Researchers develop intelligent segment routing scheme for network management

Site visitors engineering has attracted a lot analysis consideration, particularly in recent times as networks develop in dimension and complexity. Community operators more and more want higher methods to handle the huge quantities of information flowing by means of their networks. A staff of researchers has proposed an clever routing scheme for site visitors engineering that achieves load balancing with restricted management overheads.

Their analysis is printed within the journal Massive Knowledge Mining and Analytics.

Historically, researchers have studied site visitors engineering associated to conventional web protocol (IP) networks with a give attention to IP routing protocols, routing optimization issues, and overlaying in an IP community. With the arrival of the software-defined community, researchers started to focus extra on site visitors engineering points, together with site visitors splitting and protocol design.

In software-defined networks, the community might be centrally managed utilizing software program purposes. Software program-defined networks allowed researchers to realize extra environment friendly community administration, fixing among the large site visitors engineering points which might be troublesome to handle in conventional networks.

But even with software-defined networks, researchers struggled with scalability points. So researchers turned their consideration to phase routing. Phase routing is a way that permits researchers to simplify site visitors engineering throughout community domains by organizing collections of knowledge referred to as packets.

Researchers realized that by combining phase routing with software-defined networks, they may very well be able to fixing among the challenges within the software-defined community. Nonetheless, there have been nonetheless some unresolved points as a result of phase routing brings with it management overheads, that means further packet headers should be inserted. The overheads vastly cut back a big community’s effectivity when the phase headers grow to be too lengthy.

“Phase routing has been a novel structure for site visitors engineering, nevertheless it additionally brings management overheads and reduces forwarding effectivity. So we targeted on the best way to optimize the hyperlink load balancing efficiency with restricted management overhead primarily based on phase routing,” mentioned Laizhong Cui, a professor with the Faculty of Laptop Science and Software program Engineering at Shenzhen College.

To beat these challenges, the analysis staff proposed an clever routing scheme for site visitors engineering. This enables for load balancing with restricted management overheads. The staff began by formulating the issue as a mapping downside that maps totally different flows to key diversion factors. Subsequent they proved the issue is nondeterministic polynomial arduous, a method of defining the issue in computational complexity principle.

Then to unravel the issue, they developed an improved ant colony optimization algorithm. Ant colony optimization is a way that makes use of chance in fixing community optimization issues. In addition they designed a second algorithm, a load balancing algorithm, and so they analyzed its theoretical efficiency.

“We proposed two algorithms to understand our load balancing goal and keep away from forwarding overload. The idea of ant colony optimization and linear programming offered concepts and instructions for the algorithms,” mentioned Laizhong Cui.

The staff evaluated their clever routing scheme for site visitors engineering in several real-world topologies. Topology describes how the weather of networks are organized and linked. The staff’s outcomes present that their algorithms outperform conventional algorithms. With their clever routing scheme for site visitors engineering, the utmost bandwidth is 24.6 p.c decrease than that of conventional algorithms, when evaluated on the Bell Canada community topology.

Looking forward to future analysis, the staff is getting ready to check and optimize their algorithms in an actual community atmosphere. In addition they plan to additional develop their scheme by including the tactic of synthetic intelligence in software-defined large space networks. “Our final purpose is to develop and apply our options to most community architectures to enhance community transmission efficiency,” mentioned Cui.

Supplied by
Tsinghua College Press