When receivers rely on stochastic energy harvesting, the receiver processing may
be interrupted by energy outages that affect the reliable communication rate. To
model the receiver, we decompose the processing tasks in two parts: first is
sampling or Analog-to Digital-Conversion (ADC) which includes all RF front end
processing, and second is decoding. In periods when the energy harvesting rate
is low, the receiver may choose to sample and store the received signal but
defer message decoding. In this work, we consider a source that transmits
codewords from a fixed rate codebook through a memoryless channel. Based on the
relative energy costs of sampling and decoding, we determine the policy that
maximizes the receiver's message decoding rate.