Performance Metrics Of Interest Research Articles

Beyond energy savings: Investigating the co-benefits of heat resilient architecture

Heat is a growing concern in cities around the word, especially in the face of climate change. Because buildings are an important component of the built environment vis-à-vis both energy use and heat resiliency in cities, we explored their climate mitigation and adaptation potential. Specifically, we investigated how design decisions interact with regard to three heat-related factors—namely, energy use/CO2 emissions, passive survivability, and heat rejection to the urban climate. We selected an archetypical building as our test case, created various design permutations, and used whole-building simulations to analyze their performance. Our simulations show that permutations of the building with a smaller carbon footprint also emit less heat to ambient air and had a better passive survivability. However, we also noted potential trade-offs (e.g., where ventilation is inadequate, increasing insulation levels for energy efficiency may hurt passive survivability). Based on our findings, we argue that, at least at a policy level, it is imperative to take advantage of the synergies, and their collective benefits. Moreover, building regulations or incentive programs should look beyond energy as the sole performance metric of interest and consider passive survivability as well as thermal interactions with urban climate.

Outage and Throughput Analysis of Cognitive Users in Underlay Cognitive Radio Networks With Handover

Interference characterization in cognitive networks with handover has received less attention in stochastic geometry-based interference management and control, especially in cognitive radio networks, because of the possibility of complicating the analysis of various performance metrics of interest, such as outage probability and throughput. However, because of the possible mobility that is observed in real practical systems, some of the receivers may be located outside the coverage regions of their paired transmitters. In order to ensure that any receiver located outside the coverage region of its paired transmitter continues to receive its required service from its paired transmitter while still achieving tractable analysis for various performance metrics of interest, we adopted multiuser diversity via packet relaying. With this approach, any secondary nodes waiting to transmit can be used to sustain coverage between any typical active transmitter and receiver pair, while reducing their own waiting period in the process. We obtained tractable analysis for outage probability, spectral efficiency and throughput and showed the effect of handover rate over the network performance. The outcomes of the numerical results show that the proposed approach is capable of improving the overall network performance by improving coverage and throughput among network users in the cognitive radio networks.

Joint Device Association and Power Coordination for H2H and IoT Communications in Massive MIMO Enabled HCNs

The integration of massive multiple-input and multiple-output (MIMO) enabled heterogeneous cellular networks (HCNs) and Internet of Things (IoT) is a promising treatment for future networking paradigms. In such networks, the human-to-human (H2H) devices and IoT devices have some distinct service requirements. Specifically, the former mainly concentrates on downlink throughput, but the latter pays more attention to uplink power consumption. Based on this, we design a device association mechanism (scheme) to achieve a tradeoff between these two performance metrics under devices' association requirements. Through some appropriate adjustments of weighting parameters, different types of devices can improve their performance metrics of interest. At last, such a scheme is formulated as a network-wide logarithmic utility maximization problem in a nonlinear and combinatorial form. To solve it, we develop two types of association algorithms, whose primary difference lies in the treatment of weighting parameters. Then, we show the corresponding convergence and computation complexity analyses for them. Finally, we investigate the impacts of weighting parameters, the BS power and the number of antennas on some certain performance metrics of designed algorithms and other existing one. The simulation results show that the designed algorithms can meet different devices' association requirements by properly adjusting weighting parameters.

