ICP2010-98

Simulation of 1.25 Gb/s Downstream Transmission Performance of GPON-FTTx

Hesham A. Bakarman, Sahbudin Shaari, Member, IEEE, and Mahamod Ismail, Member, IEEE
data from a single optical line termination (OLT) toward multiple optical network units (ONUs) or optical network terminations (ONTs). A burst-mode link in the 1310-nm window collects all ONU/ONT upstream traffic toward the OLT as variable-length packets at a 1.25-Gb/s aggregate rate, in a P2MP time-division multiple-access (TDMA) scheme. This paper focuses on the downlink part only. It presents transmission performance of downstream link GPON network with 1.25 Gb/s bit rate.

Abstract In this paper, 1.25 Gb/s GPON downstream link is presented. All the optical distribution network (ODN) classes are implemented, using Optisystem, to investigate the transmission capability and performance of the proposed downstream physical media (PM) GPON model. Some of the design constraints involved in an optical network design such as fiber span analysis, power budget and margin calculations are taken into consideration with worst case. The quality or performance of a digital communication system is specified by its BER or Q value with respect to other parameters such as receiver sensitivity. The simulated model can support 18, 50 and 128 number of users for classes A, B, and C respectively. Index Terms Bit error rate, fiber to the home, GPON, passive optical network

I. INTRODUCTION passive optical network (PON) technology is based on Tpassive star fiber network and offers a cost effective
HE

optical access solution with point-to-multipoint (P2MP) nature. With rapidly growing customer bandwidth requirements and proliferation of bandwidth in metro networks, broadband passive optical networks (BPONs) [1], [2] and the emerging gigabit-capable passive optical networks (GPONs) are expected to prevail as the leading optical access technology eliminating the bandwidth bottleneck in the last mile. The full-services access networks (FSAN) GPON can provide high bandwidth services to customers following different fiber-to-the premises/ cabinet/building/home/user (FTTx) scenarios [3]. Recently, the FSAN initiated GPON network standardization via recommendations for the GPON physicalmedia-dependent (PMD) layer and the transmission convergence (TC) layer [4], [5]. Figure 1 illustrates a symmetric 1.25 Gb/s GPON access system. A continuous downlink in the wavelength band of 1480 1500 nm carries 1.25 Gb/s time-division-multiplexed (TDM)
Hesham A. Bakarman is with the Photonics Technology Laboratory (PTL), Institute of Micro Engineering and Nanoelectronics (IMEN) Universiti Kebangsaan Malaysia, Bangi, 43600 UKM Bangi Malaysia (phone: 03-87360705; e-mail: hesham@ vlsi.eng.ukm.my). Sahbudin Shaari , was with Photonics Technology Laboratory (PTL), Institute of Micro Engineering and Nanoelectronics (IMEN) Universiti Kebangsaan Malaysia, Bangi, 43600 UKM Bangi Malaysia (e-mail: sahbudin@vlsi.eng.ukm.my). Mahamod Ismail is with the Electrical, Electronics and Systems Engineering Department, Universiti Kebangsaan Malaysia, Bangi, 43600 UKM Bangi Malaysia (e-mail: mahamod@vlsi.eng.ukm.my).

Fig. 1. GPON network architecture for FTTx scenarios.

II. GPON ACCESS NETWORK Due to the limitation of the ADSL (asymmetric digital subscriber line) service, which suffers from limited transmission speed and distance, because it uses conventional metallic cables, optical access is expected to become the default broadband access system in the future. For this reason, ITU-T (International Telecommunication Union Telecommunication Standardization Sector) has discussed a standard for optical access systems called G-PON (Gigabit passive optical network), which is an optical access system with gigabit per second-class transmission capability; it is suitable as the next-generation optical access system. A. Previous Optical Access System Standards ITU-T has created several standards for optical access systems. One of the most important is the BPON (Broadband PON) standard. PON is a network topology that shares a

978-1-4244-7187-4/10/$26.00 2010 IEEE

ICP2010-98
single optical fiber among two or more customers. Figure 2 shows its basic structure. The main feature is that network equipment, called OLT placed in a central office, is connected to the optical network terminal equipment, called ONU installed in a customers premises, via an optical splitter. Since several customers share the optical fiber and OLT, PON can offer economical services by reducing subscriber (or customer) cost. For these reasons, a PON system is considered to be eminently suitable for the future optical access system.

