Showing papers by "Min-You Wu published in 2011"
10 Apr 2011
TL;DR: The experiments on the test bed show that RDCM handles packet loss without obvious throughput degradation during high-speed data transmission, and is implemented as a user-level library onWindows platform.
Abstract: Multicast benefits data center group communication in both saving network traffic and improving application throughput. The SLA (Service Level Agreement) of cloud service requires the computation correctness of distributed applications, translating to the requirement of reliable Multicast delivery. In this paper we present RDCM, a novel reliable Multicast approach for data center network. The key idea of RDCM is to minimize the impact of packet loss on the Multicast performance, by leveraging the rich link resource in data centers. A Multicast-tree-aware backup overlay is purposely built on group members for peer-to-peer packet repair. Riding on Unicast, packet repair not only achieves complete repair isolation, but also has high probability to bypass the pathological links in the Multicast tree where packet loss occurs. The backup overlay is organized in such a way that it causes little individual repair burden, control overhead, as well as overall repair traffic. We have implemented RDCM as a user-level library onWindows platform. The experiments on our test bed show that RDCM handles packet loss without obvious throughput degradation during high-speed data transmission.
41 citations
TL;DR: VStore, a cooperative storage solution in vehicular sensor networks for mobile surveillance, which has been designed to support redundancy elimination and storage balancing throughout the network, is presented.
23 citations
26 Sep 2011
TL;DR: The experimental results show that the proposed approaches can substantially improve the performance of irregular computation on GPUs and could be easily applied to many other irregular problems to improve their performance.
Abstract: While graphics processing units (GPUs) show high performance for problems with regular structures, they do not perform well for irregular tasks due to the mismatches between irregular problem structures and SIMD-like GPU architectures. In this paper, we explore software approaches for improving the performance of irregular parallel computation on graphics processors. We propose general approaches that can eliminate the branch divergence and allow runtime load balancing. We evaluate the optimization rules and approaches with the n-queens problem benchmark. The experimental results show that the proposed approaches can substantially improve the performance of irregular computation on GPUs. These general approaches could be easily applied to many other irregular problems to improve their performance.
15 citations
TL;DR: It is demonstrated that any solution of the scheduling subproblem can be built with |E| + 1 or fewer schedulable sets, where|E| is the number of links of the network, and due to the hardness of the problem, a polynomial time scheme is proposed.
Abstract: Recent advances in the physical layer have enabled the simultaneous reception of multiple packets by a node in wireless networks. We address the throughput optimization problem in wireless networks that support multipacket reception (MPR) capability. The problem is modeled as a joint routing and scheduling problem, which is known to be NP-hard. The scheduling subproblem deals with finding the optimal schedulable sets, which are defined as subsets of links that can be scheduled or activated simultaneously. We demonstrate that any solution of the scheduling subproblem can be built with |E| + 1 or fewer schedulable sets, where |E| is the number of links of the network. This result is in contrast with previous works that stated that a solution of the scheduling subproblem is composed of an exponential number of schedulable sets. Due to the hardness of the problem, we propose a polynomial time scheme based on a combination of linear programming and approximation algorithm paradigms. We illustrate the use of the scheme to study the impact of design parameters on the performance of MPR-capable networks, including the number of transmit interfaces, the beamwidth, and the receiver range of the antennas.
9 citations
01 May 2011
TL;DR: A new method to select the stop threshold for the image reduction process, which reduces the total computation by half, is given, and a color image output mode is provided to let the system enjoy more human-oriented design.
Abstract: This paper proposes techniques for face detection and gives the implementation details for an FPGA development board. We analyze and discuss the relation between the system computation cost and selection of the image scaling factor. We give a new method to select the stop threshold for the image reduction process, which reduces the total computation by half. We also provide a color image output mode to let our system enjoy more human-oriented design. Test results show that the system achieves real-time face detection speed (100fps) and a high face detection rate (87.2%) for an SVGA (600 × 800) video source. The low power consumption (3.5W) is another advantage over previous work.
8 citations
05 Jun 2011
TL;DR: A distributed scheduling scheme based on a k-Connected k-Dominating Set backbone to eliminate bottleneck effects on intermediate relay nodes as to enhance the network throughput and shows the effectiveness of the scheme by comparing its performance with the upper bound and node-disjoint routing.
