Software-Defined Networking proposes to fundamentally change the current practice of network control. The two basic ideas are Centralized State Control and Uniform Device Abstraction, which support... Show More

Software-Defined Networking proposes to fundamentally change the current practice of network control. The two basic ideas are Centralized State Control and Uniform Device Abstraction, which support the Software-Defined promise. SDN has made significant progress. The opportunities of SDN in carrier access networks have been largely ignored by both industry and academia. In access networks, Quality-of-Service (QoS) oriented bandwidth management is more critical; the flexible QoS provisioning could be the most important opportunity for SDN. In this position paper, the authors show that the unique characteristics of access networks pose significant challenges to the two basic ideas. Contrary to the common agreement on “match-action” abstraction, the authors argue that the object-oriented abstraction might be a better choice for access networks to make a better software-defined implementation.

Forming a virtual backbone has been always a significant technique for ad hoc networks to achieve high efficiency in various applications, including mobility management, broadcast based information... Show More

Forming a virtual backbone has been always a significant technique for ad hoc networks to achieve high efficiency in various applications, including mobility management, broadcast based information dissemination, etc. This is especially true for Vehicular Ad Hoc Network (VANET), because most of VANET applications rely on broadcasting messages. Among others, constructing the Connected Dominating Set (CDS) is a popular approach to forming virtual backbone in ad hoc networks, including VANETs. Although quite a lot of work has been done on CDS algorithms for ad hoc networks, almost all existing algorithms focus on minimizing the size of CDS, i.e. the number of nodes in CDS. In this paper, the authors consider to improve the stability of CDS, i.e. to reduce the changes of CDS during the runtime of a VANET. By making use of vehicle navigation route, the authors propose a new metric to selecting CDS node with high stability. Then, based on the new metric the authors design a CDS algorithm, which can construct a CDS with higher stability than the CDS by existing algorithms. Such advantage is validated via extensive simulations.

Quality of service (QoS) optimization for end-to-end (e2e) services always depends on performance analysis in cloud-based service delivery industry. However, performance analysis of e2e services... Show More

Quality of service (QoS) optimization for end-to-end (e2e) services always depends on performance analysis in cloud-based service delivery industry. However, performance analysis of e2e services becomes difficult as the scale and complexity of virtualized computing environments increase. In this paper, the authors present a novel hierarchical stochastic approach to evaluate the QoS of e2e virtualized cloud services using Quasi-Birth Death structures, where jobs arrive according to a stochastic process and request virtual machines (VMs), which are specified in terms of resources, i.e., VM-configuration. To reduce the complexity of performance evaluation, the overall virtualized cloud services are partitioned into three sub-hierarchies. The authors analyze each individual sub-hierarchy using stochastic queueing approach. Thus, the key performance metrics of e2e cloud service QoS, such as acceptance probability and e2e response delay incurred on user requests, are obtained.

Analyzing attack graphs can provide network security hardening strategies for administrators. Concerning the problems of high time complexity and costly hardening strategies in previous methods, a... Show More

Analyzing attack graphs can provide network security hardening strategies for administrators. Concerning the problems of high time complexity and costly hardening strategies in previous methods, a method for generating low cost network security hardening strategies is proposed based on attack graphs. The authors' method assesses risks of attack paths according to path length and the common vulnerability scoring system, limits search scope with a threshold to reduce the time complexity, and lowers cost of hardening strategies by using a heuristic algorithm. The experimental results show that the authors' method has good scalability, and significantly reduces cost of network security hardening strategies with reasonable running time.

The number of services on the Internet is growing rapidly. Thus, the problem of selecting proper services for most users becomes serious and service recommendation is widely needed. Besides... Show More

The number of services on the Internet is growing rapidly. Thus, the problem of selecting proper services for most users becomes serious and service recommendation is widely needed. Besides functions, QoS information is also an important factor to be considered when making recommendations to users. However, QoS changes with time. To address and solve these challenges, this paper proposes a temporal-aware QoS-based service recommendation framework, and also comes up with a prediction algorithm based on Tucker decomposition. Moreover, the authors use real-world datasets to verify our method with results better than traditional methods.