Performance evaluation of an on demand tunable resource allocation apparatus using constrained active network

Abstract

Resource allocation strategy must offer a robust way for the best assignment of resources to network while at the same time enabling equal share of network resources. This work proposes an active network conceptual model suited for the network resource allocation problem. This constrained active network allocates network resources to each demand performing capacity allocation and bandwidth reservation. Swarm-based active packets continuously communicate with active nodes by using the Split Agent Routing Technique (SART) and inform active nodes about available resources on previously visited node(s) and links. This apparatus enables adaptation of the system to new conditions (bandwidth reservation/capacity allocation). It also enables additional information to be passed to neighboring nodes for which information is embodied in transmitted packets. Thorough examination is made, for the performance of the proposed scheme in the network and the QoS offered, taking into account a number of performance metrics. Simulation results show that this swarm-based active network scheme offers a decentralized control and an efficient way to increase overall performance enabling at the same time optimized throughput.