Achievable Physical-Layer Security Over Composite Fading Channels

We investigate the physical layer security limits of Wyner's wiretap model over Fisher-Snedecor .T composite fading channels. .T fading conditions have been recently shown to provide an accurate characterization of multipath fading and shadowing effects in emerging wireless transmission scenarios such as body centric, cellular and vehicular communications. To this end, we utilize a redefined analytic expression for the Fisher-Snedecor .T distribution in order to ensure unconstrained validity and reliability when used in the analysis of various performance metrics of interest. In this context, we assume that the main channel (i.e., between the source and the legitimate destination) and the eavesdropper's channel (i.e., between the source and the illegitimate destination) undergo independent quasi-static Fisher-Snedecor .T composite fading. Novel exact analytic expressions are then derived for the corresponding average secrecy capacity (ASC), secure outage probability (SOP) and probability of strictly positive secrecy capacity (SPSC) along with their insightful asymptotic representations. In addition, analytical expressions for the ASC, SOP and SPSC over mixed fading channels such as Nakagami-m/Nakagami-m, Nakagami-m/Fisher-Snedecor .T and Fisher-Snedecor .T/Nakagami-m are derived. The new formulations are validated through comparisons with Monte-Carlo simulations and analyzed to gain useful insights into the impact of the fading parameters on the achievable accuracy and the overall system performance.

An End-to-End Performance Analysis for Service Chaining in a Virtualized Network

Future mobile networks supporting Internet of Things are expected to provide both high throughput and low latency to user-specific services. One way to overcome this challenge is to adopt Network Function Virtualization (NFV) and Multi-access Edge Computing (MEC). Besides latency constraints, these services may have strict function chaining requirements. The distribution of network functions over different hosts and more flexible routing caused by service function chaining raise new challenges for end-to-end performance analysis. In this paper, as a first step, we analyze an end-to-end communication system that consists of both MEC servers and a server at the core network hosting different types of virtual network functions. We develop a queueing model for the performance analysis of the system consisting of both processing and transmission flows. We propose a method in order to derive analytical expressions of the performance metrics of interest, i.e., end-to-end delay, system throughput, task drop rate. Then, we show how to apply the similar method to a larger system and derive a stochastic model for such systems. We observe that the simulation and analytical results are very close. By evaluating the system under different scenarios, we provide insights for the decision making on traffic flow control and its impact on critical performance metrics.

Quantifying the Impact of Common Feeding Interventions on Nutritive Sucking Performance Using a Commercially Available Smart Bottle.

An estimated 25% to 40% of infants experience difficulties with learning to breast- or bottle-feed. Yet, guidelines and evidence-based support for common feeding practices are limited. The objective of this case report was to quantify the impact of feeding interventions on nutritive sucking performance after discharge in an outpatient setting. This observational case series involved 2 infants. To determine the impact of cumulative interventions, pre- and postintervention effect sizes were calculated. Sucking performance metrics of interest included nipple movement peak sucking amplitude, duration, frequency, and smoothness. Interventions included positional changes and changes in nipple flow rate, among others. For both infants, cumulative interventions had the greatest impact on suck frequency; postintervention, infants were able to increase their rate of nutritive sucking per burst. Other aspects of sucking performance were differentially impacted for each baby. Researchers agree that neonatal and infant feeding has been understudied and that the evidence for common interventions needs to be strengthened. We have demonstrated the implementation of readily available technology that can be used to quantify the direct impact of any intervention on actual sucking performance. In doing so, we can individualize care to support skill development and improve outcomes for infants at risk for ongoing feeding challenges.

Real-Time Status Updates for Markov Source

For timely sensor update, the traditional approach is to send new information at every available opportunity. Recent research has shown that with limited receiver feedback, sensors can improve the update timeliness by transmitting differential information for slowly varying correlated sources. One can elect to transmit either the actual or the differential state information based on a differential encoding threshold for a general Markov source. This threshold captures the natural trade-off between differential transmission opportunities and the coding gains. Using matrix-geometric method , we find the limiting age distribution for a Markov source as a function of the encoding threshold, from which several other performance metrics of interest, such as mean age, peak age, and the probability of decoding failure can be derived.