Fig. 3. GPON generic physical configuration of the optical distribution network

Fig. 2. Basic composition of PON

Fig. 4. Reference physical medium model

B-PON was developed as a PON system that uses ATM cells for transmission and has a maximum access speed of 155 Mbit/s upstream and 622 Mbit/s downstream. By using ATM cells, B-PON can accommodate various services, such as Internet or CATV services. B. Network Architecture of the GPON The PON access technology is a passive tree network in which one OLT serves up to 128 customers [6]. Figure 3 depicts the reference points and the optical interfaces of the generic physical configuration of the ODN (G.983.2). The two directions for optical transmission in the ODN (Optical Distribution Network) are identified for the symmetric GPON as follows: (1) Downstream direction for signals traveling from the OLT to the ONU(s): Wavelength: 1480-1500 nm (basic band) Physical link rate: 1.24416 Gbit/s, TDM (2) Upstream direction for signals traveling from the ONU(s) to the OLT: Wavelength: 1260-1360 nm bands Physical link rate: 1.24416 Gbit/s, TDMA C. Physical Media Building Blocks Figure 4 show the purely optical layer includes the optical fiber, splitters, WDM multiplexers/demultiplexers, connectors, attenuators, optical filters and optical amplifiers (not used in this simulation).

Just above the purely optical layer there is a layer for electrical-to-optical and optical-to-electrical conversion; the electrical-to-optical conversion function is performed by a semiconductor laser diode, turning an electrical current signal into an optical power signal. At the other side of the link, the optical-to-electrical conversion is performed by an optical receiver comprising a semiconductor photodiode and an electrical (pre) amplifier. A further layer is then added above the analogue electrical layer for the conversion from/to the electrical digital layer. Digital-to-analogue conversion is performed by the laser driver (including an electrical filter) in one direction and by the decision stage in the opposite direction. The digital layer is very useful for the link performance evaluation since it allows the BER evaluation. This model can be used for every fiber optics digital transmission system [3]. D. 1.25 Gb/s Downstream PMD Layer Specifications The optical parameters defined in [5] refer to values measured at S/R and R/S points, as shown in Figure 3. In this paper, all the ODN classes are considered, therefore Class A, Class B, and Class C parameters are reported. III. NETWORK DESIGN AND MODELING A network planner needs to optimize the various electrical and optical parameters to ensure smooth operations of an optical network. Whether the network topology is that of a point-to-point link, a ring, or a mesh, system design inherently can be considered to be of two separate parts: optical system design and electrical or higher-layer system design. This section explores some of the design constraints involved in an

ICP2010-98
optical network design such as power budget and margin calculations. To ensure that the fiber system has sufficient power for correct operation, network designer needs to calculate the spans power budget, which is the maximum amount of power it can transmit [7]. From a design perspective, worst-case analysis calls for assuming minimum transmitter power and minimum receiver sensitivity. This provides for a margin that compensates for variations of transmitter power and receiver sensitivity levels
Power budget ( P ) = Minimum transmitte r power ( PT

Simulation schematic in Optisystem software. In the left side, where the OLT transmitter part is located, an external modulated transmitter is used. It consists of continuous wave laser source, Mach-Zehender modulator, NRZ pulse generator and pseudo random sequence generator. The right hand side an ONT receiver should have photodiode, low pass filter and clock and data recovery. The link, sometimes called channel, is consists of 20 km of single mode fiber SMF28 and optical attenuator so as to add the loss for all the ODN classes.

min )

(1)

Minimum receiver sensitivit y ( PR min )

The span losses can be calculated by adding the various linear and nonlinear losses. Factors that can cause span or link loss include fiber attenuation, splice attenuation, connector attenuation, chromatic dispersion, and other linear and nonlinear losses
Span loss ( P ) = ( Fiber attenuatio n * Km ) + ( Splice attenuatio n * number of

splices ) + ( Connector attenuatio n * number of connectors ) + ( In line device losses ) + ( Nonlinear losses ) + ( Safety m arg in )

