will determine the status of FFp concerning ACK


(a) (b) (c)

Figure 8 Eye patterns for the proposed tree-based hybrid SAC-OCDMA system at (a) 1 (b) 15 and (c) 2Gbps of data

OS

1

2MAC

f

OLTTx

OLTRx

FFp

FFs

Figure 9 Self-healing mechanism for failure at FF

messages from ONUs that is if OLT receives ACK from allor some ONUs for the corresponding DG packets it willcontinue its normal operation However if no ACKmessagesare received OLT will retransmit the DG packets to confirmthe fault at the connected FFp After several attempts if noACKmessages are received from the access arena OLTMAClayerwill confirm that FFp is faulty and flip the switch towardsport 2 MAC layer will check the status of FFs through DGpackets Section 51 before transmitting information Forpositive ACK messages MAC layer will direct the OLT tostart transmission through the newly switched FFs as shownin Figure 9

OLT will continue its normal operation while using FFsand listen to the LOS until FFp is repaired It must be notedthat fault at FF will also disconnect the ONUs in accessdomain Consequently they will also initiate their recoverymechanism to resolve the flow of affected traffic whichwill lead to unnecessary switching at ONUs Therefore itis imperative to increase resolution time and mechanism offault detection in ONU modules as compared to OLT Thiswill avoid the unnecessary switching atONUs before the issueis resolved at the affected OLT

53 Distribution Fiber Failure Effect of failure at theDF is notas severe as that of the FF since the flow of traffic is disruptedto a single ONU Nevertheless this needs to be addressedin order to provide high subscriber satisfaction and reduceservice interruption penalty costs [11] in case of businesssubscribers at the access domain To explain the self-healing

mechanism at distribution level of the network ONU2 isconsidered which is under normal mode of operation OS2is at position 1 and all traffic is sent and received throughthe corresponding DF During transmission ONU2 will alsolisten to LOS which represents the absence of downstreamsignals from connected OLT In case of no LOS ONU2 willcontinue its normal operation through its corresponding DFwith OS2 at port 1 In case of failure at point 10158401198881015840 in the DFONU2 will not receive any downstream traffic from OLTSimilarly OLT will not receive any upstream traffic fromONU2 This will trigger LOS flag at ONU2

Thus ONU2 will initiate its recovery mechanism andMAC layer will send DG packet towards OLT in order toconfirm the status of connected DF On the reception ofACK messages from OLT ONU2 will continue its normaloperation with OS2 at position 1 If no ACK message isreceived from OLT ONU2 will perform several attempts toconfirm the status of connected DF The number of DGpackets sent by ONU2 must be twice as that of the ones sentby OLT in order to avoid any unnecessary switching

For repeated negative ACKs ONU2 will confirm thestatus of failure and flips the switch towards port 2 Con-sequently the flow of traffic will shift from DF to the DRbetween ONU2 and ONU3 as shown in Figure 10 Now alltraffic from ONU2 will be sent and received by the DF ofONU3 through the intermediate PF

It must be noted that each affected ONU can restore theflow traffic by reconnecting with OLT through its neighbor-ing ONUs in case of failure at the DF Based on the sameprinciple if failure or cut occurs at several DFs the affectedONUs can share the DFs with working ONUs to reconnectwith OLT

54 Restoration Time The time required for fault detec-tion and restoration can be determined by referring to theautodiscovery process in IEEE 8023ah [20] Let119879dg representthe time required by OLT to trigger LOS flag and send aDG packet towards the ONU modules Fault discovery isstarted after OLT sends the DG packets and takes 2119879cycle tocomplete the registration process from a single ONU In caseof negative acknowledgment OLT will retransmit the DG


2

1

MAC

c

c

OCP

DFs PF

OS

ONU1 ONURx

ONUTx

ONU2

ONU2

ONU3

ONU4

ONUM

2N

OCP

Figure 10 Self-healing mechanism for failure at DF

packet and wait for 2119879cycle to receive ACK from at least oneONU module If no ONU register to the DG packet aftertwo queries OLT MAC layer will flip the switch positionConsequently the total time between fault detection andswitching operation at feeder fiber failure is 119879dg + 2119879cycle +2119879cycle + 119879switching

ONU MAC layer adapts the same DG packets forfault detection at the distribution level Therefore the timerequired for fault detection and restoration at ONUmodulescan be determined through same methodology However itis necessary to increase the number of DG packets at ONUmodules in order to avoid unnecessary switching for the eventof failure at FFp Therefore the time required by ONUMAClayer for fault detection and switching operation is to be givenby 119879dg + 2119879cycle + 2119879cycle + 2119879cycle + 2119879cycle + 119879119904witching55 Power Budget and Network Capacity Power budgetrepresents the sum of losses at each component between OLTand ONUs The purpose of power budgeting in PONs is toensure the efficient delivery of transmitted data with nominalpower ratings since low power at the PIN photodiode cangenerate substantial amount of short and thermal noiseMoreover it also determines the OCP splitting ratio atRN which translates the number of subscribers in PONTherefore it is necessary to analyze power budget of theproposed system in order to ensure high capacity

For efficient recovery of transmitted information it isimperative that power received at the PIN photodiode mustbe greater than its 119877sen Table 3 shows the insertion losses atdownstream of the proposed tree-based protection architec-ture for SAC-OCDMA system [14]

If 119875CO represents the power drop across the central officewhich is the sum of insertion losses and gains of opticalcomponents at the CO as shown in Table 3 119875RN gives thepower loss across the FF and 2 119873OCP and 119875DN is the powerloss across the distribution network components then thedownstream power budget 119875DS can be calculated as

