Correction of Misalignments in Molecular Index Modulated Systems

Molecular communications employs messenger mo-lecules to establish communication between small scale transmit-ter and receivers in a liquid environment. Messenger molecules released by the transmitter propagate through the medium via diffusion and some of them arrive at the receiver. The receiver counts the messenger molecules that have reached its surface and tries to estimate the transmitted information. Recently, in order to increase the channel communication capacity, multiple-input multiple-output molecular communication systems have been proposed in the literature. One recent example is the molecular index modulation systems, where the information is carried on the index of the active transmitter. Alignment of the transmitter and receiver planes is a crucial problem for molecular index modulation systems. Although there are analytical models for the single-input single-output communication systems, there is no such result for multiple-input multiple-output communication systems, which makes it impossible to study the misalignment problem analytically. In this work, a machine learning-based method is introduced to estimate the amount of misalignment with high accuracy.