Journal of Engineering Sciences [600728]

242
Journal of Engineering Sciences
Assiut University
Faculty of Engineering
Vol. 4 3
No. 2
March 2015
PP. 242 – 252

* Corresponding author.
E-mail address: tamer_abdelrahman@cic -cairo.com
ALL OPTICAL OVSF CODE GENERATOR REALIZATION
USING OPTICAL FLIP FLOP AND OPTICAL HARDLIMITTERS
Tamer A. Moniem * and Noha EL Mosilhy
Canadian International College CIC, New Cairo, Egypt
Noha_elmosilhy@cic -cairo.com
(Received 7 March 2015; Accepted 22 April 2015)
ABSTRACT
The orthogonal variables spreading factor (OVSF) provides variables data rates to flexible support
many applications with different bandwidth. Th is paper presents an optical OVSF code generator
using all optical gates and a set of all optical flip -flops based on two coupled polarization switches
(PSWs). The optical OVSF code generator helps to use the OVSF c ode technique in optical
communication s. The general objective is to make the OVSF code tree suitable to support a lot of
users in the optical communication links . The overall design and the results are discussed through the
realization and the computing numerical simulation to confirm its operation and feasibility . The
proposed scheme has been theoretically demonstrated for a spreading factor of 4 and 8.
Keyw ords: Optical Flip Flops, Optical Gates, Bragg Grating, OVSF, SOA.
1. Introduction
The Modern Code Division Multiple Access (CDMA) uses the OVSF technology codes
to provide multiuser access. It has a lot of uses in different applications mainly with digital
communications and test pattern generations .
In the OVSF , a single code using one transceiver can support higher and variable data
rates with less complexity than the multi -code orthogonal constant spreading factor
(OCSF ) [1]. The OVSF techniques of code placement and allocation in CDMA guarantee
the orthogonally between all users i n different communication channels.
The emergence of increasingly high speed and the number of user s in optical
communication system s demands the OVSF codes technique in the optical communication
links . The performing of signal processing operations entirely within the optical domain
would exploit the speed and parallelism inherent to optics [2].
The overall optical design of linear feedback shift register (LFSR) of pseudorandom
binary sequence (PRBS) generator of CDMA using the hardlimiter and a set of series all

243
JES, Assiut University, Faculty of Engineering, Vol. 4 3, No. 2, March 2015, pp. 242 – 252
optical D flip flops has been demonstrated by Tanay and Tamer et al. (2013) [3]. Zoirios et
al. (2011) [4] have demonstrated an all optical pseudorandom generators using the
Terahertz optical asymmetric Demultiplexer (TOAD) .
In this paper, the basic operation and design of all-optical OVSF code generator of the modern
CDMA used in Universal Mobile Telecommunication System (UMTS) standard are proposed .
The overall optical design in the optical communication system eliminate s the conversion from
optical to electrical and vice versa and reduce s the latency due to this conversion [3].
An overview of the OVSF codes and its code placement methods are presented in
Section 2. The digital electronic circuit of OVSF code is represente d in Section 3. The
building construction and operation of optical devices such as all optical gates and all
optical flip flops are proposed in Section 4 . Finally, the overall optical design of OVSF
code generator and the results output from this generator are displayed in Section 5.
2. OVSF codes in the modern CDMA

The CDMA system allows multiple users to transmit at the same time in the same
frequency band [3]. The CDMA suffers from the high blocking probability because the
number of orthogonal codes is limited. There are two steps applied to users in a CDMA
system. The first step is transforming every bit into a code sequence, where the code
sequence has a length called the spreading factor (SF). The second step is called the
scrambling process, where the scrambling codes are used to separate the signal from different
sources [5 , 6]. The OVSF in CDMA achieves a flexible support of mixed and va riable data
rates at the same bandwidth , where the OVSF code s are used as the channelization codes [7] .

The OVSF codes can be represented and defined by the code tree shown in Figure 1 . Each
OVSF code can be presented by a channelization code CSF,N, where SF is the spreading factor
in the range [4 -512]=[22-29] and N is the branch identification number of user , 1 < N < (SF-1).
The number of codes at each level is equal to the value of SF. All codes in the same layer are
orthogonal, while codes in dif ferent layers are orthogonal if they do not have an ancestor –
descendant relationship such as C 4,2 and C 2,2. Leaf codes have the minimum data rate, which is
denoted by 1R and t he data rate is doubled whenever we go one level up the tree [8].