119875DS = 119875119879 minus ILOLT minus ILOCco+ 119875EDFA minus ILOSco minus 120572LFF

minus 119883ILOCP2119873RN minus 120572LDF minus ILOCPonuminus ILOSonu

minus ILOConuminus ILONU

(18)

Table 3 Description and specification of network components withpower loss

Description Parameters ValuesTransmitter power 119875119879 0sim10 dBmInsertion loss OLT ILOLT 3d BmReceiver sensitivity 119877sen minus16simminus25 dBmInsertion loss OS ILos 05 dBAmplifier gain 119875EDFA 25 dBInsertion loss OC ILOC 025 dBInsertion loss ONU ILONU 1 dBPropagation loss 120572 025 dBkmInsertion loss 1 2 OCP 10 log(2) 3 dBTable 4 Downstream power budget analysis for proposed architec-ture

System components DownstreamBBS power [dBm] 0CO insertion loss [dB] 37525Km fiber loss [dB] 6252 64 OCP insertion loss [dB] 18Insertion loss [dB] 475Total losses minus3275EDFA gain [dB] 25Received power [dB] minus775Receiver sensitivity [dBm] minus24

For efficient recovery of transmitted information it isimperative that

119875DS ge 119877sen (19)

Table 4 shows the downstream power budget for theproposed tree-based hybrid protection architecture whileusing insertion losses fromTable 4 It is shown that utilizationof EDFA at CO can elevate power of the encoded spec-trum Therefore it can be detected and efficiently recoveredat the receiving photodiode Moreover without the useof EDFA at the transmitter module the value of 119875sr =minus10 dBm This value is significantly lower than requiredfor an error-free recovery Therefore a gain of 10 dB andabove is required for error-free transmission between OLTand ONU receiver module Same analogy applies on theupstream traffic which uses identical encoding and decodingarchitecture at ONU transmitter and OLT receiver modulerespectively

Furthermore downstream power budget in Table 4is determined with 2 64 OCP between OLT and ONUsIt is evident that end-face of RN represents the numberof ONUs in the access domain Therefore the proposedsystem can easily support 64 subscribers with nominalpower ratings at the receiving photodiode Furthermoreincrease in EDFA gain within minus5simminus30 dB can provide sup-port for desired number of subscribers in the proposedarchitecture


Table 5 Components description connection availability and cost [11 12]

System components Cost ($) Energy consumptionOLT 12100 20W 25W (with EDFA) 2W (standby)ONU 350 1W 025W (standby)Optical circulator 50Optical switch 501 2 2 2 CPR 501 119873 2 119873CPR 800Fiber (Km) 160

6 Cost Analysis

Cost figures of protection architectures are an importantparameter showing the economic benefits for a common enduser at the access domain This section determines overallexpenditure of the proposed architecture in comparison withexisting ITU-T type C D and hybrid architectures whileusing component costs in Table 5 [11 24] Cost figuresfor PON can be categorized as CAPEX and OPEX [1425] CAPEX represents the investment that is required forthe installation and deployment of PON in access domainCAPEX is calculated by computing the cost of networkequipment and fiber infrastructure Moreover CAPEX persubscribers is determined by dividing the sum of networkequipment and fiber infrastructure costs over the number ofsubscribers

OPEX is related to the cost of network operation fromthe time of deployment till replacement by a new technologyOPEX mainly consists of failure reparation cost serviceinterruption penalties and power consumed by active com-ponents at OLT and ONUs Reparation cost depends on thenumber of failures during the network lifetimeMean lifetimeandMTTRof each component determine the total downtowntime per year which is multiplied by the number of tech-nicians required and their salaries in order to calculate theyearly failure reparation cost Service interruption penaltiesinclude the expenditure that is spent on the fine defined inservice level agreement (SLA) between network operatorsand subscribers If the service interruption time is higherthan a threshold mentioned in SLA the operator has to pay afee depending on the interval that the service is unavailablePower consumption for OPEX is calculated by multiplyingthe sum of energy consumption of all active componentsduring the network lifetime with the unit price of electricalenergy [11 12 25 26]

Analysis is performed for residential customers only andlife span of the system is taken as 20 years Furthermore

(i) each OLT contains 4 ports with 16 subscribers perport

(ii) each architecture contains 20 km FF 5 km DF and1 km protection fiber between adjacent ONUs

(iii) cost of TDM and OCDMA based OLT and ONUs istaken to be the same

(iv) due to high variation in fiber digging cost it isassumed that fiber is laid in existing trenches at bothfeeder and distribution level

(v) no penalty cost is considered as only residentialcustomers are considered

(vi) repairing cost is 1000 $h(vii) per hour cost of electricity is taken to be 025 $kWh

If 119872 represents the number of PONs and 119873 is the totalnumber of subscribers then CAPEX equations (base onRBDs) for the considered protection architectures can bewritten as

119862[119862] = (119872 timesOLT timesOLT) + (20 times FF times FF)+ (119872 times 2 times 2 119873 OCP)+ (5 times 119873 times DF times DF)+ (119873 timesONU timesONU)

119862[119863] = (119872 timesOLT timesOLT) + (20 times FF times FF)+ (119872 timesOCP timesOCP)+ (119872 times 2 times 2 119873 OCP)+ (5 times 119873 times DF times DF)+ (119873 timesONU timesONU)

119862[119867] = (119872 timesOLT timesOLT) + (20 times FR)+ (119872 timesOCP) + (119872 times 2 119873 OCP)+ (5 times 119873 times DF) + (119873 timesONU)

