Information Technology Reference
In-Depth Information
Figure 7. Experiment scenario
We are interested in measuring and comparing
the e2e performance including delay, reliability,
and fairness associated with delay and reliability.
In this chapter, we adopt Jain's fairness measure
(Jain, Chiu, & Hawe, 1984), which is defined as
f x x
the least e2e delay, while LEAST has the greatest
delay, whit PROBE in the middle. The reason why
MQMAC-CQBR-HS is better than PROBE is that
it has a hybrid scheduler to guarantee delay; the
Packet Error Rate (PER) is reduced by exploiting
channel diversity to allow simultaneous transmis-
sions and avoid interference, and by selecting
the best route with best link quality; conversely,
PROBE does not have the scheduler and can only
select the best route with good link quality without
the option of choosing better channels. On the
other hand, PROBE is better than LEAST as the
latter chooses the route with the shortest number
of hops, leading to the severest interference from
neighbor nodes while PROBE tends to select bet-
ter links for packet forwarding.
Our proposal provides effective priority ser-
vices based on the vital sign's e2e delay require-
ments. As for e2e delay, we have ECG < CO 2 <
=
2
(
x
)
for a set of values x 1 ,
x 2 , …, x n , and compare these protocols. It ranges
from 1/n (worst case, i.e., most unfair) to 1 (best
case, i.e., maximum fairness for all these n values).
(
,
, ...,
x n
)
i
i
1
2
2
n
(
x
)
i
i
Delay and Reliability
The curves for our solution in the following figures
are denoted by “X-MQMAC-CQBR-HS”, where
X represents the patient's data type (ECG/CO 2 /
CVP). As for e2e delay, our experiment (Figure
8(a)) shows that MQMAC-CQBR-HS < PROBE
< LEAST. On average, MQMAC-CQBR-HS has
Search WWH ::




Custom Search