(Invited) Nanostructured Silicon Integrated Circuit Thermoelectric Generators with High Specific Power Generation Capacity

Microelectronic thermoelectric (TE) devices aimed at integrated circuit (IC) and sensor applications such as on-chip thermal management, bio-thermal power for wearable electronics, and microwatt level sustainable power sources for internet-of-things (IoT) sensors have grown in recent interest. While silicon CMOS by far dominates microelectronics technology, Si has historically not been considered as a viable TE material because bulk Si has a very low TE figure-of-merit ZT ~ 10–3 to 10–2 near room temperature. By contrast, modern “high ZT” materials such as the (Bi,Sb)2(Se,Te)3 family have ZT ≈ 1. However, these materials are incompatible with Si CMOS processing and so have not gained widespread adoption into microelectronic ICs. In 2008 two groups [Hochbaum, et al. Nature 451, 163; Boukai, et al. Nature 451, 168] reported that Si nanowires (NWs) can have ZT up to 0.6. Nevertheless, numerous efforts to develop microelectronic TE generators (TEG) using Si NWs or other nano-structured Si thermopiles have not achieved efficiencies better than what is consistent with ZT ~ 10–2. In practice, efficiency is not necessarily the primary concern in many applications but rather how much power a TEG of given size generates from a temperature difference ∆T, as well as cost per Watt generated. Since the maximum power P max generated is proportional both to (∆T)2 and to cross-sectional area A for heat flow, a performance metric of more direct practical interest is the specific power generation capacity Γ P = P max/[A·(∆T)2], as this measures TEG output power scaled to device size and to ∆T. Γ P is an extrinsic device metric that can be engineered by circuit design and management of parasitic impedances, in contrast to ZT which is an intrinsic material characteristic. In this presentation we will describe IC TEGs using nanostructured silicon thermopiles fabricated on an industrial “65 nm” technology node Si CMOS process line following standard CMOS process design rules. The thermopile elements are nanostructured n- and p-type Si “blades” nominally 80 nm wide x 750 nm long x 320 nm tall, with a typical total thermopile area of 48 µm x 36 µm. Depending on design layout and doping, these Si IC TEGs show high Γ P, up to 29 µW·cm–2K–2, with most designs having Γ P near 8.5 µW·cm–2K–2 using a cold side heat sink at room temperature. In terms of power generation ability a Γ P of order 10 µW·cm–2K–2 makes these Si IC TEGs competitive with (Bi,Sb)2(Se,Te)3 based TEGs, although the net efficiency of these Si IC TEGs is still low. Engineering a high Γ P despite low efficiency results from the ability of Si CMOS processing to fabricate a high areal density of thermocouples while still maintaining a low areal packing fraction, that is, to make uniformly a very large number of very small things in a given area, and to carefully control both intrinsic and parasitic electrical and thermal impedances. We find that TEG power increases significantly with decreasing thermocouple width, giving a fabrication path to further improvements. Unlike (Bi,Sb)2(Se,Te)3 TEGS, these Si IC TEGs can be seamlessly integrated into large scale Si CMOS microelectronic circuits at very low marginal cost.

Modeling Curbside Parking as a Network of Finite Capacity Queues

Paid curbside parking can be advantageously modeled as a network of interdependent queues. To this end, we introduce methods for analyzing a special class of networks of finite capacity queues where drivers arrive from an exogenous source, join the queue if there is an available parking space, or continue to search at an adjacent queue for an available space. Furthermore, we apply this model to estimate the proportion of drivers cruising in the neighborhood of Belltown, Seattle, WA, USA. Using occupancy approximated by parking transaction data, we estimate the percentage of drivers cruising for curbside parking by comparing the rate of drivers unable to find parking to bulk through-traffic measurement data. We find percentages of up to 50% for a Belltown's 1st Ave. depending on the time, day, and direction of travel. We then calculate a per vehicle travel-time cost to social welfare incurred by this proportion: upward of a 10% increase in travel time to all drivers along 1st Ave. Last, we introduce a simulation tool and test assumptions made when estimating interesting performance metrics like the probability of a block-face being full. Our results suggest that while assuming exponential service time distributions is not justified, mean rate solutions under a Markovian relaxation of the problem is comparable to service times representative of parking transaction data in simulation.

FpgaConvNet: Mapping Regular and Irregular Convolutional Neural Networks on FPGAs.