Abstract: Multi-Packet Reception (MPR) enables simultaneous receptions from different transmitters to a single receiver, which has been demonstrated to bring capacity improvement in wireless network. However, MPR does not improve the transmission capability of intermediate relay nodes in a multi-hop routing and thus these nodes may become the bottlenecks for increasing throughput despite of great reception capability. We investigate the scheduling for multi-hop routing with MPR to improve the network throughput under multiple data flows. We formulate the optimization problem under K-MPR model and analyze the performance upper bound with ideal scheduling. We propose a distributed scheduling scheme based on a k-Connected k-Dominating Set backbone to eliminate bottleneck effects on intermediate relay nodes as to enhance the network throughput. We show the effectiveness of our scheme by comparing its performance with the upper bound and node-disjoint routing.
2 citations
21 Mar 2011
TL;DR: Experimental results showed that speedup increases with the problem size, and a FPGA-based parallel algorithm to compute the transitive closure of the relation matrix on a fixed-size PE array achieves acceleration rate of 3.7 and 376 under the worst and best situations, respectively.
Abstract: In this paper, we propose a FPGA-based parallel algorithm to compute the transitive closure of the relation matrix on a fixed-size PE array. Experimental results showed that speedup increases with the problem size. The speedup against a single PE is between 11.3 and 195.9. Compared to a general CPU solution, this algorithm achieves acceleration rate of 3.7 and 376 under the worst and best situations, respectively.
2 citations
01 Dec 2011
TL;DR: The proposed solution enables the sensors to interfere the localization and map generation on mobile targets so as to expose them in the sensor field and enhance the target detection performance in a WSN.
Abstract: In surveillance related Wireless Sensor Networks (WSNs) applications, mobile targets can utilize intelligent motion planning in robotics to avoid detection, which could impose great threats on the sensor field. In this paper, we propose countermeasures against mobile target motion planning. The proposed solution enables the sensors to interfere the localization and map generation on mobile targets so as to expose them in the sensor field. We compare the behaviors of a mobile target in naive mode, smart mode with Simultaneous Location And Mapping (SLAM) and interfered mode where sensor nodes use different puzzle algorithms to disturb it. The simulation results show that the puzzle algorithms can interfere SLAM on mobile targets and enhance the target detection performance in a WSN. We also build an experimental testbed with Iris motes and a mobile robot to validate our solution.
1 citations
05 Jun 2011
TL;DR: This paper introduces target bypassing routing methods applied on omnidirectional and directional sensor communication models to achieve the goal of reducing sensor communication exposure to targets.
Abstract: In a target detection application, rational adversary targets that are conscious of the deployed location of sensor nodes are capable of planning a path in order to avoid being detected by sensor nodes. Probing sensor communication is one of the means that are used by adversary targets to get the necessary information. Therefore, this paper investigates how to reduce the sensor communication exposure. We introduce target bypassing routing methods applied on omnidirectional and directional sensor communication models to achieve the goal of reducing sensor communication exposure to targets. The simulation results show that our bypassing methods can decrease the possibility of communication exposure to a large extent. We also implement a prototype system with Iris motes to verify our solution.
01 Dec 2011
TL;DR: This work combined couple of previous approaches to a single model with some improvements and successfully implemented it into a new simulator which can be used to evaluate other traffic applications upon the driver behavior model.
Abstract: Understanding the human behaviors in driving is a highly complex topic but on the other hand, extremely important for improving human's driving competence by developing driver assistant systems accordingly and finally accomplishing fully autonomous vehicles and road systems. There are many excellent researches in this area for decades. However, by building a comprehensive system to evaluate human driver's behavior, we found that none of the existing model can reflect the same outcome as our observation from the real world in the multiplex traffic scenario. But this is indeed important for us to find the right way to improve road traffic efficiency as well as develop autonomous vehicles. In this work, we combined couple of previous approaches to a single model with some improvements and successfully implemented it into a new simulator which can be used to evaluate other traffic applications upon the driver behavior model. We focused on lane changing decision of the driver behavior since it is the most complex part and causes the most problem and inconsistency in previous researches. After comparing the result getting from the simulation based on the new model with the real world observation, we can conclude that the new driver behavior model does reflect the real world scenario. And the model and simulator also helps us in our further researches in intelligent traffic control area.