Optics Optik Optik Optik 120 (2009) 959–962 A new family of two-dimensional codes for optical CDMA systems Jaswinder Singh a, , Maninder Lal Singh b a Department of Electronics & Communication Engineering, Beant College of Engineering & Technology, Gurdaspur, India b Department of Electronics Technology, Guru Nanak Dev University, Amritsar 143005, India Received 30 October 2007; accepted 12 March 2008 Abstract The design of a new family of two-dimensional single pulse per column codes for optical code division multiple access (OCDMA) networks is reported. The 1-D modified pseudo-noise codes have been known to be orthogonal and their generation and system design based on these codes is rather simple. But their performance is limited due to the bandwidth constraints if the code length increases. Hence, using these 1-D modified pseudo-noise codes, modified 2-D pseudo-noise matrix codes (MPMCs) are generated. The system performance is evaluated for two, three and four simultaneous users using the link with all the sources responsible for degradation included: attenuation, chromatic dispersion, non-linear refractive effects, non-linear scattering and four-wave mixing. The effect of the non-linear and lossy dispersive medium over the system performance is shown by plotting the BER with respect to the link length for the systems designed using encoders/decoders base on 1-D modified pseudo-noise codes and our MPMCs. The performance is compared for the two types of codes by finding the crosstalk due to interfering users simultaneously operating in the network. r 2008 Elsevier GmbH. All rights reserved. Keywords: Differential detection; Multi-access interference; Optical code division multiple access (OCDMA); Single pulse per row (SPPR) codes; Two-dimensional W-T codes 1. Introduction For high-speed access and local-area networks, the passive broadcast optical networks have been consid- ered the most promising. Optical code division multiple access (OCDMA) is one such scheme with many attractive features (Fig. 1). The codes used for designing the encoders/decoders for fiber-optic communication systems are the (0, 1) codes because of the non- negativity of the optical fiber channel. The PN codes like gold codes as well as the maximal length codes suffer from the multi-access interference problem as the number of 0’s and 1’s in these codes are not equal so the modified PN codes have been developed [1], but these belong to the class of one-dimensional codes. Because of the poor performance with the one- dimensional codes, many two-dimensional code families have been proposed [2–4] and designed for different kinds of detection schemes such as IM/DD and differential detection [5]. In [2], the prime codes have been used for time-spreading and optical orthogonal codes for frequency hopping patterns to obtain better crosstalk properties. But still the BER performance is not very convincing for 1 Gbps data rates as only a few users can be behaviour accommodated for BER of ARTICLE IN PRESS www.elsevier.de/ijleo 0030-4026/$ - see front matter r 2008 Elsevier GmbH. All rights reserved. doi:10.1016/j.ijleo.2008.03.031 Corresponding author. E-mail addresses: j_singh73@rediffmail.com (J. Singh), mlsingh7@yahoo.co.uk (M.L. Singh).