On Generality of the Data Plane and Scalability of the Control Plane in Software-Defined Networking

The control and data planes are decoupled in software-defined networking(SDN),which enables both planes to evolve independently,and brings about many advantages such as high flexibility,programmability,and rapid implementation of new network protocols.However,in order to improve the scalability of the control plane at present,some control functionalities are added to the data plane,which is probably to impact on the generality of the data plane.The key challenge of adding control functionalities to the data plane is to strike a careful balance between the generality of the data plane and the scalability of the control plane.We propose some basic principles that both control and data planes should comply with,based on the evolutionary trend of SDN.Moreover,we take two approaches for reference according to the principles,viewed from the control messages in OpenFlow-based SDN.Our evaluations demonstrate that the approaches can maintain the generality of the data plane and improve the scalability of the control plane.

Optimizing the Resource Utilization of Datacenter Networks with OpenFlow

Decreasing the flow completion time(FCT) and increasing the throughput are two fundamental targets in datacenter networks(DCNs), but current mechanisms mostly focus on one of the problems. In this paper, we propose OFMPC, an Open Flow based Multi Path Cooperation framework, to decrease FCT and increase the network throughput. OFMPC partitions the end-to-end transmission paths into two classes, which are low delay paths(LDPs) and high throughput paths(HTPs), respectively. Short flows are assigned to LDPs to avoid long queueing delay, while long flows are assigned to HTPs to guarantee their throughput. Meanwhile, a dynamic scheduling mechanism is presented to improve network efficiency. We evaluate OFMPC in Mininet emulator and a testbed, and the experimental results show that OFMPC can effectively decrease FCT. Besides, OFMPC also increases the throughput up to more than 84% of bisection bandwidth.

A quality requirements model and verification approach for system of systems based on description logic

System of systems engineering(So SE) involves the complex procedure of translating capability needs into the high-level requirements for system of systems(So S) and evaluating how the SoS quality requirements meet their capability needs. One of the key issues is to model the So S requirements and automate the verification procedure. To solve the problem of modeling and verification, meta-models are proposed to refine both functional and non-functional characteristics of the So S requirements. A domain-specific modeling language is defined by extending Unified Modeling Language(UML) class and association with fuzzy constructs to model the vague and uncertain concepts of the SoS quality requirements. The efficiency evaluation function of the cloud model is introduced to evaluate the efficiency of the SoS quality requirements. Then a concise algorithm transforms the fuzzy UML models into the description logic(DL) ontology so that the verification can be automated with a DL reasoner. This method implements modeling and verification of high-level So S quality requirements. A crisp case is used to facilitate and demonstrate the correctness and feasibility of this method.

Fast counting the cardinality of flows for big traffic over sliding windows

Counting the cardinality of flows for massive high-speed traffic over sliding windows is still a challenging work under time and space constrains, but plays a key role in many network applications, such as traffic management and routing optimization in software defined network. In this pa- per, we propose a novel data structure （called LRU-Sketch） to address the problem. The significant contributions are as follows. 1） The proposed data structure adapts a well-known probabilistic sketch to sliding window model; 2） By using the least-recently used （LRU） replacement policy, we design a highly time-efficient algorithm for timely forgetting stale information, which takes constant （O（1）） time per time slot; 3） Moreover, a further memory-reducing schema is given at a cost of very little loss of accuracy; 4） Comprehensive ex- periments, performed on two real IP trace files, confirm that the proposed schema attains high accuracy and high time efficiency.ferences including IEEE TPDS, ACM ToS, JCST, MIDDLEWARE, CLUSTER, NAS, etc. Currently, his research interests include big data management, cloud storage, and distributed file systems.

ON PAIRED PRIORITIZATIONS OF CRITERIA IN THE PERSPECTIVE OF DIAGRAPHS

It is not usually independent among criteria in multi-criteria decision making （MCDM）, and various dependences of criteria greatly influence the results of decision making. If an exact decision is desired, we must make clear the role of the dependences of criteria. Prioritizations, a new kind of dependences of criteria proposed recently, imply that the importance weights of criteria with lower priority for an alternative rely on whether the alternative satisfies the decision maker under criteria with higher priority. It has been validated that there exist lots of relevant applications in our daily activities. However, most existing literatures focus on how to deal with the problems of MCDM with ordered prioritizations among criteria （a special form of prioritizations）. The characteristics of prioritizations are not dug deep. This paper constructs a new form of prioritizations, called paired prioritizations, so as to reduce or even avoid imperfect rationality of decision makers hidden in the ordered prioritizations. We first represent binary paired prioritizations as a digraph, based on which we discover two kinds of imperfect rationality （inconsistency and incompleteness） produced in the period that the decision maker supplies the binary paired prioritizations. After the given paired prioritizations are consistent and complete, we develop an approach to transform the paired prioritizations to ordered prioritizations. The latter can be used to handle prioritized MCDM problems. Moreover, uncertainty, another kind of imperfect rationality, is considered when the decision maker provides the fuzzy paired prioritizations based on a set of linguistic labels. We construct a fuzzy digraph whose fuzzy relations are just the fuzzy paired prioritizations. The ordered prioritizations can then be derived with the aid of the fuzzy digraph. Two use cases are taken to show the process of transformations from binary/fuzzy paired prioritizations to ordered prioritizations.

