http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Shedding Light on the Use of AS Relationships for Path Inference
Wenping Deng,Wolfgang Mühlbauer,Yuexiang Yang,Peidong Zhu,Xicheng Lu,Bernhard Plattner 한국통신학회 2012 Journal of communications and networks Vol.14 No.3
Autonomous system (AS) business relationships and their inference have been widely studied by network researchers in the past. An important application of inferred AS relationships can be the prediction of AS paths between a source and destination AS within a model. However, besides knowing the topology and inferred AS relationships, AS path prediction within a model needs to be understood in order for us to know how we can derive border gateway protocol (BGP) policies from AS relationships. In this paper, we shed light onto the predictive capabilities of AS relationships by investigating whether they can be translated into BGP policies such that inferred AS paths are consistent with real AS paths, e.g., paths observed from BGP routing tables. Our findings indicate that enforcing constraints such as the well-known valley-free property and the widely assumed preference of customer routes always results in a very low consistency for AS path inference. In addition, this is true irrespective of whether customer,peer, or provider routes are preferred. Apparently, applying such constraints eliminates many “correct” paths that are observed in BGP routing tables and that are propagated in a simple shortest path model where AS relationships are ignored. According to our findings, deriving BGP routing policies for predicting with high accuracy AS paths in a model directly from AS relationships is still difficult.
Shedding Light on the Use of AS Relationships for Path Inference
Deng, Wenping,Muhlbauer, Wolfgang,Yang, Yuexiang,Zhu, Peidong,Lu, Xicheng,Plattner, Bernhard The Korea Institute of Information and Commucation 2012 Journal of communications and networks Vol.14 No.3
Autonomous system (AS) business relationships and their inference have been widely studied by network researchers in the past. An important application of inferred AS relationships can be the prediction of AS paths between a source and destination AS within a model. However, besides knowing the topology and inferred AS relationships, AS path prediction within a model needs to be understood in order for us to know how we can derive border gateway protocol (BGP) policies from AS relationships. In this paper, we shed light onto the predictive capabilities of AS relationships by investigating whether they can be translated into BGP policies such that inferred AS paths are consistent with real AS paths, e.g., paths observed from BGP routing tables. Our findings indicate that enforcing constraints such as the well-known valley-free property and the widely assumed preference of customer routes always results in a very low consistency for AS path inference. In addition, this is true irrespective of whether customer, peer, or provider routes are preferred. Apparently, applying such constraints eliminates many "correct" paths that are observed in BGP routing tables and that are propagated in a simple shortest path model where AS relationships are ignored. According to our findings, deriving BGP routing policies for predicting with high accuracy AS paths in a model directly from AS relationships is still difficult.
Community Detection in Complex Networks based on Improved Genetic Algorithm and Local Optimization
Kun Deng,XingYan Liu,WenPing Li 보안공학연구지원센터 2016 International Journal of Grid and Distributed Comp Vol.9 No.10
This paper proposes the community detection in complex networks based on improved genetic algorithm and local optimization (IGALO) in terms of the defect that traditional community detection approaches based on genetic algorithm have strong randomness and weak searching ability in the process of community detection. Taking modularity function Q as the objective function, IGALO algorithm adopts label propagation method of one-iteration to initialize population so as to generate initial population with certain precision. Then, anti-destructive one-way crossover strategy is proposed to ensure the crossover operation to develop in the direction of making community structure increase to modularity function. Finally, mutation strategy of node local optimization is proposed to improve the searching efficiency of algorithm. This algorithm effectively overcomes the defect that traditional algorithms have weak searching ability and improves the community detection accuracy. Tests are made on benchmark networks and real-world networks and comparative analysis is also made with various classic algorithms. The results show that IGALO algorithm is effective and feasible.