Dynamic Bandwidth Allocation in ATM Networks

Today's new applications such as World Wide Web, video conferencing and multimedia have introduced a large amount of traffic into the network. Additionally new applications are also heading towards real time process. Instant access to the network, greater level of performances and higher degree of satisfaction has become the main concerns of users using these new applications. Although current transmission mediums have advanced in capacity through means such as optical fiber and Gigabit Ethernet, future and unknown new services tend to consume up the available bandwidth. ATM network is the new technology used to support a wide variety of services including data, voice, video and most possibly other future applications. Its flexibility, efficiency and high throughput have gained popularity but with greater complexity due to different approaches in handling different type of services.A high-speed network such as ATM networks must have an effective traffic management scheme in order to gain high data throughput with the least cost of operation. Thus, simulation and modeling are the effective methods used to design the trade-off between network parameters and their performances. Effective sharing of network resources such as bandwidth and buffer are studied through the dynamic allocation method. Static allocation scheme has been proven inefficient to provide high resources utilization as can be seen in STM networks compared to A TM networks. However, ATM networks should provide different dynamic allocation methods according to its different services and traffic characteristics. Four dynamic allocation strategies have been designed, evaluated and compared for their performances. They are called Static Bandwidth Allocation, Bandwidth Allocated Proportional to Expected Queue Length, Bandwidth Allocated Proportional to Expected Queue Length with Threshold Value and Bandwidth Allocated with Threshold Interrupt. Bandwidth Allocated with Threshold Interrupt is proven to be the most effective strategy as it could response to congestion immediately