119862[PT] = (119872 timesOLT) +OS + EDFA + (20 times FF times FF)+ (119872 times 2 119873 OCP) + (5 times 119873 times DF)+ (119873 times DR) + (119873 timesOCP) + (119873 timesOS)+ (119873 timesONU)

(20)

Figure 11(a) shows the deployment cost of different pro-tection architectures It is observed that the proposed archi-tecture requires minimum CAPEX in comparison with typesC D and hybrid schemes It is evident from the fact that typesC and D duplicate the entire PON which requires significant


[D] [H] [PT][C]Protection architectures

2000

2200

2400

2600

2800

3000

3200

3400

3600

3800

4000

Capi

tal e

xpen

ditu

re ($

)

(a)

Reparation costEnergy consumptionTotal


0

50

100

150

200

250

300

350

400

450

Ope

ratio

nal e

xpen

ditu

re ($

)(b)

Figure 11 CAPEX and OPEX of existing architectures in comparison with proposed tree-based hybrid SAC-OCDMA system

amount of CAPEX for deployment Furthermore duplicationof OLT in hybrid architecture increases the overall cost ofdeployment as OLT is the most expensive active componentin PON shown in Table 5 Deployment cost of the proposedtree-based protection architecture is relatively low since itavoids duplication of OLT at the COMoreover efficient star-ring redundancy at the distribution level further reduces thecost of deployment by avoiding extensive fiber duplicationAs this architecture utilizes minimum components in systemconstruction therefore energy consumption in Figure 11 isalso significantly low

Figure 11(b) shows expenditure spent on the operation ofPONs after deployment It is observed that hybrid schemerequires the highest OPEX since no protection is provided atthe distribution level This significantly elevates the networkdowntime per year andmore resources are spent on detectingand recovering faults Figure 11(b) shows that the proposedarchitecture along with types C and D requires minimumreparation cost owing to desirable connection availabilityand minimal downtime per year Moreover no duplicationat OLT and ONU significantly reduces energy consumptionof the proposed architecture as compared to types C D andhybrid networks

Consequently the proposed tree-based hybrid SAC-OCDMA system requires minimum cost of deployment andoperation as shown in Figure 12 due to

(i) efficient star-ring topology that reduces the amountof DF and hence the cost required for deployment

(ii) no redundant OLT and ONU which reduce thedeployment cost and energy consumption over thenetwork life time

(iii) fault detection and restoration at both feeder andthe distribution level which significantly reduce the

CAPEXOPEXTotal cost


0

500

1000

1500

2000

2500

3000

3500

4000

4500

Tota

l cos

t ($)

Figure 12 Total cost for different protection architectures

network downtime per year Consequently minimumresources are spent on repairing

7 Conclusion

Efficient operation of PON requires a reliable architecturewhich is able to provide high network capacity along withfault detection and restoration mechanism This paper pro-poses a novel tree-based hybrid SAC-OCDMA system withEMD code and two-tier protection EMD code significantlyelevates performance of the proposed system by negatingMAI and PIIN through efficient correlation properties along


with nominal values of 119873 and 119908 Analysis shows that EMDcode along SDD technique enables the proposed system tosupport high data rates at desirable number of subscribersMoreover protection at both feeder and distribution net-works ensures ubiquitous transmission of affected traffic fromfaulty to working fiber Furthermore star-ring protection andMAC controlled switching enable the proposed system toprovide the necessary protection at minimum power lossCAPEX and OPEX Thus the proposed tree-based hybridSAC-OCDMA system is highly suitable for future PONsowing to its significant advantages like support for highcapacity in terms of data and the number of subscribers easeof implementation relatively simple architecture immediatefault detection and restoration at both feeder and distributionnetworks and minimum cost as compared with existingsolutions

Conflicts of Interest

The authors declare that they have no conflicts of interest

References

[1] E Wong ldquoNext-generation broadband access networks andtechnologiesrdquo Journal of Lightwave Technology vol 30 no 4Article ID 6094146 pp 597ndash608 2012

[2] XWangNWada TMiyazaki G Cincotti andK-I KitayamaldquoHybrid WDMOCDMA for next generation access networkrdquoin Asia-Pacific Optical Communications pp 678328-678328nternational Society for Optics and Photonics 2007

[3] H Yin and D J Richardson ldquoOptical code division multipleaccess communication networksrdquo chap vol 1 pp 36-37 2008

[4] H Ghafouri-Shiraz and M M Karbassian Optical CDMANetworks Principles Analysis and Applications John Wiley ampSons 2012

[5] N Din Keraf S A Aljunid A R Arief and P Ehkan ldquoTheevolution of double weight codes family in spectral amplitudecoding ocdmardquo Advanced Computer and Communication Engi-neering Technology pp 129ndash140 2015 Springer

[6] F Su and H Jin ldquoResearch of code construction for ocdmasystemrdquo in Advances in Mechanical and Electronic Engineeringpp 265ndash270 Springer 2013

[7] T H Abd S A Aljunid H A Fadhil R A Ahmad and NM Saad ldquoDevelopment of a new code family based on SAC-OCDMA system with large cardinality for OCDMA networkrdquoOptical Fiber Technology vol 17 no 4 pp 273ndash280 2011

[8] H A Fadhil S A Aljunid and R B Ahmad ldquoPerformance ofrandom diagonal code for OCDMA systems using new spectraldirect detection techniquerdquoOptical Fiber Technology vol 15 no3 pp 283ndash289 2009

