Implementing beneficial prioritization of traffic flow in complex networks

Abstract

Despite the reality of traffic prioritization in real networks as the internet, this has almost been ignored in designed routing protocols. In this work, we propose a priority policy where packets are prioritized according to their destination. This procedure is applied to a fraction f of nodes; packets are classified as High Priority if their destinations are among the fraction f, otherwise the packets are treated as Low Priority. Using numerical simulation we found that the prioritization of nodes with high degree (hubs) is always more efficient than the prioritization of nodes with small degree or the random prioritization of nodes. Moreover, we observed three regimes based on the prioritized fraction f in the network: the first one is characterized by an improvement of the High Priority Traffic (HPT) flow without any degradation of the flow of the Low Priority Traffic (LPT). In the second regime, the HPT gains some performance at the expense of a loss of the performance in the LPT flow. While in the last regime the LPT experiences a low performance without any noticeable improvement in HPT compared to the normal flow.

Efficient edge rewiring strategies for enhancement in network capacity

“…Jiang et al [7] improved the network capacity by removing the connection between two nodes with the maximum sum of the degree together with betweenness centrality (BC) and by adding a connection between two unconnected nodes with the minimum sum of the degree in addition to BC. Alweimine et al [9] prioritized the packets according to the destination node and found that the prioritization of nodes with hubs is always more efficient than the prioritization of nodes with a small degree or a random prioritization. Ma et al [10] worked on a two-layer network model and proposed a delivery capacity allocation strategy based on the degree distributions of both the layers.…”