Since neural networks renaissance, convolutional neural networks (ConvNets) have demonstrated a state-of-the-art performance in several emerging artificial intelligence tasks. The deployment of ConvNets in real-life applications requires power-efficient designs that meet the application-level performance needs. In this context, field-programmable gate arrays (FPGAs) can provide a potential platform that can be tailored to application-specific requirements. However, with the complexity of ConvNet models increasing rapidly, the ConvNet-to-FPGA design space becomes prohibitively large. This paper presents fpgaConvNet, an end-to-end framework for the optimized mapping of ConvNets on FPGAs. The proposed framework comprises an automated design methodology based on the synchronous dataflow (SDF) paradigm and defines a set of SDF transformations in order to efficiently navigate the architectural design space. By proposing a systematic multiobjective optimization formulation, the presented framework is able to generate hardware designs that are cooptimized for the ConvNet workload, the target device, and the application's performance metric of interest. Quantitative evaluation shows that the proposed methodology yields hardware designs that improve the performance by up to 6.65× over highly optimized graphics processing unit designs for the same power constraints and achieve up to 2.94× higher performance density compared with the state-of-the-art FPGA-based ConvNet architectures.

A Tractable Framework for the Analysis of Dense Heterogeneous Cellular Networks

This paper investigates the downlink performance of dense $K$ -tier heterogeneous cellular networks (HCNs) under general settings. First, Gaussian approximation bounds for the standardized aggregate wireless interference (AWI) in dense $K$ -tier HCNs are obtained for when base stations (BSs) in each tier are distributed over the plane according to a spatial and general Poisson point process. The Kolmogorov–Smirnov (KS) distance is used to measure deviations of the distribution of the standardized AWI from the standard normal distribution. An explicit and analytical expression bounding the KS distance between these two distributions is obtained as a function of a broad range of network parameters, such as per-tier transmission power levels, per-tier BS intensity, BS locations, general fading statistics, and general bounded path-loss models. Bounds achieve a good statistical match between the standardized AWI distribution and its normal approximation even for moderately dense HCNs. Second, various spatial performance metrics of interest, such as outage capacity, ergodic capacity, and area spectral efficiency in the downlink of $K$ -tier HCNs for general signal propagation models are investigated by making use of the derived distribution approximation results. Considering two specific BS association policies, it is shown that the derived performance bounds track the actual performance metrics reasonably well for a wide range of BS intensities, with the gap among them becoming negligibly small for denser HCN deployments. Finally, both analytical and numerical results on the area spectral efficiency reveal a non-linear growth trend with diminishing returns of HCN performance. Hence, the ${\sf SIR}$ invariance property does not hold under bounded path-loss models, which is a critical finding from the HCN design perspective. In particular, it points out a critical BS density beyond which the HCN performance starts to decline due to excessive wireless interference.

Exploring the Utility of Developer Exhaust.

Using machine learning to analyze data often results in developer exhaust - code, logs, or metadata that do not define the learning algorithm but are byproducts of the data analytics pipeline. We study how the rich information present in developer exhaust can be used to approximately solve otherwise complex tasks. Specifically, we focus on using log data associated with training deep learning models to perform model search by predicting performance metrics for untrained models. Instead of designing a different model for each performance metric, we present two preliminary methods that rely only on information present in logs to predict these characteristics for different architectures. We introduce (i) a nearest neighbor approach with a hand-crafted edit distance metric to compare model architectures and (ii) a more generalizable, end-to-end approach that trains an LSTM using model architectures and associated logs to predict performance metrics of interest. We perform model search optimizing for best validation accuracy, degree of overfitting, and best validation accuracy given a constraint on training time. Our approaches can predict validation accuracy within 1.37% error on average, while the baseline achieves 4.13% by using the performance of a trained model with the closest number of layers. When choosing the best performing model given constraints on training time, our approaches select the top-3 models that overlap with the true top- 3 models 82% of the time, while the baseline only achieves this 54% of the time. Our preliminary experiments hold promise for how developer exhaust can help learn models that can approximate various complex tasks efficiently.

