XTP and Vmtp on Multiprocessor Architectures

The processing of data units inside a communication node became to be the bottleneck in communications, because protocol processing did not increase in the same order of magnitude as the networks' bandwidth. For that reason, new and efficient approaches to surmount this bottleneck have been proposed. First, the development of new protocols and, second, the use of dedicated hardware platforms have a number of advantages. Throughout this paper the integration of these advantages will be presented using the light-weight protocols XTP [1] and VMTP [2] on multiprocessor architectures. Measurement and simulation results proof the performance gain of the selected approach. The results show that XTP and VMTP are more appropriate for implementations on parallel architectures than comparable OSI protocols.