(2)

Fig. 5. GPON downstream simulation schematic in Optisystem

The next calculation involves the power margin ( Pm ), which represents the amount of power available after subtracting linear and nonlinear span losses ( Ps ) from the power budget ( Pb ). A Pm greater than zero indicates that the power budget is sufficient to operate the receiver. The formula for power margin ( Pm ) is as follows:
Power m arg in ( Pm ) = Power budget ( Pb ) Span loss ( Ps )

(3)

To prevent receiver saturation, the input power received by the receiver, after the signal has undergone span loss, must not exceed the maximum receiver sensitivity specification ( PR max ). This signal level is denoted as ( P Iin ). The

maximum transmitter power ( PR max ) must be considered as the launch power for this calculation. The span loss ( Ps ) remains constant.
Input power ( P

B. Bit Error Rate The worst-case probability of error (BER) as function of the received input optical power Pr is shown in figure 6. The curve given in this figure is usually referred as performance curve or BER curve of a conventional optical transmission system with high intensity modulation and direct detection (IM/DD). Due to intensity modulation, this system is completely insensitive to laser phase noise. The dominant sources of noise are thermal noise and not shot noise, which considered being approximately Gaussian distributed [8]. The quality or performance of a digital communication system is specified by its BER or Q value. The BER is specified as the average probability of incorrect bit identification. In general, the higher the received Q-value, the lower the BER probability will be. Mathematically, equation (5) gives the Q-factor of an optical signal.
Q= I1 I 0 I1 + I 0

(5)

in

) = Maximum transmitte r powre ( PT

max

) Span loss

(4)

The design equation: Input power ( P Iin ) <= Maximum receiver sensitivity ( PR max ) must be satisfied to prevent receiver saturation and ensure system viability. If the input power ( P Iin ) is greater than the maximum receiver sensitivity ( PR max ), passive attenuation must be considered to reduce signal level and bring it within the dynamic range of the receiver. A. GPON Downstream Simulation Model Figure 5 shows the proposed GPON Downstream

where, I 1 is the value of the 1-bit current, I 0 is the value of the 0-bit current, 1 is the standard deviation of the 1-bit current, and 0 is the standard deviation of the 0-bit current. The relationship of Q-factor to BER is defined as:
BER = 1 Q erfc ( ) 2 2

(6)

If noise is Gaussian, the BER is also expressed as [9]:

ICP2010-98
-30 -29 -28 -27 -26 -25 -24 -23 -22 -21 -20 0.01 1E-06 1E-10 1E-14 1E-18 1E-22 1E-26 1E-30 1E-34 1E-38 1E-42 1E-46 1E-50

BER

every 1012 bits can be corrupted during transmission. Figure 7 shows the minimum BER for class A, B and C with transmitted power of 4, 1 and 5 dBm respectively. Typical requirements for optical receivers used in this simulation are optimized to be BER < 10 10 (less than one error in 1010 bits). The receiver sensitivity is the minimum averaged received optical power required to achieve BER = 10 10 . From this explanation, it becomes evident why optical system design considers power budget and power margins (safety margins for good design) so important.

Reveived input optical power

Fig. 6. BER as a function of the received input optical power for a direct detection system

BER =

Q 1 [1 erfc ( )] 2 2

1 2

Q2 2
Q

(7)

Based on this Q value of 6.1 corresponds to 10 9 , and a Q value of 7.2 to BER = 10 12 .

IV. RESULTS AND DISCUSSION The specified optical levels at the optical interface Old / Ord as shown in figure 3, for each class are listed in Table 1. The values take into account the worst case, for application in a single or dual fiber ODN (with or without coarse WDM).
Fig. 7. Minimum BER obtained from BER analyzer

Figure 8 shows the relation between receiver sensitivity and BER. It is obvious; to achieve the required BER a photodetector should have sensitivity power of -26.5 dBm for class A and B and -27.5 dBm for class C. In this range, even a very small rise in optical signal power improves BER by some order of magnitudes.
0.01 1E-06 1E-10 1E-14 1E-18 BER 1E-22 1E-26 1E-30 1E-34 1E-38 1E-42 1E-46 1E-50 -32 -31 -30 -29 -28 -27 -26 -25 -24 -23 -22 Receiver sensitivity (dBm)
Class A Class B Class C

