Fleksibilni Internet servisi na bazi kontrole kašnjenja i

5. Scheduling for proportional delay differentiation
Proportional delay differentiation assumes that the ratio of average queuing delays
of throughput-sensitive and delay-sensitive packets should be approximately equal to a
specified delay differentiation parameter. We consider a work-conserving scheduler
serving two FIFO queues managed by an active queue management (AQM) algorithm,
which makes drop decisions upon packet arrivals (once being admitted, the packet is
never dropped from the queue). The work-conserving property assumes that the scheduler
does not idle when there are packets in a queue waiting to be transmitted.
Scheduling algorithms for proportional delay differentiation provide relative per-
hop delay guarantees. One may argue that real-time applications require absolute end-to-
end guarantees because their performance depends on probability of being bellow certain
end-to-end delay threshold [13]. Proportional delay differentiation decreases the
probability of exceeding the threshold for the delay-sensitive class. This makes it more
suitable for real-time applications than the existing best-effort service. It has been also
shown that per-hop proportional delay differentiation easily translates into proportional
end-to-end delay differentiation [23]
In this Section, we discuss two schedulers for proportional delay differentiation:
the well-known Backlog-Proportional Rate scheduler (BPR) scheduler [12] and a newly
proposed idealized scheduler named Optimized Backlog-Proportional Rate (BPR + )
scheduler [38] that we introduced in order to evaluate the performance of BPR compared
29