Abstract:To solve the dynamic preemptive scheduling problem of a fuzzy uncertain task set under an unpredictable environment, a scheduling algorithm, by applying fuzzy rules and fuzzy scheduling theory, is presented based on fuzzy feedback control, and its corresponding scheduling structure, composed of a basic scheduler and a fuzzy feedback controller, is established. The fuzzy scheduling algorithm is used as the scheduling algorithm of the basic scheduler. By classifying tasks into different priority levels and scheduling the task with high priority level first, more critical tasks can be scheduled. A fuzzy controller constitutes the part of the fuzzy feedback control with the adjusting policy of task flow. Simulation results show that the proposed fuzzy feedback-control scheduling can effectively control the missed deadline rate of tasks, the scheduling problem with fuzzy characteristics or unpredictable environment can be solved, and the successful scheduling rate of critical tasks can be improved.

Abstract:Schedulability test is an essential issue in real-time scheduling theory. Rate monotonic(RM) algorithm is one of the scheduling algorithms in real-time systems However, a general review on this field can hardly be seen. This paper presents a review of the various schedulability tests under RM algorithm, starting from the simplest ideal RM scheduling model and then going into the more complicated ones. Three subjects are covered: (1) schedulable tasks?CPU utilization least bound and the sufficient and necessary conditions; (2) condition of schedulable tasks taking the scheduling time cost into account; (3) condition of schedulable tasks taking priority inversion into account. Some appropriate examples are provided to illustrate and compare the relative merits among the discussed algorithms.

Abstract:Static priority scheduling is widely used in real-time systems. But its schedulability will be reduced if priority levels of the system are insufficient. A task set may require more priority levels than the system can support. In this case, more than one task must be grouped into the same priority. This paper presents necessary and sufficient conditions for analyzing the schedulability of static priority algorithms on resources with limited priority levels. A static priority assignment algorithm (AGP) with limited priority levels is developed. As it turns out, AGP is optimal for the basic task set in the sense that the number of priority levels required by AGP is minimal and no other static priority rule can schedule a basic task set which cannot scheduled by AGP. Simulation results show that the schedulability of AGP is much higher than that of Constant Ratio Grid algorithm. AGP is significant for solving the problem of assigning priorities of tasks in embedded real-time systems.

Abstract:Distributed control systems are a category of high complex systems that include a large number of devices controlled and harmonized by computer systems. Their reliability and functional correctness always need to be guaranteed as their mission-critical feature. The analysis process for complex control systems consists of proving or verifying that the designed system indeed meets certain specifications. However, both the design and analysis may be formidable due to the complexity and magnitude of the system. From an analysis perspective, the complexity of a system can be reduced by imposing a hierarchical structure and abstraction on the architectural design. Currently, model checking has been demonstrated by more and more successes. It is an effective way to verify that the construction of a complex system satisfies to the requirements of reliability and correctness. In this paper, an approach for formally analyzing distributed control systems at architectural level by applying software architecture description and model checking techniques is presented. Through study on a building comprehensive control system, it is shown that the method could improve the quality of design of distributed control systems.

Abstract:Software overhead in interconnection network communication has currently become the bottleneck of a cluster system. To reduce it, a user-level communication software UMPS based on real-time OS RTLinux is designed and implemented, which is comfortable with VIA. A new concept of semi-polling driven is presented. With the semi-polling driven mechanism, the interrupts frequency is lowered and the processing performance for short message is significantly ameliorated. By means of the address translation and buffer managing algorithm based on the resource-mapping graph, applications bypass OS and interact with network interface directly using asynchronous DMA. So the overhead and latency in communication are efficiently reduced. Experimental results indicate that the throughputs of UMPS for 64 byte and 1500 byte messages are 394 Mbps and 895 Mbps respectively, and the performance of UMPS surpasses that of other mechanisms.