Fig. 1. Tree of OVSF codes till spreading factor SF=8

244
Tamer A. Moniem and Noha EL Mosilhy , All optical OVSF code generat or realization using …
There are many techniques used for code placement, where the code placement
techniue is allocation policy for a new user requesting for a code of rate kR. There are two
possible prbabilties that occur at no free codes exists, the first one is reject ing this call (i .e.
Code Blocking) and the second is to relocate the existing codes used for a new user [8].
The OVSF code placement has been investigated as follow s. Chao et al. [8] proposed a
simple algorithm based on single and multi – code s placement and replacement schemes ,
but it possibly incurs many fragmented codes which produce a code blocking problem. The
proposed strategies in Chao [8] and Chen et al . [10] algorithm s are based on three
strategies of Random, Leftmost and Crowded –first. The OVSF code placement is
assigned for IMT -2000 for multi-rate with less complexity by Cheng and Lin [ 9].
Recently, Che n and Lung [6] proposed un-sequence property of linear -code cha ins to
design a new code placement and replacement mechanism by identify ing a linear -code
chains (LCCs) and nonlinear -code trees (NCTs) in the R otated OVSF code tree . The logic
circuit of OVSF code tree generator was designed by Dong [11] and Boris [12].
3. The d igital logic circuit of OVSF codes generator

This section represents the electrical logic circuit of OVSF codes generator. Boris et al .
[12] proposed a logic circuit for OVSF code generator. The hardware of the OVSF codes
generator consists of a set of logic gates and T flip flops. The chip sequence is specified in
binary number set { -1, +1} for digital logic operat ing in the set {0, 1} . The OVSF codes
generator is shown in Figure 2 [12]. This generator consists of 9 bits OVSF code ID
register (n 8 – n0), this ID register represent the identification number of user N, suppos ing
the channelization code C 8,6 is required
N=6=n 3n2n1n0=0110
SF=8

Also, the OVSF generator contain s a set of AND gates, XOR gates, and 8 bits binary
ripple counter as shown in Figure 2. That code generator is capable of producing codes
with a spreading factor over range SF=4 to 512 [12]. The logical expression of OVSF code
can be expressed as:

𝑂𝑉𝑆𝐹 𝐶𝑜𝑑𝑒 =(𝑛0∩𝑏8)⊕(𝑛1∩𝑏7)⊕(𝑛2∩𝑏6)⊕(𝑛3∩𝑏5)⊕(𝑛4∩𝑏4)⊕
(𝑛5∩𝑏3)⊕(𝑛6∩𝑏2)⊕(𝑛7∩𝑏1)⊕(𝑛8∩𝑏0) (1)

⊕ and ∩ indicate s the XOR and AND operation , respectively .
The design counter is counting incrementally from 0 to SF – 1, where b 8 is always the most
significant bit (MSB), and the least significant bit (LSB) is specified by the variable spreading
factor SF. The MSB of ID code number N is always enabled by the LSB of the counter [12]. In
this case, a regular binary ripple counter is required to count between 0 and SF – 1. Based on
the logic circuit of OVSF code generator, the code generated for all channelization codes in the
level of SF=8 for binary counting b 8b7b6 is depicted in Table 1. The logic circuit of binary
ripple counter is depicted in Figure 3 [12]. The counter consists of a series T flip flops of 9 bits.
The LSB counting is controlled by SF register as depicted in Table 2.
The MSB of counter is always n 8, and the LSB is controlled by SF register code. The
stored value in SF register represents the information describing the spreading factor and
the code ID required to generating the code.

245
JES, Assiut University, Faculty of Engineering, Vol. 4 3, No. 2, March 2015, pp. 242 – 252

Fig. 2. The logic circuit of OVSF codes generator [ 12].
Table1.
Generation of channelization code for SF=8

246
Tamer A. Moniem and Noha EL Mosilhy , All optical OVSF code generat or realization using …

Fig. 3 . The logic circuit of 9 bits ripple counter with variable location of the LSB [12]
Table 2.
The contents of SF register for different spreading factors
SF SF Register
S9S8S7S6S5S4S3S2S1S0 Counter Range
1 0000000000 0
2 0000000000 0
4 0000000100 0-4
8 0000001000 0-8
16 0000010000 0-16
32 0000100000 0-32
64 0001000000 0-64
128 0010000000 0-128
256 0100000000 0-256
512 1000000000 0-512
4. All optical hardware components

This section represents the optical components used in the realization of the proposed
optical OVSF generator .