Dimensionality reduction via kernel sparse representation

Dimensionality reduction （DR） methods based on sparse representation as one of the hottest research topics have achieved remarkable performance in many applications in recent years. However, it＇s a challenge for existing sparse representation based methods to solve nonlinear problem due to the limitations of seeking sparse representation of data in the original space. Motivated by kernel tricks, we proposed a new framework called empirical kernel sparse representation （EKSR） to solve nonlinear problem. In this framework, non- linear separable data are mapped into kernel space in which the nonlinear similarity can be captured, and then the data in kernel space is reconstructed by sparse representation to preserve the sparse structure, which is obtained by minimiz- ing a ~1 regularization-related objective function. EKSR pro- vides new insights into dimensionality reduction and extends two models： 1） empirical kernel sparsity preserving projec- tion （EKSPP）, which is a feature extraction method based on sparsity preserving projection （SPP）; 2） empirical kernel sparsity score （EKSS）, which is a feature selection method based on sparsity score （SS）. Both of the two methods can choose neighborhood automatically as the natural discrimi- native power of sparse representation. Compared with sev- eral existing approaches, the proposed framework can reduce computational complexity and be more convenient in prac- tice.

Multi-view dimensionality reduction via canonical random correlation analysis

Canonical correlation analysis （CCA） is one of the most well-known methods to extract features from multi- view data and has attracted much attention in recent years. However, classical CCA is unsupervised and does not take discriminant information into account. In this paper, we add discriminant information into CCA by using random cross- view correlations between within-class samples and propose a new method for multi-view dimensionality reduction called canonical random correlation analysis （RCA）. In RCA, two approaches for randomly generating cross-view correlation samples are developed on the basis of bootstrap technique. Furthermore, kernel RCA （KRCA） is proposed to extract nonlinear correlations between different views. Experiments on several multi-view data sets show the effectiveness of the proposed methods.

An OpenFlow-based performance-oriented multipath forwarding scheme in datacenters

Although dense interconnection datacenter networks(DCNs)(e.g.,Fat Tree) provide multiple paths and high bisection bandwidth for each server pair,the widely used single-path Transmission Control Protocol(TCP)and equal-cost multipath(ECMP) transport protocols cannot achieve high resource utilization due to poor resource excavation and allocation.In this paper,we present LESSOR,a performance-oriented multipath forwarding scheme to improve DCNs' resource utilization.By adopting an Open Flow-based centralized control mechanism,LESSOR computes near-optimal transmission path and bandwidth provision for each flow according to the global network view while maintaining nearly real-time network view with the performance-oriented flow observing mechanism.Deployments and comprehensive simulations show that LESSOR can efficiently improve the network throughput,which is higher than ECMP by 4.9%–38.3% under different loads.LESSOR also provides 2%–27.7% improvement of throughput compared with Hedera.Besides,LESSOR decreases the average flow completion time significantly.

An improved fully integrated,high-speed,dual-modulus divider

A fully integrated 2n/2n＋1 dual-modulus divider in GHz frequency range is presented. The improved structure can make all separated logic gates embed into correlative D flip-flops completely. In this way, the complex logic functions can be performed with a minimum number of devices and with maximum speed, so that lower power consumption and faster speed are obtained. In addition, the low-voltage bandgap reference needed by the frequency divider is specifically designed to provide a 1.0 V output. According to the design demand, the circuit is fabricated in 0.18 μm standard CMOS process, and the measured results show that its operating frequency range is 1.1- 2.5 GHz. The dual-modulus divider dissipates 1.1 mA from a 1.8 V power supply. The temperature coefficient of the reference voltage circuit is 8.3 ppm/℃ when the temperature varies from -40 to ＋ 125 ℃. By comparison, the dual-modulus divide designed in this paper can possess better performance and flexibility.