The complex, distributed nature of data centers have spawned the adoption of distributed, multi-tiered software architectures, consisting of many inter-connected microservices. These microservices exhibit extremely short request service times, often less than 250μs. We show that these “killer microsecond” service times can cause state-of-the-art dynamic power management techniques to break down, due to short idle period length and low power state transition overheads. In this paper, we propose μDPM, a dynamic power management scheme for the microsecond era that coordinates request delaying, per-core sleep states, and voltage frequency scaling. The idea is to postpone the wake up of a CPU as long as possible and then adjust the frequency so that the tail latency constraint of requests are satisfied just-in-time. μDPM reduces processor energy consumption by up to 32% and consistently outperforms state-of-the-art techniques by 2×.

Achieving energy proportionality in data centers supporting latency-sensitive applications is challenging because of the strict Service Level Agreements. Previous works individually focus on making the server energy proportional or reducing the data center network's power consumption for latency-tolerant applications. In this paper, we propose EPRONS to minimize the overall data center's power consumption with latency-sensitive applications by trading-off network slack in favor of providing additional slack for computations. We utilize the linear programming model to consolidate latency-sensitive search queries and latency-tolerant background flows to a minimal subnet of the topology by turning off unused switches and links without violating the application deadlines. Servers take advantage of the additional 'network-provided' slack to allow slowing down request processing. For servers, we design a novel power saving technique using Dynamic Voltage and Frequency Scaling (DVFS) based on the average tail latency of a request. If needed, we turn on a minimal number of additional network links and switches to reduce network latency while still maximizing entire data center's power saving. Experimental results show that our scheme saves up to 31.25% of a data center's total power budget.

Web search constitutes an important class of data-intensive online services in data centers. Optimizing search systems for energy efficiency, timely response and high search quality (i.e., how relevant the returned results are to a search query), however, is very challenging, as a search system involves a distributed architecture with hundreds of thousands of index serving nodes (ISNs) that return searching results to an aggregator through multiple interdependent retrieval stages in a partition-aggregate fashion. In this paper, we discover through experiments two important characteristics that can affect the system performance: (1) response time and energy consumption are greatly impacted by a small fraction of queries with long processing times; (2) the quality contribution of the ISN is independent of the query processing time. Based on our observation, we propose TailCut, which judiciously discards long query executions and enables ISN-aggregator coordination to minimize energy consumption subject to latency and quality constraints. Our experimental results show that TailCut can achieve up to 39% power saving, while satisfying the tail latency and quality constraint.

Servers running in datacenters are commonly kept underutilized to meet stringent latency targets. Due to poor energy-proportionality in commodity servers, the low utilization results in wasteful power consumption that cost millions of dollars. Applying dynamic power management on datacenter workloads is challenging, especially when tail latency requirements often fall in the sub-millisecond level. The fundamental issue is randomness due to unpredictable request arrival times and request service times. Prior techniques applied per-core DVFS to have fine-grain control of slowing down request processing without violating the tail latency target. However, most commodity servers only support per-core DFS, which greatly limits potential energy saving. In this paper, we propose DynSleep, a fine-grain power management scheme for datacenter workloads through the use of per-core sleep states (C-states). DynSleep dynamically postpones the processing of some requests, creating longer idle periods, which allow the use of deeper C-states to save energy. We design and implement DynSleep with Mem-cached, a popular key-value store application used in datacenters. The experimental results show that DynSleep achieves up to 65% core power saving, and 27% better than the per-core DVFS power management scheme, while still satisfying the tail latency constraint. To the best of our knowledge, this is the first work to analyze and develop power management technique with CPU C-states in latency-critical datacenter workloads

As processor power density increases, thermal and power control becomes critical for application processing. In this paper, we consider network applications which feature ON/OFF execution pattern, that causes frequent temperature and power consumption changes in the processor. A novel power aware thermal management algorithm is designed to achieve power saving in multicore processors by employing a vacation scheme. We implement the scheme through the idle states (C-state) provided by the OS in the CPU and show their effectiveness both through analysis and experimental data. Then, we apply our scheme with the thermal constraint and propose a heterogeneous load distribution, which creates more opportunities for power saving. Besides maintaining processor temperature below the temperature constraint, our technique achieves higher sustainable load and better power saving with minimum latency increase compared to existing thermal management techniques. To the best of our knowledge, this is the first work to discuss and develop vacation algorithm considering power, temperature and latency for network application on a general purpose multicore processor.

