ALRP: scalability study of ant based local repair routing protocol for mobile adhoc networks
WSEAS Transactions on Computer Research
Maximum entropy method and underdetermined systems applied to computer network topology and routing
AIC'09 Proceedings of the 9th WSEAS international conference on Applied informatics and communications
A heuristic algorithm for the network design problem
NN'10/EC'10/FS'10 Proceedings of the 11th WSEAS international conference on nural networks and 11th WSEAS international conference on evolutionary computing and 11th WSEAS international conference on Fuzzy systems
An algorithm for the network design problem based on the maximum entropy method
AMERICAN-MATH'10 Proceedings of the 2010 American conference on Applied mathematics
| Hi-index | 0.00 |
Computer network routing is a very important and interesting optimization problem. Many different routing algorithms have been used over the years on the Internet, often with unexpected problems. Dynamic systems, i.e. systems that change over time, can be optimized statically with a fixed solution that corresponds to some average system state, or dynamically where the solution tries to follow the system change over time. It is a normal expectation that dynamic optimization has to give better results than a static one. Dynamic optimization is more complex, requires more computation, more advanced methods, but is superior to static optimization because it can always be transformed to the static case simply by neglecting change of the system in time and selecting a single state as a representative. However, that expectation that dynamic optimization gives better results than static one applies only to the perfect dynamic optimization, which is impossible in practice. It takes some time to collect information about the system current state, and optimization is always done with that obsolete information. This situation is examined on computer network routing. By complete mathematical analysis of a simple network, we show that dynamic routing gives better results than static, as expected, but that the margin is much smaller then intuitively expected. Further analysis shows that that minor advantage can easily be lost if there is even a small error in the dynamic routing tables, and actually dynamic routing can easily become worse than static. It takes time to collect information about network traffic. By the time routing tables are calculated, they are already obsolete; they are about some previous condition on the network, not the current one. Quantitative analysis shows that delays in building routing tables can affect dynamic routing performance unexpectedly strongly. This leads to the qualitative recommendation: "Trying to optimize too hard will make things worse. Dynamic routing should not try to adapt to traffic changes very fast." This hypothesis is accepted today and implemented in routing algorithms.