Orthogonal Optical Codes for optical CDMA

Submitted by

Arishankari R (ID-2018H1240081H)

Rutuja Gomase (ID-2018H1240073H)

For

MidTerm project

In

Optical Communication

Department of Electrical and Electronics Engineering

October, 2019
 Abstract

Code division multiple access (CDMA) is a type of spread spectrum multiple access
technique where the user signal is spread over a bandwidth larger than the minimum required
for transmission. This is achieved by multiplying the user signal with an encoding sequence
that is some kind of a random signal. In CDMA system, all the users operate using the whole
system bandwidth and transmit at random timings for random durations unlike as in time
division multiple access or frequency division multiple access (TDMA/FDMA). There is
least coordination required in CDMA as opposed to TDMA/FDMA which require
cumbersome protocols to deal with issues like synchronization, slot allocation etc.

Optical CDMA is a very useful multiple access technique due to the features like
simultaneous network access, soft capacity and unique addressing of the users. Another
useful feature is that it can minimize the delay in accessing the network. Other benefits are
minimum network protocol and simultaneous access

Optical CDMA is based on assigning unique codes to all the users for an error-free operation.
Many parameters affect the performance of the optical CDMA system in addition to the
properties of the codes. The code used should have the following desired properties.

i. The code should be distinguishable from the shifted version of itself

ii. Maximum peak auto-correlation and minimum cross-correlation.
iii. Ability of faster synchronization at the receiver

In Optical CDMA each user at one end is connected to an optical star coupler (OSC). This
OSC is connected to other optical star couplers from other end through optical fibres. Each
user has its own transmitter and receiver section with separate assignment of optical
orthogonal codes (OOCs). The codes assigned to the transmitter section of a user
(information source) will also be provided to receiver section of other user as information
sink and vice versa. These optical orthogonal codes within a set are designed by an optical
orthogonal code design scheme. Since the spectral width of CDMA signal is large, this
scheme is also called spread spectrum communication.

Literature Survey
Code division multiple-access techniques for use in optical-fibre local-area networks

 In this literature survey we are discussing about CDMA and we know that Optical
fibre offers a virtually limitless communications bandwidth when used as the physical
medium for a local area network (LAN). However, conventional multiple-access
schemes are limited in their ability to exploit this potential.
 Direct-sequence spread-spectrum systems are a class of code division multiple-access
(CDMA) systems that encode data using high-rate sequences to spread data over a
much wider bandwidth than normally required for transmission. Therefore, media
access schemes better suited to the optical medium.
 CDMA Principle : CDMA covers systems in which the multiple-access capability is
achieved primarily by coding and the need for stringent co-ordination in time or
frequency between stations as in time division multiple-access (TDMA) or frequency
division multiple-access (FDMA) is removed by the mapping of data to a complex
time-frequency representation.
 Incoherent Optical CDMA: It has two types of codes have been developed for use in
incoherent Optical CDMA systems.
i) The first type is prime codes, is formed by time-mapping of one-coincidence
sequences originally developed for frequency-hopped systems. These codes
have cross correlation [functions with a maximum of one or two].
ii) The second type is optical orthogonal codes (OOCS), are so named to
emphasise their unipolar (optical) and orthogonal-like nature; hence their
suitability to incoherent optical correlation. These codes can be developed
with desirable auto- and cross correlation functions.
 Optical CDMA using conventional codes and electronic processing: Sequence
inversion keyed CDMA: A scheme has been proposed where direct-sequence spread-
spectrum techniques have been applied to an optical fibre CDMA LAN.
i) The system is based on sequence inversion keying and uses unipolar code
sequences to spread each bit of unipolar data sequences. While at the receiver
the unipolar channel sequences are correlated with a bipolar version of the
reference to recover the data. .
 Optical CDMA using sparse code and optical processing: Systems using Optical Fibre
Signal Processing :We have two asynchronous CDMA systems which have been
proposed based on a star network architecture, using OOK modulation, sparse optical
codes and optical-fibre processing elements - one using prime codes, the other using
optical orthogonal codes.
i) Data is encoded using the appropriate code type, either electronically via
direct modulation of a laser diode or optically by splitting a single laser
pulse and recombining the output of selected optical-fibre delay lines,
before being fed to the network via a star coupler.
ii) While at the receiver an incoherent optical correlation is performed by an
optical delay line matched filter whose impulse response is the time-
reversed, complex conjugate of the input sequence.
 Research which is going on in optical code division multiple-access local area
networks (CDMA LANs) is developing in two distinct areas, the majority of systems
concentrating on direct-sequence spread-spectrum methods.
i) In the first we have conventional CDMA techniques which are being
adopted.
ii) The second area of research is concerned with the development of CDMA
systems employing optical processors. Optical switching speeds offer much
higher data rates and the support of more users.
The use of CDMA in an optical LAN environment would allow simultaneous channel access
to a group of asynchronous users without the need for network scheduling and without access
delay. Although the LAN would be inefficient in terms of its bandwidth utilisation, the
effectively infinite bandwidth provided by optical fibre means that very impressive capacity
could still be achieved.

Optical Orthogonal Codes: Design, Analysis, and Applications

An optical orthogonal code is a family of (0, 1) sequences with good auto- and cross-
correlation properties

i. the autocorrelation of each sequence exhibits the ‘thumbtack’ shape – facilitates

detection of desired signal
ii. the cross correlation between any two sequences remains low throughput – reduces
interference from unwanted signals

Optical orthogonal codes enable efficient and reliable transmission.

