Iterative Nonlinearity Mitigation and Decoding for LED Communications

A Deep Learning Approach to Universal Binary Visible Light Communication Transceiver

“…However, in some cases such as mobile communication, the communication channel may vary at each transmission and cannot be perfectly known to the receiver. Furthermore, material properties of practical LEDs, such as LED memory effect and nonlinear electro-to-opto transfer function [35], [36], also have nonlinear impacts on the performance. The resulting received signal can be expressed as…”

“…In IM/DD-based systems, the input current modulated by the binary electrical signal based on binary codeword s b,d is applied to the LED [35], [36], whereas the corresponding LED output g(s b,d ) is the optical power transmitted through the optical channel (28). Thus, it is still necessary to force the LED input current s b,d to be of binary alphabets, which can be accomplished by the proposed stochastic binarization.…”

“…Due to a wall located at 3 m, there exists a LED-wall-PD reflection path incurring ISI to a LoS path from the LED to the PD. A commercial Kingbright blue T-1 3/4 LED [37] is considered with electro-to-opto transfer function g(z) modeled as (29) with K = 4, ζ = 0.1, a 1 = 34.11, a 2 = −29.99, a 3 = 6.999, and a 4 = −0.1468 [36]. The position of the LED is fixed at (1.5 m, 3 m), while the PD suffering from the delayed ISI signal reflected by the wall is located at random on the floor as (0 m, P m) with P being the uniform random variable within According to [25], the received signal r in (1) for the ISI models can be written as…”

Extreme-Learning-Machine-Based Noniterative and Iterative Nonlinearity Mitigation for LED Communication Systems

“…It can be seen that all of them depend on the LED nonlinearity, i.e., the memory polynomial coefficients {a k,m }. In [20], the iterative receiver is designed with the perfect knowledge of of LED nonlinearity and memory effects, i.e., assuming that the memory polynomial coefficients {a k,m } are perfectly known. However, they are usually unknown in practice.…”

Primitive Polynomials for Iterative Recursive Soft Sequential Acquisition of Concatenated Sequences

“…However, adopting any hard decision based coherent approach becomes unrealistic, since the SNR is usually very low prior to despreading [7], [9]. For this reason, we advocate the recursive Soft Input Soft Output (SISO) detection principle, which has its roots in the classic turbo channel decoding philosophy [18], [20], [24], [25]. Before delving into the new contributions of this paper we have summarized the evolution of the field in Fig.…”