Bit-error rate analysis for PIM-CDMA optical wireless communication systems Chun K. See * , Zabih Ghassemlooy ** , John M. Holding Optical Communications Research Group, School of Engineering, Sheffield Hallam University, Pond Street, Sheffield, S1 1WB. U.K. ABSTRACT A hybrid pulse interval modulation – code-division multiple-access (PIM-CDMA) is proposed for optical wireless communication systems. It offers higher transmission rate compared with the hybrid pulse position modulation (PPM) – CDMA due to its non-fixed frame structure. Strict optical orthogonal code is used as a signature sequence located at the start each frame. Expressions for bit error rate are given for both PPM-CDMA and PIM-CDMA schemes, assuming that the dominant noise and interference are due to the self-interference and optical multiple access interference. The numerical results are compared with both PPM-CDMA and OOK-CDMA, and it shows that the proposed scheme is a viable scheme for optical wireless communications. Keywords: PIM, PPM, OOK, CDMA, Strict-OOC. 1. INTRODUCTION There have been a growing number of research interests in using optical wireless for future indoor communication systems 1-3 . Compared with RF systems it offers many advantages such as higher bit rate, reuse of same frequency in the adjacent rooms, security and free from spectrum regulation and licensing 4 . However, there are two major impairments associated with optical wireless systems when employing intensity modulation with direct detection (IM/DD). The first is the additive noise due to sunlight and artificial light, which degrades the signal-to-noise (SNR) performance, and the second is the multi-path dispersion associated with the non-direct line of sight infrared systems, which results in inter-symbol interference (ISI) 1-3 . Wong et al 5 proved that the use of a signature sequence from the spread spectrum/code-division multiple-access (CDMA) technique overcomes the effect of multi-path dispersion and artificial light interference. On the other hand, using a family of signature sequences, each having unique signature sequence, will enable a multiple and asynchronous transmission in a communication system 6-11 . There are two families of signature sequences have been proposed for optical systems: (i) optical orthogonal code (OOC) 8-9 and (ii) Prime codes 10-11 . The former is preferable since it offers improved error performance 12 , and provides potential solution suited for application in the optical wireless multiples access environment. However, CDMA schemes employing OOC address code are designed to support constant-bit-rate (CBR) applications 17 . There exists an alternative CDMA address code, known as strict-OOC’s, which is more suitable for variable-bit-rate applications and can guarantee the minimum value for the correlation constrains 9 . This is achieved without the need for bandwidth expansion and increased system complexity, compared with the conventional OOC’s. In this work strict-OOC address code has been adopted 13 . 1.1. On-off keying (OOK)-CDMA This was first proposed for fibre optic communication systems of N users using (n,w,λ a , λ c ) OOC as signature sequences 6-7 , where n is the sequence length of w weight and λ a and λ c are the auto- and cross-correlation constraints, respectively. Both λ a and λ c are usually set to 1 to achieve the optimum bit error rate (BER) performance 6 . The signature sequence C i , for the i th user can be represented in binary format as ( ) i n i i a a a ,..., , 2 1 , where i =1,2,…,N and ( ) 1 , 0 ∈ i j a is the binary patterns in the j th * C.K. See; Email: c.k.see@shu.ac.uk; Tel.: (44) 0114 225 3301 ** Z.Ghassemlooy; Email: z.f.ghassemlooy@shu.ac.uk; Fax: (44) 0114 225 3433