[9] H Y Ahmed and K S Nisar ldquoDiagonal Eigenvalue Unity(DEU) code for spectral amplitude coding-optical code divisionmultiple accessrdquoOptical Fiber Technology vol 19 no 4 pp 335ndash347 2013

[10] M H Kakaee S Seyedzadeh H Adnan Fadhil S BarirahAhmad Anas andMMokhtar ldquoDevelopment of Multi-Service(MS) for SAC-OCDMA systemsrdquo Optics and Laser Technologyvol 60 pp 49ndash55 2014

[11] L Wosinska J Chen and C P Larsen ldquoFiber access networksReliability analysis and Swedish broadband marketrdquo IEICE

Transactions on Communications vol E92-B no 10 pp 3006ndash3014 2009

[12] K Grobe M Roppelt A Autenrieth J-P Elbers and M EiseltldquoCost and energy consumption analysis of advanced WDM-PONsrdquo IEEE CommunicationsMagazine vol 49 no 2 pp S25ndashS32 2011

[13] W A Imtiaz Y Khan and K Mahmood ldquoDesign and analysisof self-healing dual-ring spectral amplitude coding optical codedivisionmultiple access systemrdquoArabian Journal for Science andEngineering vol 41 no 9 pp 3441ndash3449 2016

[14] A Waqas Design and feasibility of optical code division multipleaccess system for fiber-to-the-home networks [PhD Disserta-tion] 2016

[15] W A Imtiaz M Ilyas and Y Khan ldquoPerformance optimizationof spectral amplitude coding OCDMA system using newenhancedmulti diagonal coderdquo Infrared Physics andTechnologyvol 79 pp 36ndash44 2016

[16] M Kavehrad and D Zaccarin ldquoOptical Code-Division-Multiplexed Systems Based on Spectral Encoding of Noncoher-ent Sourcesrdquo Journal of Lightwave Technology vol 13 no 3 pp534ndash545 1995

[17] S Ayotte and L A Rusch ldquoIncreasing the capacity of SAC-OCDMA Forward error correction or coherent sourcesrdquo IEEEJournal on Selected Topics in Quantum Electronics vol 13 no 5pp 1422ndash1428 2007

[18] K-I Kitayama X Wang and N Wada ldquoOCDMA over WDMPON - Solution path to gigabit-symmetric FTTHrdquo Journal ofLightwave Technology vol 24 no 4 pp 1654ndash1662 2006

[19] S S Pawar and R K Shevgaonkar ldquoModelling of FBG forencodingdecoding in SAC-OCDMA systemrdquo in Proceedings ofthe International Conference on Next Generation Networks pp1ndash4 ind September 2010

[20] B T Lee M S Lee and H Y Song ldquoSimple ring-type passiveoptical network with two-fiber protection scheme and perfor-mance analysisrdquo Optical Engineering vol 46 no 6 Article ID065002 2007

[21] P Lafata and J Vodrazka ldquoSimulation of ring-based passiveoptical network and its experimental verificationrdquo Elektronikair Elektrotechnika vol 19 no 5 pp 93ndash98 2013

[22] C-H Yeh and S Chi ldquoSelf-healing ring-based time-sharingpassive optical networksrdquo IEEE Photonics Technology Lettersvol 19 no 15 pp 1139ndash1141 2007

[23] S Mostafa A E-N AMohamed F E Abd El-Samie and A NZ Rashed ldquoPerformanceAnalysis ofDiagonal EigenvalueUnity(DEU) Code Using NAND Subtraction and Spectral DirectDetection Techniques and Its Use with Triple-Play-Services inSAC-OCDMArdquoWireless Personal Communications vol 85 no4 pp 1831ndash1849 2015

[24] X Zhao X Chen and X Fu ldquoA novel protection switchingscheme for pons with ring plus tree topologyrdquo in Asia-PacificOptical Communications pp 60223Hndash60223H InternationalSociety for Optics and Photonics 2005

[25] MMahloo ldquoReliability versus Cost in Next Generation OpticalAccess Networks [PhD thesis]rdquo Tech Rep KTH Royal Insti-tute of Technology 2013

[26] J Chen and L Wosinska ldquoAnalysis of protection schemes inPON compatible with smooth migration from TDM-PON tohybrid WDMTDM-PONrdquo Journal of Optical Networking vol6 no 5 pp 514ndash526 2007

RoboticsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014


Active and Passive Electronic Components

Control Scienceand Engineering

Journal of


International Journal of

RotatingMachinery


Hindawi Publishing Corporation httpwwwhindawicom

Journal of

Volume 201

Submit your manuscripts athttpswwwhindawicom

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



FF

PIN

PIN

PIN

BBS

BBS

BBS

DFsDecoder1Encoder1

Decoder2Encoder2

Decoder3Encoder3

1N

OCP

N1

OCP

Figure 1 Conventional SAC-OCDMA system

the use of suitable binary sequence is still an open issue[3 5 7ndash10]

Another problem that limits the deployment of SAC-OCDMA system at access domain is the lack of a suitableprotection mechanism [13 14] Conventional SAC-OCDMAsystems are built with tree architecture that utilize a longfeeder fiber (FF) to connect optical line terminal (OLT) atCO with remote node (RN) at the subscriber premises asshown in Figure 1 End-face of the passive RN connectsmultiple optical network units (ONUs) through short spandedicated distribution fibers (DFs) Figure 1 shows the basicarchitecture of SAC-OCDMA system where no protection isprovided between CO andONUs [3 4]This significantly ele-vates the loss of data in case of failures at optical components(including fibers) Therefore it is imperative to design anefficient protection mechanism that facilitates the detectionand recovery of cutsfailures throughout the network inorder to ensure the success of SAC-OCDMA technology atthe access domain

