Modulation and coding techniques for infrared wireless local area networks

Because of low-cost optical devices and virtually unlimited bandwidth, optical wireless communications (OWC) for indoor wireless local area networks (WLANs) have recently become an attractive alternative to radio frequency systems. Since optical signals cannot penetrate through walls or other opaque barriers, the security of infrared WLANs is very high and there is no interference between rooms. Subsequently, cell planning is simple and easy, and the potential capacity of an optical-based network in a building is extremely high. However, the system link is susceptible to path loss and multipath dispersion. In addition, the average transmit power is constrained by eye-safety regulations and power consumption concerns. Hence, most recent research deals with the physical layer aspects such as modulation, equalization and error-control coding in order to cope with these drawbacks, especially the effects of multipath dispersion. The objective of this thesis is to study practical signaling techniques capable of eliminating the effects of intersymbol interference (ISI).; Regarding the drawbacks of OWC, modulation schemes which are power and bandwidth efficient are considered. Pulse-position modulation (PPM) has been employed for IrDA and IEEE802.11 standards because it offers high power efficiency. However, it requires high bandwidth so that its performance is considerably degraded when the channel is more corrupted by ISI. A number of modified PPM techniques have been proposed to improve bandwidth efficiency. This thesis introduces a hybrid between pulse-amplitude modulation (PAM) and differential pulse-position modulation (DPPM), named differential amplitude pulse-position modulation (DAPPM), in order to gain a better compromise between power and bandwidth efficiency. It yields better bandwidth and/or power efficiency than PAM, PPM and DPPM depending on the number of amplitude levels (A), and the maximum length (L) of a symbol.; The channel capacity of PPM, DPPM and DAPPM systems is investigated. Since these modulation schemes over an ISI channel can be represented by a trellis diagram, their channel capacity is determined using a method for calculating the capacity of a Markov process channel. Over a soft-output channel, DAPPM achieves a higher capacity and is less sensitive to multipath dispersion than PPM and DPPM. Moreover, the comparison of hard-decision decoding (HDD) and soft-decision decoding (SDD) for PPM and DPPM systems shows that the performance of SDD is superior to that of HDD, especially when the channel is dispersive.; Then, some soft-decision techniques for DPPM system are considered. Although maximum-likelihood sequence detection (MLSD) is the optimal SDD for DPPM system, its complexity is extremely high. This thesis examines SDDs which are less complex than MLSD, but have performance close to that with MLSD. As the DPPM system is a Markov process, maximum a posteriori (MAP), Max-Log-MAP and the soft-output Viterbi algorithm (SOVA) are adopted. In addition, a novel very low complexity soft-decision decoding algorithm is introduced. The performance of the proposed algorithm is independent of the knowledge of the channel model, while the performance of the optimal and suboptimal MAP algorithms is impaired when the receiver has no information about the channel.; Finally, to achieve lower power requirements, error-control coding in an OWC system is investigated. Because insertion and deletion errors exist in DPPM systems, conventional coding techniques cannot be used. This thesis presents the concatenation of marker and Reed-Solomon codes which is able to correct such errors. The coded systems with HDD and SDD are examined by analysis and simulation.