Abstract:Two parallel programming models of shared-memory and message-passing are widely adopted. The programmability of message-passing is poor, while that of shared-memory is good. The OpenMP Application Programming Interface is an emerging standard for shared-memory. OpenMP on cluster supplies an OpenMP computing environment on cluster of workstations or PCs, which combines the friendly programmability of shared-memory with the fine scalability of cluster. Taking 7 well-known parallel applications on a cluster of PCs, this paper compares the performance of OpenMP/JIAJIA, an OpenMP system on cluster, with that of MPI, a typical message passing system. Experimental results show that the performance of OpenMP is averagely equal to 81% of MPI for the 7 applications running on 8-nodes, but the former is easier to use than the latter.

Abstract:It's impossible for traditional operating systems to separate the abstracts for data storage (process virtual address space), for data computation (thread), and for resource management (process itself). This paper first analyzes the problems due to not able to separat these three abstracts. On the base of the analysis, the idea that the three abstracts should be separated is proposed, and then, the operating systems based on virtual address spaces on files (OS-BVASF) is developed. Then, OS-BVASF抯 architecture model thread migration and instruction accessing file that implement separation of the three abstracts is investigated. Finally, its implementation, test, and performance evaluation are discussed. The work in this paper shows it is feasible to separate data storage, data computation, and resource management in operating systems.

Abstract:Fuzzy partitional clustering algorithms are widely used in pattern recognition field. Until now, more and more research results on them have been developed in the literature. In order to study these algorithms systematically and deeply, they are reviewed in this paper based on c-means algorithm, from metrics, entropy, and constraints on membership function or cluster centers. Moreover, the advantages and disadvantages of the typical fuzzy partitional algorithms are discussed. It is pointed out that the standard FCM algorithm is robust to the scaling transformation of dataset, while others are sensitive to such transformation. Such conclusion is experimentally verified when implementing the standard FCM and the maximum entropy clustering algorithm. Finally, the problems existing in these algorithms and the prospects of the fuzzy partitional algorithms are discussed.

Abstract:A palmprint is a relative new biometric feature for personal authentication. Palm-lines, including the principal lines and wrinkles, are one of the most important features used in palmprint recognition. This paper proposes a novel approach of line feature representation and matching for palmprint recognition. To represent palm-lines, a vector, called line feature vector (LFV), is defined by using the magnitude and orientation of the gradient of the points on these lines. A LFV contains information about both the structure and thickness of the lines, thus its capability to distinguish between palmprints, including those with similar line structures, is strong. A correlation coefficient is employed to measure the similarity between LFVs of palmprints during the matching phase. 99.0% and 97.5% accurate rates are obtained in the one-to-one matching test and one-to-many matching test, respectively. The results show that LFV is robust to some extent in rotation and translation of the images. The accuracy, speed and storage of the proposed approach can meet the requirements of an online biometric recognition.

Abstract:The choices of discrete time step for Euler method and trapezoidal method and terminating condition of iteration in trapezoidal method are discussed in this paper for numerical implementation of continuous time Hopfield network. The decreasing conditions of an energy function are investigated by the use of convex function. By utilization of the primal convex function, the conditions are analyzed under which its conjugate function minus a quadratic function is also convex. Based on the analysis of the proof for convergence of the continuous time Hopfield network model, a generalized model is proposed. For the common Euler and trapezoidal methods, the choice of their discrete time step is discussed for numerical implementation of the continuous time Hopfield network. As the trapezoidal method is an implicit scheme, its realization needs an iterated procedure. The conditions to terminate the iterated procedure are analyzed. According to the special forms of the continuous time Hopfield network model, an improved iterated algorithm for trapezoidal method is proposed and analyzed. The numerical results show that choosing a suitably large discrete time step will be helpful not only to accelerate the numerical implementation but also to improve the optimization performance.

Abstract:A model for detecting salient regions in an image based on location shift and extent trace is proposed in this paper. In this model, the detection is divided into two stages: the first one is location shift at which the location of a new salient region is found according to the global saliency measurement, and the second is extent trace at which the size of the salient region is chosen according to the local saliency measurement. In this way, a series of salient regions are detected through the alternation of location shift and extent trace. Based on this model, a novel algorithm is presented. The results of experiments with many real images show the algorithm is effective and efficient.

