With the advent of high-speed networking technologies such asfiber optics, a traditional bottleneck in communication has disappeared.Now, the speed at which a processor can execute a communication protocolis the most significant limiting factor in protocol performance. Thequestion is, how can a multimedia application take best advantage ofthese high-speed networks? We adopt an optimized implementation approachthat allows efficient use of the bandwidth available on today'shigh-speed networks. In this paper, we propose a new horizontalprocessing architecture that processes data as soon as it arrives fromthe network. This model increases the processing speed of the OSI modelby using a multiple-instruction, single-data (MISD) scheme. However, theOSI protocol stack often imposes ordering constraints that preventconcurrent processing of the protocol layers. By using a fixed packetformat, all layers from the network layer through to the presentationlayer are processed in parallel. A prototype shows that the proposedmodel has a performance improvement of up to 60% more than theconventional approach
展开▼