This paper proposes an efficient SAC-OCDMA sys-tem with high transmission capacity through enhancedmultidiagonal (EMD) code Moreover a robust tree-basedhybrid protection architecture is proposed for survivabilityenhancement along with self-healing capacity at both feederand distribution fibers The proposed solution includes anew star-ring protection architecture that provides pre-planned redundancy without duplicating the entire fibersetup at distribution network Performance of the proposedtree-based hybrid system is analyzed using comprehen-sive analytical and simulation analysis by referring to biterror rate (BER) and eye patterns Redundancy operationof the proposed system is also observed for failures atboth feeder and distribution fibers Furthermore capitalexpenditure (CAPEX) and operation expenditure (OPEX)of the proposed setup are also computed to determine thefeasibility of implementation in comparison with existingsolutions

Rest of the paper is organized as follows Section 2 dis-cusses the construction of EMD code followed by a detaileddesign and implementation of the proposed architecture inSection 3 Analytic and simulation analysis for transmissioncapacity of the proposed setup is performed in Section 4Section 5 covers the survivability analysis of the proposedtree-based SAC-OCDMA system for failures at the feederand distribution levels Detailed cost analysis is presented inSection 6 that is followed by the final conclusion in Section 7

2 Enhanced Multidiagonal Code

EMD code is used in the proposed system to achieve therequired transmission capacity in terms of data reach andnumber of subscribers EMD code is designed with two-matrix structure including data matrix (119863119899) and code matrix(119862119899) Chips in 119863119899 are used to encode the data bits afterspectral translation whereas 119862119899 chips refer to the code bitsthat are embedded to elevate the performance of EMD basedSAC-OCDMA system [15] Moreover EMD coding algo-rithm is designed with three basic performance parametersincluding119873119908 and120582119888 between adjacent codesThe followingsection outlines the properties of119863119899 and 119862119899 followed by theimplementation of EMDmatrix

EMD = [119863119899 | 119862119888]119870times119873 (1)

21 Data Matrix 119863119899 consists of a 119870 times 119870 diagonal matrixwhere119870 rows and119870 columns of thematrix represent numberof subscribers and length of 119863119899 respectively Furthermore119863119899 is designed with multiple rows of length 119870 119908 = 1 and119888 = 0 Therefore the basic 2 times 2 119863119899 can be written as

119863119899 = [1 00 1] (2)

22 Code Matrix 119862119899 consists of a 119870 times 119869 matrix where 119870rows and 119869 columns represent the number of subscribersand length of the matrix respectively 119862119899 for EMD code isdesigned with multiple rows of 119869 = 119870 + 1 119908 = 2 and 119888 = 1between the codes in adjacent rowsThe basic 2times3 119862119899 can bewritten as

119862119899 = [0 1 11 1 0] (3)

Consequently the proposed EMD code while using (1)and the properties of119863119899 119862119899 is given by

EMD2 = [1 0 | 0 1 10 1 | 1 1 0]2times5

(4)

Equation (4) shows that 10158401 2 11015840 combination is main-tained between adjacent EMD codes The number of sub-scribers in EMD code can be increased through simplediagonal mapping while maintaining the respective prop-erties of 119863119899 and 119862119899 This increases the cardinality of SAC-OCDMA systemwithminimum impact on119873 for EMD codeConsequently EMD code matrix for 4 subscribers is given by

EMD4 =[[[[[[

1 0 0 0 | 0 0 0 1 10 1 0 0 | 0 0 1 1 00 0 1 0 | 0 1 1 0 00 0 0 1 | 1 1 0 0 0

]]]]]]4times9

(5)


OCP

OCP

PINLPFDM

PINLPFDM

OCP

MZM

MZM

MZM

MZM

PIN LPF DM

LED

Data

MZM

1

OS

DFsPF

ONU

os3

21

2

2

Encoder

Encoder

Encoder

Encoder

Encoder

EDFA

OCP

1

BPF

BPF

BPF

ONU1

LEDM

LED1

ONU2

ONU3

ONU4

ONU5

ONU6

ONUMminus1

ONUM

OLTTx

OLTRx

FFp

FFs

2N

OCP






119873sum119894=1

119861119896 (119894) 119861119897 (119894) =


(6)

























MZM

Data

FFPINEncoder Filter

1N

OCP

N1

OCP






SNR = [ 11986821205902] (7)


1205902 = 2119890119861119868 + 1198682119861120591119888 + 4119870119887119879119899119861119877119871 (8)


120591119888 = intinfin119900119866 (120592)2 119889120592

[intinfin119900119866 (120592) 119889120592]2 (9)



119873sum119894=1

119861119896 (119894) 119861119897 (119894) = 119908 forall119896 = 1198970 forall119896 = 119897 (10)


119868 = Rintinfin119900

119866 (120592) 119889120592 (11)


119866 (120592) = 119875srΔ120592119870sum119896=1

119889119896119873sum119894=1

119861119896 (119894) 119861119897 (119894) [119906 (Δ120592119873 )] (12)




intinfin0

119866 (120592) 119889120592

= intinfin0

119875srΔ120592119870sum119896=1

119889119896119873sum119894=1

119861119896 (119894) 119861119897 (119894) 119906 [Δ120592119873 ]119889120592(13)


intinfin0

119866 (120592) 119889120592 = 119875sr119908119873 (14)


119868 = R119875sr119908119873 (15)


