Quality of service mechanisms in optical burst-switched networks

The explosive growth of the Internet demands a high-speed transmission technology for supporting rapidly increasing bandwidth requirements. Currently, wavelength division multiplexing (WDM) technology enables multiplexing of 160–320 wavelengths into a single fiber, with a transmission rate of 10–40 Gb/s per wavelength. In order to efficiently utilize the raw bandwidth in WDM networks, an all-optical transport method, which supports fast resource provisioning and asynchronous transmission, must be developed. Optical burst switching (OBS) is such a promising bufferless WDM switching technology. An important issue in OBS networks is the development of quality of service (QoS) mechanisms in order to offer better services for diverse Internet applications. In this dissertation, we propose an absolute QoS model which ensures that an upper bound on the burst loss probability is guaranteed for the high-priority class. We develop two mechanisms for providing loss guarantees at OBS core nodes: an early dropping mechanism and a wavelength grouping mechanism. We also discuss admission control and resource provisioning for OBS networks, and propose a path clustering technique to further improve the network-wide loss performance. We develop analytical loss models for the proposed schemes and verify the results by simulation. We also propose a burst retransmission scheme in OBS networks in which bursts lost in the OBS network are retransmitted in the OBS transport layer. Our retransmission scheme reduces burst loss probability significantly. We develop an analytical model for obtaining the burst loss probability over an OBS network that uses the retransmission scheme. TCP carries a majority of Internet traffic, thus understanding and improving the performance of TCP implementations over OBS networks is critical. Due to the bufferless nature of OBS networks, randomburst losses may occur, even at low traffic loads. For OBS networks in which TCP is implemented at a higher layer, these random burst losses may be mistakenly interpreted by the TCP layer as congestion in the network, leading to serious degradation of the TCP performance. We adopt the burst retransmission scheme to reduce the probability that the TCP layer falsely detects congestion, thereby improving the TCP throughput. We analyze the TCP throughput when OBS networks employ the burst retransmission scheme and develop a simulation model to validate the analytical results.