Signal detection and ISI cancellation for quantity-based amplitude modulation in diffusion-based molecular communications

Simulation study of molecular communication systems with an absorbing receiver: Modulation and ISI mitigation techniques

“…In the literature, the ISI effect is not considered adequately, instead just one symbol ISI is assumed and some semi-analytical formulations are derived [22]. In [23], a 1-D environment is considered to derive the analytical formulations for ISI mitigation. We need to analyze the ISI effect in a 3-D environment hence, to have more realistic results, we send the symbols consecutively without making any assumptions on the ISI.…”

Performance Analysis of Amplitude Modulation Schemes for Diffusion-Based Molecular Communication

“…The quantity S k has a binomial distribution with parameters β x and p 0 , when hypothesis H x is true. For large β x , we approximate the distributions of S k and I i by Gaussian distributions N (β x p 0 , β x p 0 (1 − p 0 )) and N (β x p i , β x p i (1 − p i )), respectively [15], [21], [30]. 1 To make the interference independent of the transmitted symbol x, x ∈ X , we average over all possible transmissions to obtain the mean µ Ii and variance σ 2…”

“…Each of these works leaves out one or another critical effect; for example, [15] accounts for ISI only due to one previous symbol, [21], [22] perform analysis without considering the effect of external noise, and [16] provides only a lower bound and not a closed form expression. Authors in [17] consider finite life expectancy but provide analysis for a binary scheme only.…”

“…Considering ISI from one previous symbol only and additive white Gaussian noise, capacity for binary and 4-ary schemes has been evaluated in [15]. In the absence of noise, M -ary detection for diffusion-based communication has been studied in [21], [22]. A lower bound on the probability of error for binary and 4-ary schemes has been computed in [16].…”

Local Convexity Inspired Low-Complexity Noncoherent Signal Detector for Nanoscale Molecular Communications

“…Note that, for the non-Gaussian noises [23], the corresponding likelihood density will be required when performing the Bayesian estimation. 1 The sampling time may be smaller than the symbol duration for coherent detection [17]. As will be seen later, for the considered non-coherent detector, an over-sampling technique would also be adopted where Ts < T b .…”