⟨1198942shot⟩ = 2119890119861R119875sr119908119873 (16)



SNR = [R119875sr119908119873]2119890119861R119875sr119908119873 + 4119870119887119879119899119861119877119871 (17)



SNRBER

0

80

160

240

320

400

480

560

640

720

800

Sign

al to

noi

se ra

tio (S

NR)

80 8856 64 7240 48 9632 112 120 128104


1E minus 33

1E minus 30

1E minus 27

1E minus 24

1E minus 21

1E minus 18

1E minus 15

1E minus 12

1E minus 9

1E minus 6

1E minus 3

Bit e

rror

rate

(BER

)


36 40 44 48 52 56 60 6432


1E minus 40

1E minus 36

1E minus 32

1E minus 28

1E minus 24

1E minus 20

1E minus 16

1E minus 12

1E minus 8

1E minus 4

Bit e

rror

rate

(BER

)

15 Gbps2Gbps25 Gbps

3Gbps4Gbps5Gbps







125 200 225100 150 250 275175

Data rate (Gbps)

1E minus 34

1E minus 32

1E minus 30

1E minus 28

1E minus 26

1E minus 24

1E minus 22

1E minus 20

1E minus 18

1E minus 16

1E minus 14

1E minus 12

1E minus 10

1E minus 8

Bit e

rror

rate

(BER

)














(a) (b) (c)


OS

1

2MAC

f

OLTTx

OLTRx

FFp

FFs











2

1

MAC

c

c

OCP

DFs PF

OS

ONU1 ONURx

ONUTx

ONU2

ONU2

ONU3

ONU4

ONUM

2N

OCP









(18)





119875DS ge 119877sen (19)






6 Cost Analysis















(20)




2000

2200

2400

2600

2800

3000

3200

3400

3600

3800

4000

Capi

tal e

xpen

ditu

re ($

)

(a)



0

50

100

150

200

250

300

350

400

450

Ope

ratio

nal e

xpen

ditu

re ($

)(b)








CAPEXOPEXTotal cost


0

500

1000

1500

2000

2500

3000

3500

4000

4500

Tota

l cos

t ($)



7 Conclusion






References




























RoboticsJournal of





Journal of



RotatingMachinery



Journal of

Volume 201


VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation










OCP

OCP

PINLPFDM

PINLPFDM

OCP

MZM

MZM

MZM

MZM

PIN LPF DM

LED

Data

MZM

1

OS

DFsPF

ONU

os3

21

2

2

Encoder

Encoder

Encoder

Encoder

Encoder

EDFA

OCP

1

BPF

BPF

BPF

ONU1

LEDM

LED1

ONU2

ONU3

ONU4

ONU5

ONU6

ONUMminus1

ONUM

OLTTx

OLTRx

FFp

FFs

2N

OCP






119873sum119894=1

119861119896 (119894) 119861119897 (119894) =


(6)

























MZM

Data

FFPINEncoder Filter

1N

OCP

N1

OCP






SNR = [ 11986821205902] (7)


1205902 = 2119890119861119868 + 1198682119861120591119888 + 4119870119887119879119899119861119877119871 (8)


120591119888 = intinfin119900119866 (120592)2 119889120592

[intinfin119900119866 (120592) 119889120592]2 (9)



119873sum119894=1

119861119896 (119894) 119861119897 (119894) = 119908 forall119896 = 1198970 forall119896 = 119897 (10)


119868 = Rintinfin119900

119866 (120592) 119889120592 (11)


119866 (120592) = 119875srΔ120592119870sum119896=1

119889119896119873sum119894=1

119861119896 (119894) 119861119897 (119894) [119906 (Δ120592119873 )] (12)




intinfin0

119866 (120592) 119889120592

= intinfin0

119875srΔ120592119870sum119896=1

119889119896119873sum119894=1

119861119896 (119894) 119861119897 (119894) 119906 [Δ120592119873 ]119889120592(13)


intinfin0

119866 (120592) 119889120592 = 119875sr119908119873 (14)


119868 = R119875sr119908119873 (15)


⟨1198942shot⟩ = 2119890119861R119875sr119908119873 (16)



SNR = [R119875sr119908119873]2119890119861R119875sr119908119873 + 4119870119887119879119899119861119877119871 (17)



SNRBER

0

80

160

240

320

400

480

560

640

720

800

Sign

al to

noi

se ra

tio (S

NR)

80 8856 64 7240 48 9632 112 120 128104


1E minus 33

1E minus 30

1E minus 27

1E minus 24

1E minus 21

1E minus 18

1E minus 15

1E minus 12

1E minus 9

1E minus 6

1E minus 3

Bit e

rror

rate

(BER

)


36 40 44 48 52 56 60 6432


1E minus 40

1E minus 36

1E minus 32

1E minus 28

1E minus 24

1E minus 20

1E minus 16

1E minus 12

1E minus 8

1E minus 4

Bit e

rror

rate

(BER

)

15 Gbps2Gbps25 Gbps

3Gbps4Gbps5Gbps







125 200 225100 150 250 275175

Data rate (Gbps)

1E minus 34

1E minus 32

1E minus 30

1E minus 28

1E minus 26

1E minus 24

1E minus 22

1E minus 20

1E minus 18

1E minus 16

1E minus 14

1E minus 12

1E minus 10

1E minus 8

Bit e

rror

rate

(BER

)














(a) (b) (c)


OS

1

2MAC

f

OLTTx

OLTRx

FFp

FFs











2

1

MAC

c

c

OCP

DFs PF

OS

ONU1 ONURx

ONUTx

ONU2

ONU2

ONU3

ONU4

ONUM

2N

OCP









(18)





