An intelligent approach to quality of service for MPEG-4 video transmission in IEEE 802.15.1

Nowadays, wireless connectivity is becoming ubiquitous spreading to companies and in domestic areas. IEEE 802.15.1 commonly known as Bluetooth is high-quality, high-security, high-speed and low-cost radio signal technology. This wireless technology allows a maximum access range of 100 meters yet needs power as low as 1mW. Regrettably, IEEE 802.15.1 has a very limited bandwidth. This limitation can become a real problem If the user wishes to transmit a large amount of data in a very short time. The version 1.2 which is used in this project could only carry a maximum download rate of 724Kbps and an upload rate of 54Kbps In its asynchronous mode. But video needs a very large bandwidth to be transmitted with a sufficient level of quality. Video transmission over IEEE 802.15.1 networks would therefore be difficult to achieve, due to the limited bandwidth. Hence, a solution to transmit digital video with a sufficient quality of picture to arrive at the receiving end is required. A hybrid scheme has been developed in this thesis, comprises of a fuzzy logic set of rules and an artificial neural network algorithms. MPEG-4 video compression has been used in this work to optimise the transmission. This research further utilises an ‘added-buffer’ to prevent excessive data loss of MPEG-4 video over IEEE 802.15.1transmission and subsequently increase picture quality. The neural-fuzzy scheme regulates the output rate of the added-buffer to ensure that MPEG-4 video stream conforms to the traffic conditions of the IEEE 802.15.1 channel during the transmission period, that is to send more data when the bandwidth is not fully used and keep the data in the buffers if the bandwidth is overused. Computer simulation results confirm that intelligence techniques and added-buffer do improve quality of picture, reduce data loss and communication delay, as compared with conventional MPEG video transmission over IEEE 802.15.1.

Name dc:type.qualificationname

phd

Level dc:type.qualificationlevel

doctoral

Grantor dc:publisher.institution

London Metropolitan University

Year dc:date.issued

2010

Author dc:creator

• Remy, Guillaume

• 620 Engineering & allied operations

Dc Identifier Grantnumber

N/A

OAI identifier oai:identifier

oai:repository.londonmet.ac.uk:8318