As processor power density increases, thermal and power control becomes critical for packet processing on a processor. In "run-to-finish" applications, power consumption is stable and temperature simply rises to saturation point and then stabilizes. But, network applications feature ON/OFF execution pattern, which causes frequent temperature and power consumption changes in the processor. We propose a thermal aware scheduler, TrafficLight, which achieves power saving by employing vacation and rate adaptation techniques. We implement these through the idle states (C-state) provided by the OS in a CPU and show their effec-tiveness through experimental data. Then we build power, thermal and latency models based on the vacation queuing theory, which estimates the performance of our proposed techniques. Finally, we design, implement and evaluate an on-line algorithm to dynamically choose the proper pow-er/thermal management technique based on the traffic variation. The technique maintains the processor temperature below the temperature constraint and achieves power saving. To the best of our knowledge, this is the first work to provide the theoretical analysis as well as the experimental results for the vacation and rate adaptation schemes considering power, temperature and latency in the packet processing on a general purpose processor.

As processor power density increases, chip/core temperature control becomes critical for building multicore systems. This paper addresses the problem of inter-core thermal coupling and periodic thermal variation while executing multi-threaded network applications in a multicore architecture.

Traditional birdsong recognition approaches used acoustic features based on the acoustic model of speech production or the perceptual model of the human auditory system to identify the associated bird species. In this paper, a new feature descriptor that uses image shape features is proposed to identify bird species based on the recognition of fixed-duration birdsong segments where their corresponding spectrograms are viewed as gray-level images. The MPEG-7 angular radial transform (ART) descriptor, which can compactly and efficiently describe the gray-level variations within an image region in both angular and radial directions, will be employed to extract the shape features from the spectrogram image. To effectively capture both frequency and temporal variations within a birdsong segment using ART, a sector expansion algorithm is proposed to transform its spectrogram image into a corresponding sector image such that the frequency and temporal axes of the spectrogram image will align with the radial and angular directions of the ART basis functions, respectively. For the classification of 28 bird species using Gaussian mixture models (GMM), the best classification accuracy is 86.30% and 94.62% for 3-second and 5-second birdsong segments using the proposed ART descriptor, which is better than traditional descriptors such as LPCC, MFCC, and TDMFCC.

This article pursues speedy packet classification with low on-chip memory requirements realized on Xilinx Virtext-6 FPGA. Based on hashing round-down prefixes specified in filter rules (dubbed HaRP), our implemented classifier is demonstrated to exhibit an extremely low on-chip memory requirement (lowering the byte count per rule by a factor of 8.6 in comparison with its most recent counterpart [2]), taking only 50% of Virtex-6 on-chip memory to store every large rule dataset (with some 30K rules) examined. In addition, it achieves a higher throughput than any known FPGA implementation, reaching more than 200 MPPS (millions packet lookups per second) with 8 processing units and 8 memory banks in the HaRP pipeline to support the line rate over 130 Gbps under bi-directional traffic in the worst case with 40-byte packets. By reducing memory probes per lookup, enhanced HaRP can further boost the classification speed to 255 MPPS.

In this paper, we will propose a 3D model retrieval approach using 2D cepstral features. First, six projection planes representing the elevation (depth) value are generated. Then, 2D cepstral features are extracted from each projection plane for searching similar 3D models. Experiments conducted on the Princeton Shape Benchmark (PSB) database have shown that the proposed 2D cepstral features outperforms other state-of-the-art descriptors in terms of the DCG score.

Minimizing energy consumption is a fundamental requirement when deploying wireless sensor networks. Accordingly, various topology control protocols have been proposed, which aim to conserve energy by turning off unnecessary sensors while simultaneously preserving a constant level of routing fidelity. However, although these protocols can generally be integrated with any routing scheme, few of them take specific account of the issues which arise when they are integrated with geographic routing mechanisms. Of these issues, the dead-end situation is a particular concern. The dead-end phenomenon (also known as the "local maximum” problem) poses major difficulties when performing geographic forwarding in wireless sensor networks since whenever a packet encounters a dead end, additional overheads must be paid to forward the packet to the destination via an alternative route. This paper presents a distributed dead-end free topology maintenance protocol, designated as DFTM, for the construction of dead-end free networks using a minimum number of active nodes. The performance of DFTM is compared with that of the conventional topology maintenance schemes GAF and Span, in a series of numerical simulations conducted using the ns2 simulator. The evaluation results reveal that DFTM significantly reduced the number of active nodes required in the network and thus prolonged the overall network lifetime. DFTM also successfully constructed a dead-end free topology in most of the simulated scenarios. Additionally, even when the locations of the sensors were not precisely known, DFTM still ensured that no more than a very few dead-end events occurred during packet forwarding.

The large sized data sets are replicated in more than one site for the better availability to the nodes in a grid. Downloading the dataset from these replicated locations have practical difficulties, due to network traffic, congestion, frequent change-in performance of the servers, etc. In order to speed up the download, complex server selection techniques, network and server loads are used. However, consistent performance is not guaranteed due to the shared nature of network links of the load on them, which can vary unpredictably. In this paper, we present a bandwidth sensitive co-allocation scheme for parallel downloading in grid economics. Objective of the proposed technique aims to service grid applications efficiently and economically in data grids. With the consideration of cost factor, we present a novel mechanism for server selection, dynamic file decomposition and co-allocation. Under considerations in costs, our mechanism for selections of servers with various techniques combined is able to significantly attenuate economic costs. We compared our scheme with the existing schemes and the preliminary results show notable improvement in overall completion time of data transfer.

