A Queuing Model for Congestion Control and Reliable Data Transfer in Cable Access Networks

A set of simple queuing theory models which can model the average response of a network of computers to a given traffic load has been implemented. The impact of variations in traffic patterns and intensities, channel capacities, and message protocols can be assessed using them because of the lack of fine detail in the network traffic rates, traffic patterns, and the hardware used to implement the networks. In order to access the networks it’s a two stage sequential procedure for data transfer from a node at customer’s premises to a central node. The stations which have data to transmit request will undergo contention stage and requests that are successfully received by the central node enter into the second stage. All the data will reside in a queue until they are scheduled by a centralized scheduler. Once the stages are set properly and all the data is queued accordingly then a congestion free communication will happen in the cable access networks. To establish a congestion free as well as reliable data transfer in cable access networks we are working with a access network which is hybridized with queuing models which has the capacity to maintain the data packets in a scheduled manner. We test on static as well as distributed (random) data which will arrive at the centralized node. From then its up to mathematical queuing model to route the data reliably without any collision free. The performance measures and various data transfer patterns are observed in this work.