An adaptive wireless communication protocol for neural data transmission in freely behaving subjects

Long-term continuous intracortical recording of neuronal ensembles in freely behaving subjects requires a reliable wireless communication channel for transmitting important biological information. The need for ultra low-power, fully implanted recording systems, however, make the design of the wireless transmission protocol more demanding. Here, we introduce an adaptive protocol that can cope with the variable characteristics of the errors in the wireless channel associated with different levels of subject mobility, for example, during rest and active states. The wireless channel is modeled as a finite-state Markov channel, in which states are binary symmetric channels with different binary error rates. A convolutional encoder with a specific code rate is incorporated into each state, for which the length of data transmission packets is optimally estimated. The protocol can switch between different states depending on subject mobility to ensure a highly reliable communication channel, while optimizing the power consumption by minimizing the average memory length required for storing packets prior to transmission.