119875DS ge 119877sen (19)






6 Cost Analysis















(20)




2000

2200

2400

2600

2800

3000

3200

3400

3600

3800

4000

Capi

tal e

xpen

ditu

re ($

)

(a)



0

50

100

150

200

250

300

350

400

450

Ope

ratio

nal e

xpen

ditu

re ($

)(b)








CAPEXOPEXTotal cost


0

500

1000

1500

2000

2500

3000

3500

4000

4500

Tota

l cos

t ($)



7 Conclusion






References




























RoboticsJournal of





Journal of



RotatingMachinery



Journal of

Volume 201


VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation
























MZM

Data

FFPINEncoder Filter

1N

OCP

N1

OCP






SNR = [ 11986821205902] (7)


1205902 = 2119890119861119868 + 1198682119861120591119888 + 4119870119887119879119899119861119877119871 (8)


120591119888 = intinfin119900119866 (120592)2 119889120592

[intinfin119900119866 (120592) 119889120592]2 (9)



119873sum119894=1

119861119896 (119894) 119861119897 (119894) = 119908 forall119896 = 1198970 forall119896 = 119897 (10)


119868 = Rintinfin119900

119866 (120592) 119889120592 (11)


119866 (120592) = 119875srΔ120592119870sum119896=1

119889119896119873sum119894=1

119861119896 (119894) 119861119897 (119894) [119906 (Δ120592119873 )] (12)




intinfin0

119866 (120592) 119889120592

= intinfin0

119875srΔ120592119870sum119896=1

119889119896119873sum119894=1

119861119896 (119894) 119861119897 (119894) 119906 [Δ120592119873 ]119889120592(13)


intinfin0

119866 (120592) 119889120592 = 119875sr119908119873 (14)


119868 = R119875sr119908119873 (15)


⟨1198942shot⟩ = 2119890119861R119875sr119908119873 (16)



SNR = [R119875sr119908119873]2119890119861R119875sr119908119873 + 4119870119887119879119899119861119877119871 (17)



SNRBER

0

80

160

240

320

400

480

560

640

720

800

Sign

al to

noi

se ra

tio (S

NR)

80 8856 64 7240 48 9632 112 120 128104


1E minus 33

1E minus 30

1E minus 27

1E minus 24

1E minus 21

1E minus 18

1E minus 15

1E minus 12

1E minus 9

1E minus 6

1E minus 3

Bit e

rror

rate

(BER

)


36 40 44 48 52 56 60 6432


1E minus 40

1E minus 36

1E minus 32

1E minus 28

1E minus 24

1E minus 20

1E minus 16

1E minus 12

1E minus 8

1E minus 4

Bit e

rror

rate

(BER

)

15 Gbps2Gbps25 Gbps

3Gbps4Gbps5Gbps







125 200 225100 150 250 275175

Data rate (Gbps)

1E minus 34

1E minus 32

1E minus 30

1E minus 28

1E minus 26

1E minus 24

1E minus 22

1E minus 20

1E minus 18

1E minus 16

1E minus 14

1E minus 12

1E minus 10

1E minus 8

Bit e

rror

rate

(BER

)














(a) (b) (c)


OS

1

2MAC

f

OLTTx

OLTRx

FFp

FFs











2

1

MAC

c

c

OCP

DFs PF

OS

ONU1 ONURx

ONUTx

ONU2

ONU2

ONU3

ONU4

ONUM

2N

OCP









(18)





119875DS ge 119877sen (19)






6 Cost Analysis















(20)




2000

2200

2400

2600

2800

3000

3200

3400

3600

3800

4000

Capi

tal e

xpen

ditu

re ($

)

(a)



0

50

100

150

200

250

300

350

400

450

Ope

ratio

nal e

xpen

ditu

re ($

)(b)








CAPEXOPEXTotal cost


0

500

1000

1500

2000

2500

3000

3500

4000

4500

Tota

l cos

t ($)



7 Conclusion






References




























RoboticsJournal of





Journal of



RotatingMachinery



Journal of

Volume 201


VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation











intinfin0

119866 (120592) 119889120592

= intinfin0

119875srΔ120592119870sum119896=1

119889119896119873sum119894=1

119861119896 (119894) 119861119897 (119894) 119906 [Δ120592119873 ]119889120592(13)


intinfin0

119866 (120592) 119889120592 = 119875sr119908119873 (14)


119868 = R119875sr119908119873 (15)


⟨1198942shot⟩ = 2119890119861R119875sr119908119873 (16)



SNR = [R119875sr119908119873]2119890119861R119875sr119908119873 + 4119870119887119879119899119861119877119871 (17)



SNRBER

0

80

160

240

320

400

480

560

640

720

800

Sign

al to

noi

se ra

tio (S

NR)

80 8856 64 7240 48 9632 112 120 128104


1E minus 33

1E minus 30

1E minus 27

1E minus 24

1E minus 21

1E minus 18

1E minus 15

1E minus 12

1E minus 9

1E minus 6

1E minus 3

Bit e

rror

rate

(BER

)


36 40 44 48 52 56 60 6432


1E minus 40

1E minus 36

1E minus 32

1E minus 28

1E minus 24

1E minus 20

1E minus 16

1E minus 12

1E minus 8

1E minus 4

Bit e

rror

rate

(BER

)

15 Gbps2Gbps25 Gbps

3Gbps4Gbps5Gbps







125 200 225100 150 250 275175

Data rate (Gbps)

