3.3.1 Throughput Performance
Once the value ?„ is known, the throughput can be computed via Eq. (12), here reported for
the convenience of the reader:
( )c success idle s success idle
success
T P P T P P
] P [ E P S
??’ ??’ + +
=
1 ??
(37)
where
( )
( ) 1 1
1
??’ ??’ =
??’ =
N
success
N
idle
N P
P
?„ ?„
?„ (38)
We recall that the denominator in Eq. (37) represents the average slot size E[slot]. If
E[P], instead of bits, is expressed in the same time unit of the parameters at the
denominator (e.g., seconds), then Eq. (37) gives the ???normalized??? saturation throughput,
defined as the fraction of channel time used to send the successful payload information.
The fundamental difference with respect to the treatment suggested in Section 4.3.2 is
that, now, ?„ is no more a variable (i.e., a generic permission probability), but it is a
numerical value function of the considered backoff model parameters, namely, the retry
limit R and the sequence of mean per-stage backoff values ??
i , for all i in (0,??¦,R).
The values Ts and Tc have been computed in Section 4.3.2.1. However, some further
remarks are needed when the analysis is applied to the DCF.
Pages:
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251