A strategy for minimising the probability of contention in optical burst-switching (OBS) networks, by
combining traffic engineering in the wavelength domain with delayed burst scheduling at the ingress nodes,
has been proposed. The implementation of this strategy only requires an increase on the processing and
electronic buffering capacities of the ingress nodes, while avoiding the use of complex and expensive optical
buffers at the core nodes. The performance evaluation study has shown that the proposed strategy
outperforms the burst-scheduling strategies described in the literature by several orders of magnitude, in
terms of burst-blocking probability. Moreover, the performance improvements have been shown to depend on
the average offered traffic load, number of wavelengths per link and maximum delay allowed at the ingress
nodes.