Fundamental Properties of optical orthogonal codes:

An (n, w, A,, A,) optical orthogonal code C is a family of (0,l) sequences of length n and
weight w which satisfy the following two properties.

i. The Autocorrelation Property: : ∑𝑛−1

𝑐 𝑥(𝑡)𝑥(𝑡 + 𝑟) <=ƛc for any x E C and any
integer T, 0 < T < n.
ii. 2) The Cross-Correlation Property: ∑𝑛−1
𝑐 𝑥(𝑡)𝑦(𝑡 + 𝑟) <=ƛc for any x # y E C and any
integer T.

Applications

The objective is to design an efficient system, with available implementation technology, to

allow the users to share the common channel. Traditional multiple-access approaches such as
frequency division, time division, collision detection, or demand assignment require elaborate
network synchronization at high speed (often optical speed), and frequent conversions
between the optical domain and the electronic domain. These requirements limit the
efficiency of such an optical multiple-access system. However, by employing a code-division
multiple-access system with optical orthogonal codes, we are able to simplify greatly the
complexity of the system, to implement it with available technology, and to achieve
potentially higher transmission efficiency

There are several other potential applications of optical orthogonal codes. In spread-spectrum
communications, frequency hopping patterns are required to have low correlation. Optical
orthogonal codes can be used to generate good hopping patterns.

Design

i. Upper bounds
 Upper bounds on the maximum size of an optical orthogonal code Ɵ(n,w,A)
can be obtained from related results in algebraic coding theory.
ii. Iterative construction
Given an (n, w, A,, A,) code C, we can construct a (tn, tw, tw, th,) code C’
with the same number of code words in the following way. For each codeword
x of C construct a codeword z of C’ by concatenating t copies of x.
iii. The greedy algorithm and General lower bounds
The “greedy” algorithm is useful in many combinatorial and computational
problems. Originally, the code is empty. In (n-w) steps, the algorithm
examines all the binary n-tuples of weight w, one at a time. If an n-tuple
satisfies the autocorrelation property and satisfies the cross-correlation
property with every codeword already included, it is added to the code;
otherwise, it is discarded.
iv. Projective geometry
Optical orthogonal codes constructed from finite projective geometries
v. Combinatorial methods
Optical orthogonal codes can be constructed by various combinatorial
methods. For the case of A = 1, the problem of constructing OOC's is
equivalent to the problem of packing difference sets
vi. Block designs and algebraic coding theory
Another general approach to constructing optimal orthogonal codes is to use a
t - (U, b, A, k, A) block design. A t -(U, b, r, k, A) design consists of U objects
and b blocks (sets) of these objects, with each object contained in r blocks,
each block containing k objects, and each pair of objects contained in A
blocks.

Some new families of optical orthogonal codes for Code-division multiple-access fibre-
optic networks

a) An optical orthogonal code (OOC) is a collection of binary sequences with good auto
and cross-correlation properties. In this paper, three new families of OOCs are
constructed from the balanced in completed block design techniques (BIBD).
a. The first family of the codes is with ideal auto- and cross-correlation
properties and they have optimal cardinalities. The second and third families
of the codes are with ideal auto-correlation and nearly ideal cross-correlation
properties.
b. The second family of the codes is demonstrated to have twice as many
codeword’s as the optimal codes.
c. The third family of the codes is demonstrated to have more codeword’s than
the quadratic congruence codes when the weight of the codeword is less than
eight.
b) Constructions:
i. Construction 1: use BIBD to design an (n, w, 1, 1) OOC for w = 3, 4 and
5.Wilson has generalised these results for arbitrary w and presented a
 construction for this w. Both for odd and even w we have given the Wilson
construction (detailed explanation in the reference paper).
ii. Construction 2: we can use essentially the same techniques as construction 1
to construct (n, w, 1, 2) codes with twice as many codeword’s as the (n, w, 1,
1) codes.
iii. Construction 3: This construction result is that we can use essentially the same
techniques as above two to construct
(n, w, 1, 2) with (w - 1) times (for odd w) or w times (for even w) as many
codeword’s as the (n, w, 1, 1) codes when w is less than eight.
c) The cardinality of the newly constructed codes from construction 3 is w times (for
even w) or (w – 1) times (for odd w) as many codeword’s as the (n, w, 1, 1) codes
when w is less than eight.
Proposed Idea
The project aims to understand and construct an orthogonal optical code for CDMA. After
constructing the code, we plan to understand its drawbacks in practical scenario. There are
different proposed ideas and methodologies to overcome the drawbacks. We plan to analyze
these proposed ideas and try to implement them. The possibility is towards two dimensional
Orthogonal Optical Codes.

Plan of work
i. Understanding CDMA
ii. Understanding the need of CDMA in FOC
iii. Understanding OOC
iv. Construction of OOC
v. Understanding the drawbacks
vi. Two dimensional OOC

References
1. M. J. Parham, C. Smythe, and B. L. Weiss, "Code Division Multiple-access
Techniques for use in Optical-fibre Local-area Networks." Electron. Commun. Ens.
Journal, pp. 206-212,Aug. 1992
2. F. R. K. Chung, J. A. Salehi, and V. K. Wei, “Optical orthogonal codes: Design,
analysis, and applications,” IEEE Trans. Inform. Theory, vol. 35, pp. 595–604, May
1989
3. G.-C. Yang, “Some new families of optical orthogonal codes for code division
multiple-access fiber-optic networks,” IEE Proc. Commun., vol. 142, pp. 363–368,
Dec. 1995