Benxiong Huang

Cyber-Individual Meets Brain Informatics

To help people live better in todays digitally explosive environment, the authors envision a Cyber-Individual (Cyber-I) that is the counterpart of a real individual in the physical world.

Bandwidth-Aware Scheduling With SDN in Hadoop: A New Trend for Big Data

Software-defined networking (SDN)xa0is a revolutionary network architecture that separates out network control functions from the underlying equipment and is an increasing trend to help enterprises build more manageable data centers where big data processing emerges as an important part of applications. To concurrently process large-scale data, MapReduce with an open-source implementation named Hadoop is proposed. In practical Hadoop systems, one kind of issue that vitally impacts the overall performance is known as the NP-complete minimum make span11Make span means the time between job’s start time and job’s finish time. problem. One main solution is to assign tasks on data local nodes to avoid link occupation since network bandwidth is a scarce resource. Many methodologies for enhancing data locality are proposed such as the Hadoop default scheduler (HDS)xa0and state-of-the-art scheduler balance-reduce scheduler (BAR). However, all of them either ignore allocating tasks in a global view or disregard available bandwidth as the basis for scheduling. In this paper, we propose a heuristic bandwidth-aware task scheduler bandwidth-aware scheduling with SDN in Hadoop (BASS)xa0to combine Hadoop with SDN. It is not only able to guarantee data locality in a global view but also can efficiently assign tasks in an optimized way. Both examples and experiments demonstrate that BASS has the best performance in terms of job completion time. To our knowledge, BASS is the first to exploit talent of SDN for job scheduling of big data processing and we believe that it points out a new trend for large-scale data processing.

Cyber-I: Vision of the Individual's Counterpart on Cyberspace

Cyber-Individual, with a short term Cyber-I, is a real individuals counterpart in cyberspace. It is closely related to human-centric computing ideology which focuses on placing human in the center of computing. The study on Cyber-I tries to re-examine and analyze human essence in the digital era. Cyber-Is vision is to create a unique, digital, comprehensive description for every real person being in the cyberspace. Humans social context, mood, temper, physical status and so on also need to be considered for such a full description. Further research on social computing, antropology, human behaviour study, psycology and other fields/disciplines are required to enrich Cyber-I concept, meanwhile, Cyber-I will also raise new problems to these fields/disciplines. The IT technologies on the whole including ubiquitous computing, pervasive sensors, wired/wireless networks and clouds will bring Cyber-I vision into practice. In this paper, we first present the Cyber-I concept, its important characteristics and basic architecture. Then, we discuss the special features of Cyber-I as compared with other related concepts and studies including AR (Augmented Reality), HCI (Human Computer Interaction), AI, artificial life, etc. The Cyber-I layered architecture and basic elements are described in detail, and the associations between Cyber-I and the corresponding real individual are explained as well. Finally, fundamental problems and challenging issues brought by Cyber-I are addressed in terms of necessary technique, security, privacy, ethic, philosophy, etc.

Web Search Results Clustering Based on a Novel Suffix Tree Structure

Web search results clustering are navigator for users to search needed results. With suffix tree clustering (STC), search results can be clustered fast, automatically, and each cluster is labeled with a common phrase. Due to the large memory requirement of suffix tree, some other approaches have been proposed, with lower memory requirement. But unlike other algorithms, STC is an incremental algorithm and a promising approach to work on a long list of snippets returned by search engines. In this paper we proposed an approach for web search results clustering and labeling, based on a new suffix tree data structure. The approach is an incremental and linear time algorithm, with significantly lower memory requirements. This approach also labels every final cluster a common phrase, thus it is suitable for quickly browsing by users. Experimental results show that the new approach has better performance than that of conventional web search result clustering.