Abstract:Concerning the features of complex objects and massive data transmission, a new XML-based method to design and implement communication protocols for WebGIS is presented. With the aid of UML, the typical requiring and responding protocols of WebGIS are analyzed through object-oriented concept. Based on object oriented analysis of the protocols, the mechanism of designing communication protocols following W3Cs XML Schema specification is illustrated. Finally, the main flow of embedding the protocols into WebGIS is given by packing and parsing XML-based protocols in a WebGIS application prototype. This kind of protocols can be used in spatial information exchange among heterogeneous WebGIS platforms in distributed environment.

Abstract:As mobile hosts dynamically change their locations, which makes the multicast transmission tree to be rebuilt frequently, the existing reliable multicast algorithms are unsuitable for mobile hosts in wireless IP networks. In this paper, a new basic mobile multicast mechanism called previous network subscription is proposed. Then, a new reliable multicast algorithm supporting mobile hosts is given. Based on hierarchical architecture, this algorithm integrates previous network subscription and remote subscription. An ACK and NAK mixed acknowledgement mechanism is adopted, the lost packets are retransmitted by multicasting, and the NAK suppression is used in the subnet. The analysis and simulation results show that the algorithm has a good performance in terms of retransmission delay, retransmission cost, multicast service disruption, and protocol cost.

Abstract:Peer-to-Peer systems have shown a great potential on file sharing in recent years, and developing efficient search techniques has become a crucial research problem. Most unstructured peer-to-peer systems with existing distributed routing algorithms can only run blind search, but not global search due to the lack of cache scheme. To address the problem, a novel key clustering algorithm is proposed, which divides the routing space into two layers, AUT layer and HUB layer. Such an algorithm can perform a well-ordered search from a global view. In order to improve scalability, theoretical results from small-world are utilized. In the improved algorithm, a few shortcuts with distant peers are inserted into the routing tables with some probabilities, and the average path length is reduced. The preliminary simulation results show that the key clustering algorithm with shortcuts is efficient and scalable.

Abstract:Dynamic peer communication is the most complicated of all the group communication mode. The paper concisely analysed five typical group key agreement protocols put forward in recent years, they are: centralized key distribution (CKD), burmester-desmedt (BD), steer et al. (STR), group diffie-hellman (GDH) and tree-based group diffie-hellman (TGDH). A new group key agreement protocol (TTS) is proposed and compared with other protocols. This protocol has the advantage in computation, and fit to common network condition.

Abstract:To support different QoS requirements in terms of bandwidth and delay constraints imposed by heterogeneous and dynamic joining receivers, this paper proposes an algorithm named QDMR-LD (QoS-based dynamic multicast routing for streaming layered data). When a new receiver joins, receiver-oriented path searching heuristic is used to find a feasible path with minimum cost from the multicast tree to the receiver. RBMF (reverse best metric forwarding) mode proposed in our previous work is adopted in this paper to increase the joining success ratio. When a receiver leaves, corresponding part of the multicast tree is pruned. Simulation results show that QDMR-LD increases the success ratio and lowers the multicast tree cost compared with other related schemes.

Abstract:The ever-increasing multicast data applications recently have aroused considerable interests in the design of congestion control scheme for multicast services. This kind of study is indeed important, especially to those multicast receivers with large propagation delays which mean the feedbacks arriving at the source node are somewhat outdated and harmful to control actions. A distributed self-tuning explicit rate algorithm is presented in this paper to overcome the vulnerability that suffers from the heterogeneous multicast receivers. It is suggested that congestion controllers be located at the source and the participating intermediate nodes to regulate the transmission rate. This network-assisted property is different from the traditional control scheme in that the router computes the appropriate transmission rate of itself and executes it rather than sends packets in best efforts. This active manner makes the control more responsive to the network status. The proposed self-tuning controller has essentially a proportional controller structure. The proportional gain is related to the extent that the router buffer occupancy deviates from the desired point. Simulation results show the efficiency of the proposed scheme in terms of fast response, high link utilization, and relatively stable buffer occupancy.