Fig. 8. BER as a function of attenuation for PIN sensitivity measurement

A. Minimum BER and Receiver Resistivity An optical link is designed by taking into account a figure of merit, which is generally the bit error rate (BER) of the system. The signal entering the decision circuit fluctuates due to the various noise mechanisms. It is the probability of incorrect bit identification by the decision circuit. For most practical optical networks, this requirement of BER is 10 12

Figure 9 shows the BER behavior as a function of the span loss, keeping the fiber length constant 20 Km and the transmitter power constant (minimum values) for each class. BER of 10 10 is obtained with over all span losses of 22.5 dB, 27.5 dB and 32.5 dB for class A, B and C respectively. Figure 9 can be expressed in terms of number of users (ONTs). It is clear from figure 10 that GPON with 1.25 Gbit/s can provide communication service from at least 18 up to 128 ONTs users with BER = 10 10 .

(~ 10 9 to 10 12 ), which means that a maximum one out of

ICP2010-98

0.01 1E-06 1E-10 1E-14 1E-18 BER 1E-22 1E-26 1E-30 1E-34 1E-38 1E-42 1E-46 1E-50 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 Losses (dB)
Class A Class B Class C

customers, GPON is the most cost-effective solution. This paper presented the downstream transmission performance of 1.25 Gb/s GPON bit rate. All the ODN classes are simulated separately. A bit error rate BER 10 10 with Q values between 6 and 7 are obtained, which are convenient to sustain a good communication. Multiple customers who are connected to the PON share the OLT costs. While EPON allows only 16 ONTs per PONs, GPON standard allows the OLT PON card to support up to 128 ONTs. This makes the GPON solution 4 to 8 times more cost effective. In this simulation, number of users (ONTs) of 18, 50 and 128 are obtained for classes A, B and C. REFERENCES
[1] [2] [3] [4]
Class A Class B Class C

Fig. 9. BER as a function of loss for 1244 Mbit/s downstream GPON

0.01 0.0001 1E-06 1E-08 1E-10 1E-12 1E-14 1E-16 1E-18 1E-20 1E-22 1E-24 1E-26 1E-28 1E-30 1E-32 1E-34 1E-36 1E-38 1E-40 1E-42 1E-44 1E-46 1E-48 1E-50 8 16 24 32 40 48 56 64 72 80 88 96 Number of ONTs users

[5]

104 112 120 128 136 144

[6] [7] [8] [9]

Fig. 10. BER as a function of number of ONTs users

V. CONCLUSIONS Due to its unprecedented offered bandwidth, GPON is the ideal technology for large-scale FTTH applications where multiple end-users are requiring an ever-growing bandwidth. Moreover, in areas populated by both business and residential

X. Z. Qiu, J. Vandewege, F. Fredricx, and P. Vetter, Burst Mode Transmission in PON Access Systems, in Proc. 7th Eur. Conf. networks Optical Communication, 2002, pp.127132. P. Vetter et al. Study and Demonstration of Extensions to the Standard FSAN BPON, in Proc. Int. Symp. Services Local Access, 2002, p.119 128. Xing-Zhi Qiu. Development of GPON Upstream Physical-MediaDependent Prototypes, Journal of Lightwave Technology, 2004, Vol. 22. ITU-T Recommendation G.984.1, General characteristics for Gigabitcapable Passive Optical Networks 2003. ITU-T Recommendation G.984.2, Gigabit-Capable Passive Optical Networks (GPON): Physical Media Dependent (PMD) Layer Specification ITU-T Recommendation G.984.3 Transmission Convergence Layer for Gigabit Passive Optical Networks, 2004. F. J. Effenberger and E. Shraga. Status of GPON and B-PON standards, Flexlight-networks, (2004). V. Alwayn, Optical Network Design and Implementation: Cisco Press, (2004). J. H. Franze and V. K. Jain , Optical Communications: Components and Systems, Narosa Publishing House, (2000). T. Antony and A. Gumaste, WDM Network Design, Cisco Press, ch. 4, (2003).

BER