Bandwidth-aware energy efficient flow scheduling with SDN in data center networks

Abstract Nowadays the energy consumption has become one of the most urgent issues for Data center networks. For general network devices, the power is constant and independent from the actual transfer rate. Therefore the network devices are energy efficient when they are in full workload. The flow scheduling methods based on the exclusive routing can reduce the network energy consumption, as the exclusive routing paths can fully utilize all their links. However, these methods will no longer guarantee the energy efficiency of switches, as they handle flows in priority order by greedily choosing the path of available links instantaneously. In a previous work we proposed an extreme case of flow scheduling based on both link and switch utilization. Herein we consider general scenarios in data center networks and propose a novel energy efficient flow scheduling and routing algorithm in SDN. This method minimizes the overall energy for data center traffic in time dimension, and increases the utilization of switches and meet the flow requirements such as deadline. We did a series of simulation studies in the INET framework of OMNet++. The experiment results show that our algorithm can reduce the overall energy with respect to the traffic volume and reduce the flow completion time on average.

High Throughput Implementation of MD5 Algorithm on GPU

Graphics Processing Unit (GPU) has evolved into a highly parallel, multithreaded, many-core processor with tremendous computational capability. The introduction of Compute Unified Device Architecture (CUDA) simplifies the software development on GPU and allows direct access to GPU resources. It’s an effective way to improve the hashing performance in high-speed network and storage systems by using GPU as a coprocessor of CPU to execute the hash encryption algorithm. This paper puts forward a CUDA-based design of the MD5 hash algorithm on GPU according to the specific application needs and presents its implementation as well as its comprehensive optimization in terms of the characteristics of GPU and CUDA. Keywords-GPU; CUDA; MD5; Hash Function

Preventing DDoS Attacks Based on Credit Model for P2P Streaming System

Distributed Denial of Service (DDoS) attack is a serious threat to the Internet communications especially to P2P streaming system. P2P streaming system is vulnerable to DDoS attacks due to its high bandwidth demand and strict time requirement. In this paper, we propose a distributed framework to defense DDoS attack based on Credit Model(CM) which takes the responsibility to identify malicious nodes and categorize nodes into different credit level. We also introduce a Message Rate Controlling Model (MRCM)to control the message rate of a node according to its credit level. Combining CM and MRCS together, our framework can improve the resistibility against DDoS for P2P streaming system.

Password-Based Access Control Scheme with Remote User Authentication Using Smart Cards

In this paper, a bilinear mapping based password-based access control scheme with remote user authentication scheme using smart cards is presented. The proposed scheme enables one user to choose his password freely in the registration phase and easily change it as needed. We take the novel technique of binding one users private key to his identity, which enables the AS (authentication server) to be unnecessary to maintain a password table to verify the login request and thus enhances the practicality and reliability of the system. Additionally, based on the difficulty of Elliptic Curve Discrete Logarithm (ECDL) problem, the proposed scheme can efficiently withstand replaying attacks, forgery attacks, Masquerade attacks, guessing and stolen verifier attacks and insider attacks.

Message Receiver Determination in Multiple Simultaneous IM Conversations

Automatically detecting instant message receivers by applying semantic relevance can help improve the usability of current IM systems and make multiple, simultaneous IM conversations more convenient.

A malicious behavior analysis based Cyber-I birth

Cyber-Individual (Cyber-I) is the digital counterpart of an individual in the real world, which aims at systematically studying and developing comprehensive individual human modeling and its associated applications. The ultimate goal of this research is to create a digital clone for each individual and to provide active desirable services. We present a part of our research work focusing on examining the basic system architecture and the birth process of Cyber-I from a security perspective. In this study, a customized honeypot is used to record multidimensional data Cyber-I is constructed for a corresponding invader. Further, assembling a Cyber-I with associated CI-Applications enables aninvader having more behaviors in the honeypot and provides a possible chance to prolong activities of the invader, which complements a loop mechanism to feed Cyber-I for its growth. The preliminary result in this paper reveals that appropriate authorization and controls are extremely necessary to prevent Cyber-I from being maliciously used and to ensure privacy of its real individual in building an open Cyber-I platform.