4.1. All optical AND, OR and XOR gates
Figure 4 [3, 13] illustrates an all-optical XOR and AND gates using the Bragg grating
as a hardlimiter . Inputs A and B get combined into a single beam. The transmitted intensity
is defined as O1 output and the reflected value as output O2. We bias the hardlimiter at a
limiting value a = 2. If one of the inputs is 0 and the other 1, t he output at O1 is 0 and at O2
is 1. If both inputs A and B have the value of 1, then 2 is transmitted and 0 reflected. If both
A and B inputs have the value of 0, then 0 is transmitted and a 0 is reflected. Thus, O1
yields the result of an AND operation a nd O2 the result of a digital XOR. The intensity
value 2 output of O1 (AND) must be normalized to yield a digital output [13, 14] .

Fig. 4. All optical XOR and AND gates [3, 13] .

247
JES, Assiut University, Faculty of Engineering, Vol. 4 3, No. 2, March 2015, pp. 242 – 252
Figure 5 [13] illustrate s the use of the Bragg grating as a hardlimiter in the construction
of OR gate. Inputs A and B are first combined into a single beam. The transmitted intensity
is defined as the O1 output and the reflected value as the O2. We bias the hardlimiter at a
limiting value a= 1. If one of the inputs is 0 and t he other 1, the output at O1 is 1 and at O2
is 0. If both inputs A and B have the value of 1, a 1 is transmitted and 1 reflected. Thus, O1
yields the result of an OR operation .

Fig. 5. All Optical OR gate [13].
4.2. All optical JK flip flop
The structure of the all -optical J -K flip -flop memory is depicted in Figure 6. It is
consists of all optical flip flop based on two coupled polarization switches [ 2, 14] and a set
of all optical logic gates based on the hardlimiters that are added to convert it into JK flip
flop [15]. The first PSW1 outputs light that is injected into the second PSW2. Hence, the
light that PSW1 outputs acts as a saturating control signal for SOA2 that can suppress
PSW2 and the light that PSW2 outputs can act as a saturating control signal for the SOA1
to suppress PSW1. The optical pulses are injected into PSW1 via Port I of PBS
(polarization beam splitter) to set the flip -flop in State1. The optical pulses are injected into
PSW2 via Port II of PBS to rese t the flip -flop in State 2 [ 14]. The hardlimiters shown in
Figure 5 are biased at a limiting value a=3 to obtain optical AND gates with three inputs
[15]. The output Q is applied to the AND gate with K terminal and clock pulse CK inputs.
The flip -flop is cleared during a clock pulse if Q was previously 1 . Similarly, output Q’ is
applied to AND gate with J terminal and clock (CK) inputs, and the flip -flop is set with a
clock pulse if Q’ was previously 1. Inputs J and K behave like inputs S and R to set and
clear the flip flop [14]. The T flip flop can be obtained from a J -K type if both inputs are
tied together [16]. The truth t able of J -K flip flops is depicted in Table 3. The logical
expression of the inputs Set (S) and Reset (R) can be expressed as:
𝑆=𝑄𝑛′∩𝐽∩𝐶𝑘 (2)
𝑅=𝑄𝑛∩𝐾∩𝐶𝑘 (3)

Fig. 6. All optical J-K flip flop based on two coupled polarization switches and optical gates.

248
Tamer A. Moniem and Noha EL Mosilhy , All optical OVSF code generat or realization using …
Table 3.
The truth table of JK flip flop
J K Qn+1 Q’n+1
0 0 Qn Q’n
0 1 0 1
1 0 1 0
1 1 Q’n Qn
5. All optical implementation of OVSF code generator
This section introduces the design of overall optical implementation , experimental
results, and the simulation analysis of the OVSF generator. The design of generator is
based on the logic design depicted in Figure 2 and 3 by using the proposed optical
components in Section 4. The performance and accuracy of the design of the optical circuit
is evaluated through experimental and numerical simulation to confirm its feasibility in
terms of the choice of the critical parameters .