1E minus 34

1E minus 32

1E minus 30

1E minus 28

1E minus 26

1E minus 24

1E minus 22

1E minus 20

1E minus 18

1E minus 16

1E minus 14

1E minus 12

1E minus 10

1E minus 8

Bit e

rror

rate

(BER

)














(a) (b) (c)


OS

1

2MAC

f

OLTTx

OLTRx

FFp

FFs











2

1

MAC

c

c

OCP

DFs PF

OS

ONU1 ONURx

ONUTx

ONU2

ONU2

ONU3

ONU4

ONUM

2N

OCP









(18)





119875DS ge 119877sen (19)






6 Cost Analysis















(20)




2000

2200

2400

2600

2800

3000

3200

3400

3600

3800

4000

Capi

tal e

xpen

ditu

re ($

)

(a)



0

50

100

150

200

250

300

350

400

450

Ope

ratio

nal e

xpen

ditu

re ($

)(b)








CAPEXOPEXTotal cost


0

500

1000

1500

2000

2500

3000

3500

4000

4500

Tota

l cos

t ($)



7 Conclusion






References




























RoboticsJournal of





Journal of



RotatingMachinery



Journal of

Volume 201


VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation













125 200 225100 150 250 275175

Data rate (Gbps)

1E minus 34

1E minus 32

1E minus 30

1E minus 28

1E minus 26

1E minus 24

1E minus 22

1E minus 20

1E minus 18

1E minus 16

1E minus 14

1E minus 12

1E minus 10

1E minus 8

Bit e

rror

rate

(BER

)














(a) (b) (c)


OS

1

2MAC

f

OLTTx

OLTRx

FFp

FFs











2

1

MAC

c

c

OCP

DFs PF

OS

ONU1 ONURx

ONUTx

ONU2

ONU2

ONU3

ONU4

ONUM

2N

OCP









(18)





119875DS ge 119877sen (19)






6 Cost Analysis















(20)




2000

2200

2400

2600

2800

3000

3200

3400

3600

3800

4000

Capi

tal e

xpen

ditu

re ($

)

(a)



0

50

100

150

200

250

300

350

400

450

Ope

ratio

nal e

xpen

ditu

re ($

)(b)








CAPEXOPEXTotal cost


0

500

1000

1500

2000

2500

3000

3500

4000

4500

Tota

l cos

t ($)



7 Conclusion






References




























RoboticsJournal of





Journal of



RotatingMachinery



Journal of

Volume 201


VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation










(a) (b) (c)


OS

1

2MAC

f

OLTTx

OLTRx

FFp

FFs











2

1

MAC

c

c

OCP

DFs PF

OS

ONU1 ONURx

ONUTx

ONU2

ONU2

ONU3

ONU4

ONUM

2N

OCP









(18)





119875DS ge 119877sen (19)






6 Cost Analysis















(20)




2000

2200

2400

2600

2800

3000

3200

3400

3600

3800

4000

Capi

tal e

xpen

ditu

re ($

)

(a)



0

50

100

150

200

250

300

350

400

450

Ope

ratio

nal e

xpen

ditu

re ($

)(b)








CAPEXOPEXTotal cost


0

500

1000

1500

2000

2500

3000

3500

4000

4500

Tota

l cos

t ($)



7 Conclusion






References




























RoboticsJournal of





Journal of



RotatingMachinery



Journal of

Volume 201


VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation










2

1

MAC

c

c

OCP

DFs PF

OS

ONU1 ONURx

ONUTx

ONU2

ONU2

ONU3

ONU4

ONUM

2N

OCP









(18)





119875DS ge 119877sen (19)






6 Cost Analysis















(20)




2000

2200

2400

2600

2800

3000

3200

3400

3600

3800

4000

Capi

tal e

xpen

ditu

re ($

)

(a)



0

50

100

150

200

250

300

350

400

450

Ope

ratio

nal e

xpen

ditu

re ($

)(b)








CAPEXOPEXTotal cost


0

500

1000

1500

2000

2500

3000

3500

4000

4500

Tota

l cos

t ($)



7 Conclusion






References




























RoboticsJournal of





Journal of



RotatingMachinery



Journal of

Volume 201


VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation












6 Cost Analysis















(20)




2000

2200

2400

2600

2800

3000

3200

3400

3600

3800

4000

Capi

tal e

xpen

ditu

re ($

)

(a)



0

50

100

150

200

250

300

350

400

450

Ope

ratio

nal e

xpen

ditu

re ($

)(b)








CAPEXOPEXTotal cost


0

500

1000

1500

2000

2500

3000

3500

4000

4500

Tota

l cos

t ($)



7 Conclusion






References




























RoboticsJournal of





Journal of



RotatingMachinery



Journal of

Volume 201


VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation











2000

2200

2400

2600

2800

3000

3200

3400

3600

3800

4000

Capi

tal e

xpen

ditu

re ($

)

(a)



0

50

100

150

200

250

300

350

400

450

Ope

ratio

nal e

xpen

ditu

re ($

)(b)








CAPEXOPEXTotal cost


0

500

1000

1500

2000

2500

3000

3500

4000

4500

Tota

l cos

t ($)



7 Conclusion






References




























RoboticsJournal of





Journal of



RotatingMachinery



Journal of

Volume 201


VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation













References




























RoboticsJournal of





Journal of



RotatingMachinery



Journal of

Volume 201


VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation









RoboticsJournal of





Journal of



RotatingMachinery



Journal of

Volume 201


VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation









design and analysis of self-healing tree-based hybrid...

Documents