Study on a self-similar traffic shaping mechanism with QoS in transport networks
Mingxia Bo, Peiyuan Lee, Xiaofei Pan, Wanyi Gu
Abstract
Mingxia Bo, Peiyuan Lee, Xiaofei Pan, Wanyi Gu
Abstract
Due to easy realization and high bandwidth utilization, SDH/WDM technology becomes the important way to carry IP traffic over the backbone network. On the other hand, the feature of the data traffic which is much different from the voice traffic is dynamic, burst and self-similar, and many proofs show that the self-similar traffic can lead to some adverse effects on the network performance due to the property of long-range dependence (LRD). For this reason it is widely recognized that self-similarity of the traffic is a significant problem as far as network engineering is concerned. So any reduction in the degree of self-similarity will be greatly beneficial. One possible strategy for mitigating the deleterious effects of the self-similarity is to reduce the burstiness of the input traffic through traffic shaping function at the edge nodes. According to this scheme, in this paper, we present a new self-similar traffic shaping mechanism with QoS in transport networks, called double threshold algorithm (DTA). Simulation results show that the proposed mechanism can effectively reduce the degree of input self-similar traffic, and performs better in the terms of network packet-loss rate and blocking probability than the non-traffic shaping schemes. At the same time it guarantees good quality of service.
A significance statement is not available in the OpenAlex record.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Due to easy realization and high bandwidth utilization, SDH/WDM technology becomes the important way to carry IP traffic over the backbone network. On the other hand, the feature of the data traffic which is much different from the voice traffic is dynamic, burst and self-similar, and many proofs show that the self-similar traffic can lead to some adverse effects on the network performance due to the property of long-range dependence (LRD). For this reason it is widely recognized that self-similarity of the traffic is a significant problem as far as network engineering is concerned. So any reduction in the degree of self-similarity will be greatly beneficial. One possible strategy for mitigating the deleterious effects of the self-similarity is to reduce the burstiness of the input traffic through traffic shaping function at the edge nodes. According to this scheme, in this paper, we present a new self-similar traffic shaping mechanism with QoS in transport networks, called double threshold algorithm (DTA). Simulation results show that the proposed mechanism can effectively reduce the degree of input self-similar traffic, and performs better in the terms of network packet-loss rate and blocking probability than the non-traffic shaping schemes. At the same time it guarantees good quality of service.
Key concepts: Burstiness, Computer science, Traffic shaping, Self-similarity, Computer network, Quality of service, Traffic generation model, Cellular traffic