Channel Estimation and Signal Detection for Optical Wireless Scattering Communication With Inter-Symbol Interference

A statistical channel model for the non-line of sight (NLOS) optical wireless scattering communication link is fundamental for developing further receiver-side digital signal processing techniques. We adopt a linear time-invariant (LTI) Poisson channel modeling method, which was originally proposed for molecular communication, to model the NLOS optical wireless scattering communication. In such a model, we characterize stochastic inter-symbol interference (ISI) based on a probabilistic delay profile. We derive the first- and second-order statistics for the received signals. Moreover, we propose channel estimation schemes to estimate the delay profile, based on the least-squares and correlation criteria, and analyze the performance of the proposed channel estimation schemes. Finally, we propose and compare two signal detection schemes, namely the simple linear minimum mean square error (LMMSE) receiver treating ISI as noise and the maximum-likelihood sequence detection (MLSD) scheme that takes ISI into account. It is shown that MLSD significantly outperforms LMMSE under channel ISI. The proposed ISI channel model and the associated channel parameter estimation schemes serve as a theoretical foundation for high symbol rate with high transmission power in optical wireless scattering communication.