Optimization of uncertain and dynamic high-rise structures for occupant comfort: An adaptive kriging approach

Performance requirements concerning serviceability often govern the design of high-rise structures. While various methods have been proposed for obtaining optimal systems that meet serviceability constraints set on the performance of non-structural components, e.g. the cladding system, methods are still lacking for optimizing under constraints aimed at ensuring occupant comfort. This is especially true for systems whose performance is assessed within modern performance-based design frameworks that consider a full range of uncertainties, including wind loads modeled as stochastic processes. This paper is focused on the development of a novel optimization framework for overcoming these limitations. The method is based on constructing an adaptive kriging metamodel of the probabilistic performance metric of interest in the low-dimensional modal space of the system. By then relating this space to the design variables, an approach is defined that can handle a wide class of occupant comfort metrics within high-dimensional spaces of design variables and uncertain parameters.

An Extension of the <inline-formula> <tex-math notation="LaTeX">$\kappa$</tex-math> </inline-formula>-<inline-formula> <tex-math notation="LaTeX">$\mu$</tex-math> </inline-formula> Shadowed Fading Model: Statistical Characterization and Applications

We here introduce an extension and natural generalization of both the \kappa-\mu$\,$shadowed and the classical Beckmann fading models: the Fluctuating Beckmann (FB) fading model. This new model considers the clustering of multipath waves on which the line-of-sight (LoS) components randomly fluctuate, together with the effect of in-phase/quadrature power imbalance in the LoS and non-LoS components. Thus, it unifies a variety of important fading distributions as the one-sided Gaussian, Rayleigh, Nakagami-m, Rician, \kappa-\mu, \eta-\mu, \eta-\kappa, Beckmann, Rician shadowed and the \kappa-\mu$\,$shadowed distribution. The chief probability functions of the FB fading model, namely probability density function, cumulative distribution function and moment generating function are derived. The second-order statistics such as the level crossing rate and the average fade duration are also analyzed. These results can be used to derive some performance metrics of interest of wireless communication systems operating over FB fading channels.

A novel hybrid priority discipline for multi-class secondary users in cognitive radio networks

Recently, service differentiation in cognitive radio networks is an important issue. Delay-sensitive traffic such as VoIP transmission is given a higher priority over delay-insensitive traffic such as data transmission. Modeling such situation using classical preemption or non-preemption priority service disciplines may result in a severe degradation of the QoS for a particular class. To address this problem, we propose a hybrid service discipline at which low priority secondary users are no longer preempted by high priority secondary users when their number of interruptions reaches a certain threshold value. We carefully design this novel service discipline taking into account that the arrival of a primary user always preempts any secondary user ongoing transmission. Analytical formulas are derived for the extended data delivery time, the overall system time and the dropping probability for each class. Based on a set of numerical results presented for a 2-class network, it is shown that adjusting the threshold value according to the performance metric of interest provides a promising performance.

A Flexible Datacenter Simulator

The current data volume and processing requirements in a modern datacenter outgrow the performance increase offered by CPU technology scaling. In light of this trend, datacenter operators have started exploiting application-specific accelerators in order to provide the performance that is needed, without exceeding power consumption constraints. The resulting heterogeneity of resources raises the question of what type of datacenter composition is most useful and under what circumstances. However, as the resulting complexity makes the testing on real-life datacenters difficult or even prohibitive, the use of simulation tools allow datacenter planners to estimate the results of their design decisions. We have developed Het-Sim, a software simulator based on the theoretical modeling of the datacenter, its components, expected workloads, and its possible deployments. Het-Sim takes as inputs a workload and a deployment description and outputs performance metrics of interest, such as task or job latency and resource utilization. In Het-Sim one can configure several parameters, including how tasks are allocated to nodes, and estimations of how fast they execute on different accelerators. Het-Sim can be used to explore how different deployments respond to different kinds of workloads, thus allowing one to determine how to best compose a datacenter based on particular workload, performance, or budgeting requirements.

