QoS Guarantee Technology In Packet Transport Network

With the exponential growth of data services in the network, the traditional with SDH technology-based bearer network needs to evolve to packet-switched network. The introduction of Packet Transport Network (PTN) makes the transport technology improve easily. PTN has the characteristics of IP networks which is fast and flexible. On the other hand, PTN retains the ability of transport multi-services, it can be reliable, scalable and manageable just like the conventional transport network. In order to provide better services, the traditional QoS (Quality of Service) guarantee technology must be improved to adapt the new network. In this paper, an in-depth study of PTN is taken first and then novel QoS guarantee technology in PTN is researched.The QoS strategies for PTN are critical, but some packet scheduling policies or models are not fair and efficient. A new packet scheduling method named buffer threshold based static priority (BTSP) is proposed. BTSP not only reduces the complexity of the static priority scheme by introducing buffer threshold but also provides fairness to packet services with different priorities. This paper presents a new algorithm for delay and bandwidth constrained path computation with traffic engineering objectives. It takes into account both the capacities and the utilization ratio on the links. The simulation result shows it has a good performance in the network supporting QoS guarantees. It can be widely adopted in PTN to ensure their QoS abilities and realize the load balancing in the network.It is important to keep the network working in synchronization when TDM services are transported across PTN. To provide TDM service QoS guarantee, service clock of the destination must be the same as the source. An improved adaptive clock recovery scheme based on packet arrival time interval is proposed in this paper. This scheme is designed to remove the stochastic jitter introduced by the PTN in the clock recovery process. Furthermore, this scheme is compared with the traditional clock recovery scheme in simulation. This method not only enables the destination clock to be synchronized with the source faster and more accurate but also avoids the buffer overflow or underflow.