Fig. 7. The overall all optical OVSF code generator including the ripple counter
of T flip flops, SF register, and optical gates
Figure 7 shows a section of the circuit diagram of all -optical OVSF code generator, it
consists of a cascaded all -optical JK flip flop s, and a set of optical hardlimiter AND gates, OR
gates, and XOR gate s which is equivalent to modulo 2 adder. The two inputs J and K of each
flip flop are tied together to obtain a T flip fl op, and this common input is set by logic input 1.
SOA’s (semiconductor optical amplifier) are used to compensate the losses of the
cascaded connections between flip flops and optical couplers of 50:50 by adjusting the
SOA current injection from 40 mA to 60 mA for different SOAs [17]. This injected
current is u sed to control on the gain of SOA.
In order to calibrate and validate the model, experimental results were used. The theoretical
parameters were heuristically refined until the best agreement between simulation and
experiment, and the values are presented in Table 4, for a central wavelength of 1550 nm.

249
JES, Assiut University, Faculty of Engineering, Vol. 4 3, No. 2, March 2015, pp. 242 – 252
Table 4.
The theoretical parameters of SOA
Parameter Definition Value
w Cavity width 1.35 µm
d Cavity height 0.21 µm
l Cavity length 670 µm
a Transversal section gain 1.29
 10-16
cm2

Ntr Transparency carrier dens. 2.1
 1018
cm-3

α Attenuation 2100/m
R Facet reflection 0.0001
β Line-width enhancement factor 5
nef Effective refractive index 3.414
αins Total insertion loss 6.3dB
Esat Saturation energy 1.25 pJ
Γ Confinement factor 0.39
The output s of SF register are applied to optical ANDs with the external clock to control the
counting of optical counter according to Table 2 . The optical counter consist s of a series of
optical T flip flops based on two coupled polarization switches which are presented in Section
4.2. The output of counter is optically AND with the identification code N. The outputs of all
optical AND s are applied to XOR gates according to Eq.(1) to finally obtain the OVSF code.
The OVSF code is numerically simulated using Beam Prop method V.7 and Optowave
simulation program V.2 . The wavelengths of each flip -flop are 1550 nm for λ1 and 1552
nm for λ2. There is an external light input at wavelength of 1600 nm (λ3) which is injected
as a clock pulse (CK). The optical clock pulses are in jected approximately at every 0.1 ns
with clock speed of 9 -10 GHz with duration equal to 0.05 ns.
Figure 8 shows the result backed by the simulation results of optical OVSF generator
C8,6 at SF=8, where the identification number is N=6=n 3n2n1n0=0110 and the SF register
code is adjusted at 0000001000 . At SF=8, the counter is counting incrementally from 0 to
8, where b8 is always the MSB, and the LSB is b 6 as shown in Figure 8. The logic
expression of OVSF code at this case can be expressed as:
𝑂𝑉𝑆𝐹 𝐶𝑜𝑑𝑒 =(𝑛0∩𝑏8)⊕(𝑛1∩𝑏7)⊕(𝑛2∩𝑏6) (4)

The output s of OVSF code generator of C8,5 and C 8,7 for N=5 and N=7 respectively are
depicted together in Figure 9.

250
Tamer A. Moniem and Noha EL Mosilhy , All optical OVSF code generat or realization using …

Fig. 8. The numerical simulation output code C8,6 at N=0110 with SF=8 and
counter range b 8 b7 b6 (0-8).

Fig. 9. The numerical simulation output code s C8,5 and C 8,7 at N=0101 and N=0111 respectively .

6. Conclusion

In this paper, we have proposed and described all -optical schemes for OVSF code generator
of spreading factor range SF=4-512. The generator consists of a series of optical T flip flops
based on two coupled polarization switches, and a set of optical logic gates. Numerical
simulation results confirming the described method are given. The theoretical model is
developed and the results obtained experimentally backed by the numerically simulation result
will be useful in the future for using the OVSF code tree techniques in all -optical computing
and optical information processing to supports a lot of users in the optical communication links .

