International Journal of Computer Applications |
Foundation of Computer Science (FCS), NY, USA |
Volume 13 - Number 7 |
Year of Publication: 2011 |
Authors: Kuda Nageswara Rao, K.Srinivas rao, P.Srinivasa Rao |
10.5120/1794-2485 |
Kuda Nageswara Rao, K.Srinivas rao, P.Srinivasa Rao . Transient Analysis of a Tandem Communication Network with Dynamic Bandwidth Allocation having Two Stage Direct Bulk Arrivals. International Journal of Computer Applications. 13, 7 ( January 2011), 14-22. DOI=10.5120/1794-2485
Communication network models play a dominant role in optimal utilization of resources. The need for congestion control and to improve quality of service, it is needed to design new communication networks. With this motivation, a two node communication network with dynamic bandwidth allocation (DBA) having two stage bulk arrivals (BA) is introduced and analyzed. The statistical multiplexing of the communication network is developed by characterizing the arrival of packets with compound Poisson process and the transmission completions with Poisson processes. To avoid burstness in buffers, the dynamic bandwidth allocation is adapted utilizing the idle bandwidth in the transmitter depending on content of the buffer. The network performance is evaluated by deriving the performance measures like mean content of the buffers, mean delays in transmitters, throughput of the nodes, utilization of the nodes etc,. The transient analysis of the model reveals that the time has a significant influence on predicting the performance measures of the communication networks. The sensitivity analysis of the model provides evidence that the dynamic bandwidth allocation strategy along with bulk arrival consideration can provide the performance of the network more close to the reality and avoid burstness in buffers. This communication network is much useful for Satellite and Mobile communication, Computer communication etc. This communication network model also includes several of the earlier communication network models as particular cases for specific values of the model parameters.