In this paper, we present ReWarn, an opportunistic relay scheme for cooperative collision warning in VANET, which helps to mitigate the impact of package loss caused by shadow fading. When safety messages in VANET, e.g., collision warning messages, are broadcast in air, a vehicle may detect collision risks both of its own and of its neighbors'. If a vehicle has detected its neighboring vehicles are experiencing a collision risk while they keep quiet, it then probabilistically relays the raw data received from the neighboring vehicles with its own detection. Simulation results show that ReWarn can achieve higher reliability in message transmission than the non-relay schemes and keep the transmission efficiency at a satisfied level as well.

Birdsongs are typically divided into four hierarchical levels: note, syllable, phrase, and song, of which syllable plays an important role in bird species recognition. To improve the recognition rate of birdsongs, in this study an enhanced syllable segmentation method based on R-S endpoint detection method was presented. Furthermore, a decision based neural network with suitable reinforcement learning rule was developed as the classifier. The proposed methods combined with the well-known MFCCs feature vector form a birdsong recognition system that was applied to two recognition problems: one is the recognition of a set of arbitrary syllables and the other is the recognition of a section of a birdsong. Experimental results show the performances of the proposed methods.

Ad hoc networks have been extensively deployed in recent years. In such networks, obtaining a high efficiency is difficult to achieve due to mobility. Many routing approaches have been proposed to improve the performance of ad hoc networks, including table-driven and demand-driven routing protocols. Geographic routing provides a suitable solution for large-scale and dynamic networks, and reduces link failure problems during packet routing. In wireless networks, nodes movement may cause a network partition, which hinders both data transmission and information availability within the network. Several schemes have been proposed for resolving the partition problem in wireless ad hoc networks. In general, the previous solutions either relied on the use of mobile devices with known trajectories or introduced additional specific mobile nodes. By contrast, this paper proposes a scheme for overcoming the problem of network partitions using only the available devices in the environment. With the scheme, the source node searches for another mobile node, designated as the helping node, which is able to meet the destination. The selected mobile node then carries the packet toward the destination. If the helping node changes its trajectory before it forwards the packet to the destination node, a hand-off scheme is activated and the data packet is transferred to another mobile node for onward transmission

With the progress of advanced technology, people anticipate better quality in the medical health care service. Due to more robust requirements in medical environments, higher availability and reliability are expected. This paper describes a recovery scheme for disconnection failures in a Web-based medical teleconsultation system. With the support, the disconnected clients can re-enter the system to resume the unfinished discussion using the synchronization mechanisms. The experiments have shown that the mechanism performed successfully. The lazy synchronization further reduced the required time that the clients had to wait for joining the consultation after network disconnections

In this paper, a genetic algorithm (GA) based optimal fuzzy controller design is proposed. The design procedure is accomplished by establishing an index function as the consequent part of the fuzzy control rule. The inputs of the controller, after scaling, are utilized by the index function for computing the output linguistic value. This linguistic value can then be used to map the suitable fuzzy control actions. This proposed novel fuzzy control rule has crisp input and fuzzified output characteristics. The index function plays a role in mapping the desired fuzzy sets for defuzzification resulting in a controlled hypersurface in the linguistic space formed by the input fuzzy variables. Two types of index functions, both linear and nonlinear, are introduced for controlling systems with different degrees of nonlinearity. The parameters of the index function are obtained by applying a simple GA with a suitable fitness function. Various controlled systems result in various parameter sets depending on their dynamics. Under the acquired optimal parameter set the optimal index function can be used to generate the desired control actions. Several simulation examples are given to verify the performance of the proposed GA-based fuzzy controller.

Location service support is required for mobile nodes using geographic ad hoc routing. The mobile nodes need location information of destinations for sending data packets. Due to mobility or node failures, queries to the location servers may not always succeed. This paper presents a distributed location service (DLS) that employs dual servers to maintain position information of mobile nodes and to achieve efficient performance. With the scheme, the number of failed queries, including no response and erroneous results, can be reduced. DLS has been implemented and evaluated using the ns2 simulator. The experimental results show that DLS not only improved service availability but also location query latency.

Many researchers on GAs lay great emphasis on the improvements of the methods of crossover, mutation and selection. These methods include dynamic crossover and mutation rates, varying population size and varying encoding length, and so on. In this paper, however, we develop methods including the extraction operator, the uniform initialization and the unbiased initialization. The extraction operator alters the inner structure of the individual, whereas the uniform initialization and the unbiased initialization methods modify the population initialization. All of these methods are simple and can be combined with the simple genetic algorithms easily. Simulation results show that these methods exhibit an evident improvement on the performance of GAs.