251
JES, Assiut University, Faculty of Engineering, Vol. 4 3, No. 2, March 2015, pp. 242 – 252
REFERENCES
[1] E. Dahlman and K. Jamal, “Wide -Band Services in a DS -CDMA Based FPLMTS System ”,
Proc. IEEE Vehicular Technology Conf. 1996, pp. 1656 -1660, (1996 ).
[2] T.A. Moniem, N.A. Rabou, E.M. Saad, “Parallel -shift register and binary multiplier using
optical hardware components ”, Opt. Eng. Vol 47, doi: 10.1117/1.2898632 , (2008 ).
[3] Tanay Chattopadhyay, T. A. Moniem, Hirak Kumar Maity “All-Optical Pseudorandom Binary Sequence
(PRBS) Generator Using The Hardlimiters ”, OPTIK Elsevier journal 124, pp. 4252– 4256, (2013) .
[4] K.E. Zoiros , M.K. Das, D.K. Gayen, H.K. Maity, T. Chattopadhyay, J.N. Roy, “All-optical pseudorandom
binary sequence generator with TOAD -based D flipflops ”, Opt. Commun. vol 284,pp. 4297–4306, (2011) .
[5] Y.C. Tseng and C.M. Chao, “Code Placement and Replacement Strategies for Wideband CDMA
OVSF Code Tree Management ”, IEEE Trans. Mobile Computing, vol. 1, no. 4, pp. 293-302, ( 2002).
[6]Yuh -Shyan Chen, and Ting -Lung Lin , “Code Placement and Replacement Schemes for WCDMA Rotated –
OVSF Code Tree Management ”, IEEE Trans. Mobile Computing, Vol. 5, no. 3, pp. 224 -239, (2006).
[7] F. Adachi, M. Sawahashi, and H. Suda, “Wideband DS -CDMA for Next -Generation Mobile
Communications Systems ”, IEEE Comm. , vol. 36, pp. 56-69, (1998 ).
[8] Yu-Chee Tseng, Member, and Chih -Min Chao , “Code Placement and Replacement
Strategies for Wideband CDMA OVSF Code Tree Management ”, IEEE Trans. Mobile
Computing, vol. 1, no. 4, pp 293 -302, (2002 ).
[9] R.G. Cheng and P. Lin, ” OVSF Code Channel Assignment for IMT -2000 ”, Proc. IEEE
Vehicular Technology Conf., vol. 3, pp. 2188 – 2192, (2000 ).
[10]W.T. Chen, Y.P. Wu, and H.C. Hsiao, “A Novel Code Assignment Scheme for W -CDMA
Systems ”, Proc. IEEE Vehicular Technology Conf., vol. 2, 1182 -1186, (2001 ).
[11] Dong H. Lee , ‘Logic design of orthogonal variable spreading factor code generator ”, J.
Circuit, Systems and Computers Vol. 14, pp. 507-513, (2005).
[12] Boris D. Andreev, Edward L. Titlebaum, and Eby G. Friedman , “Orthogonal Code Generator
for 3G Wireless Transceivers ”, GLSVLSI’03 , Washington DC – USA April 28 -29,(2003) .
[13] T.A. Moniem , M.H. Saleh , “Fuzzy logic membership implementation using optical
hardware components ”, Optics Communications vol 285, pp. 4474 –4482 , (2012).
[14]Y. Liu, M.T. Hill, H. de Waardt, G.D. Khoe, D. Lenstra, and H.J.S. Dorren , “All optical flip –
flop memory based on two polari -zation switches ”, Electron. Lett. Vo 38,pp. 904-905, (2002).
[15] Tamer A. Moniem ,” All optical Asynchronous Binary Counter Based on All Optical J -K flip -flop
Memory ”, Proceeding of ICECECE 2011 conference – Bangkok, Thailand Dec 25 -26, (2011).
[16] M.Morris Mano , “Computer engineering hardware design ”, prantice Hall international 3rd edition .
[17]G.P. Agrawal and N.K.Dutta, “Semiconductor lasers ”, Van Nostrand Reinhold , (1993 ).

252
Tamer A. Moniem and Noha EL Mosilhy , All optical OVSF code generat or realization using …
تصميم مولد ضوئي كودي لمعامل األنتشار متعامد التغير
باستخدام المتارجحات الضوئية المحدد الضوئي الصلب
الملخص العربى
معامل األنتشار متعامد التغير (OVSF) يقوم بتقديم معدالت متغيرة للبيانات ويدعم العديد من التطبيقات
المختلفة مع العديد من النطاقات الترددية المختلفة .حيث تقدم هذه الورقة مولد رمزي بصري ل OVSF
باستخدامالبوابات الضوئية ومجموعة من التأرجحات الضوئية القائمة على اثنين من مفاتيح االستقطاب
( PSWs). و يساعد المولد الرمزي OVSF الضوئي على استخدام تقنية كود OVSF في مجال االتصاالت
البصرية. وتناقش الورقة التصميم العام للمولد والنتائج من خالل المحاكاة العددية باستخدام الحاسب األلي
لتأكيد جدوى التصميم من خالل عاملي االنتشار 4 و8 .