DYNAMIC PRIORITY SERVERS - Hard Real-Time Computing Systems: Predictable Scheduling Algorithms and Applications - page 187

Periodic Load = 65% Mean Aperiodic Interarrival Time = 100

1

Polling
DSS
DPE
TBS
IPE

0.8

0.6

0.4

0.2

0

0 3 6 9 12 15 18 21 24 27 30 33

Mean Aperiodic Load (%)

Figure 6.12 Performance of dynamic server algorithms.

Note that in all graphs, TBS and IPE have about the same responsiveness when the
aperiodic load is low, and they exhibit a slightly different behavior for heavy aperiodic
loads.

All algorithms perform much better when the aperiodic load is generated by a large
number of small tasks rather than a small number of long activities. Moreover, note
that as the interarrival time T a increases, and the tasks' execution time becomes longer,
the IPE algorithm shows its superiority with respect to the others, which tend to have
about the same performance, instead.

The proposed algorithms have been compared with different periodic loads U p as well.
For very low periodic loads all aperiodic service algorithms show a behavior similar
to background service. As the periodic load increases, their performance improves
substantially with respect to background service. In particular, DPE and DSS have
a comparable performance, which tends to approach that of the Polling Server for
high periodic loads. On the other hand, TBS and IPE outperform all other algorithms
in all situations. The improvement is particularly significant with medium and high
workloads. With a very high workload, TBS is no more able to achieve the same good