Robust routing vs dynamic load-balancing a comprehensive study and new directions

Traffic engineering (TE) has become a challenging task for network management and resources optimization due to traffic uncertainty and to the difficulty to predict traffic variations. To address this uncertainty in a robust and efficient way, two almost antagonist approaches have emerged during the last years: robust routing and dynamic load-balancing. The former copes with traffic uncertainty in an off-line preemptive fashion, computing a stable routing configuration that is optimized for a large set of possible traffic demands. The latter balances traffic among multiple paths in an on-line reactive fashion, adapting to traffic variations in order to optimize a certain cost-function. Much has been said and discussed about the advantages and drawbacks of each approach, but very few works have tried to compare the performance of both mechanisms, particularly in the same network and traffic scenarios. This paper brings insight into several robust routing and dynamic load-balancing algorithms, evaluating their virtues and shortcomings and presenting new mechanisms to improve previous proposals. Among others, such a study intends to help network operators in choosing an adequate mechanism to cope with traffic uncertainty.