The effect of multiple simultaneous flows in the sensor
network is investigated next, with random selection of source
and destination nodes, while the number of flows is varied
from 1 to 6. Again, we observe the behavior of RF-MAC on
two energy and throughput metrics, when nodes experience
different levels of channel usage and traffic loads. Fig. 9 shows a
smooth and monotonic increase in the average harvested energy
of RF-MAC-opt as the number of flows increases. Even though
the RF-MAC-no-opt exhibits a similar pattern, the increase is
not as smooth as one with the frequency optimization. Evidently,
the amount of average harvested energy yield could
be almost 150% less than RF-MAC-opt. Fig. 10 depicts the
average network throughput of RF-MAC with various numbers
of data flows. Interestingly, the average network throughput
of both RF-MAC variants gracefully drops as the number of data flow increases. This reduction in the throughput is a result
of more nodes sending out RFEs as they deplete their energy
faster with increasing number of data flows. Consequently, the
network spends more time in the charging state and less time
spent in the data exchange state. However, RF-MAC-opt yields
higher average network throughput, approximately 20% more
in this case. Again, both variants of RF-MAC largely outperform
the modified CSMA. Especially, RF-MAC-opt yields
approximately 112% increase in terms of throughput.