Predictive Quality Of Service Schemes For Real-Time Multimedia Applications In Communications
In guaranteed QoS resources are reserved based on the worst-case analysis. This scheme guarantees QoS but results in low resource utilisation, subsequently depriving other users from acquiring service. In contrary, predictive QoS allocates an initial amount of resource, whereby QoS is guaranteed...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English English |
| Published: |
1999
|
| Subjects: | |
| Online Access: | http://psasir.upm.edu.my/id/eprint/11893/1/FSKTM_1999_5_A.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | In guaranteed QoS resources are reserved based on the worst-case analysis.
This scheme guarantees QoS but results in low resource utilisation, subsequently
depriving other users from acquiring service. In contrary, predictive QoS allocates
an initial amount of resource, whereby QoS is guaranteed with a certain
probability of degradation. This scheme is tailored for applications that are
adaptive and robust towards sudden fluctuations in the service provided. The
nature of multimedia data such as the variable bit rate has encouraged the
implementation of predictive Quality of Service (QoS) as compared to guaranteed
QoS.
In a typical QoS scheme, two factors contribute to the computation of
packet loss, which are : (i) a new packet dropped due to buffer overflow and (ii) a
buffered packet dropped due to expired delay. The buffer resource is increased
when the observed packet loss ratio has violated the requested level of packet loss
ratio. The limitations in this scheme is caused by the admission of all packets into the buffer, inclusive of packets that will eventually be dropped due to expired
delay. Subsequently, this results in poor resource management.
In this research, two pro-active dynamic QoS control schemes are
designed, the dynamic QoS control scheme with delay estimation, and the hybrid
dynamic QoS control scheme. In both schemes, every new packet arrival is
compared against the estimated delay it will experience, before being admitted into
the buffer. If the estimated delay expires the requested delay bound, then the
packet is dropped. In the hybrid scheme, every packet is checked before being
admitted into the buffer, and also, the packets successfully admitted into the buffer
are evaluated on the actual delay experienced before being transmitted to the
receIver.
The results obtained through the simulation models have shown that two
schemes have significantly improved the average delay for different traffic
patterns. In addition to improving the average delay in delay sensitive traffic,
improvement is seen in the average packet loss ratio, and subsequently increasing
the throughput for delay sensitive traffic. However, in packet loss sensitive traffic,
the old scheme remains beneficial. The proposed scheme can be adopted for
multimedia applications to enhance the QoS in terms of better delay and improved
utilisation. |
|---|