Experimental Study of Multicast and Unicast Video Traffic in WAN Links

In recent years there has been an exponential increase in the growth in multimedia applications, and in particular in video applications. Understanding the behavior of the video traffic and the requirements for the network helps network administrators to improve the traffic. In this work, a quantitative analysis is performed by experimentation, in order to evaluate the behavior and impact of video traffic on WAN networks. We propose a WAN test bed with fourth subescenarios that allows to inject unicast or multicast video traffic, compressed with several codecs. The generic topology used includes a computer running as a streaming server and up to 20 desktop computers (PCs) as clients connected at the ends of a network as if outside a WAN. This network WAN is formed by routers and switchs with different types of links interconnecting the same. For the operation between routers was configured the OSPF routing protocol v2. For the same routers are configured the routing multicast protocol PIM in dense mode. From capturing video traffic, we identified several interesting performance and Qo metrics, such as unicast and multicast throughput, delay, jitter, and Pareto distribution. We include detailed contributions on the impact produced by several factors, such as the codec used for the compression, and the use of unicast and multicast traffic when there are restrictions on the bandwidth, as in a corporate real WAN link of some few Mbps. The work oriented on the configurations to be taken into account for an adequate management of similar networks and an efficient use of the resources available, without compromising the performance and the expected QoS. Besides facilitates the comparison of the results with those obtained from analytical studies and modelling for different contexts.

Solute concentration at a well in non-Gaussian aquifers under constant and time-varying pumping schedule

Our study is keyed to the analysis of the interplay between engineering factors (i.e., transient pumping rates versus less realistic but commonly analyzed uniform extraction rates) and the heterogeneous structure of the aquifer (as expressed by the probability distribution characterizing transmissivity) on contaminant transport. We explore the joint influence of diverse (a) groundwater pumping schedules (constant and variable in time) and (b) representations of the stochastic heterogeneous transmissivity (T) field on temporal histories of solute concentrations observed at an extraction well. The stochastic nature of T is rendered by modeling its natural logarithm, Y=lnT, through a typical Gaussian representation and the recently introduced Generalized sub-Gaussian (GSG) model. The latter has the unique property to embed scale-dependent non-Gaussian features of the main statistics of Y and its (spatial) increments, which have been documented in a variety of studies. We rely on numerical Monte Carlo simulations and compute the temporal evolution at the well of low order moments of the solute concentration (C), as well as statistics of the peak concentration (Cp), identified as the environmental performance metric of interest in this study. We show that the pumping schedule strongly affects the pattern of the temporal evolution of the first two statistical moments of C, regardless the nature (Gaussian or non-Gaussian) of the underlying Y field, whereas the latter quantitatively influences their magnitude. Our results show that uncertainty associated with C and Cp estimates is larger when operating under a transient extraction scheme than under the action of a uniform withdrawal schedule. The probability density function (PDF) of Cp displays a long positive tail in the presence of time-varying pumping schedule. All these aspects are magnified in the presence of non-Gaussian Y fields. Additionally, the PDF of Cp displays a bimodal shape for all types of pumping schemes analyzed, independent of the type of heterogeneity considered.

Use of a Reduced Order Model (ROM) to Simulate SOFC Performance in System Models

Use of SOFC-based hybrid power generation systems has the potential to reduce emissions while producing electricity from conventional fossil energy sources at costs that are competitive with other technologies. The numerical models used to perform predictive design analysis and optimization of the power generation system require an accurate representation of the SOFC stack performance over a varying range of operating conditions. Detailed stack models may contain the required information necessary for design purposes, but are often too computationally expensive for use in the overall system model. However, other desirable information about the state of the stack, such as the internal temperature gradient, is generally not available from simple performance or thermodynamic models often used to represent the SOFC. Such additional information can be important in system optimization studies to preclude operation under off-design conditions that can adversely impact overall system reliability. Response surface techniques were used with detailed SOFC stack model results to create a computationally efficient reduced order model (ROM) of the stack. A tool was developed to randomly sample selected input parameters over the expected range of interest, perform stack solutions for these sampled points using the SOFC-MP 2D software, use regression to obtain response surfaces for the stack performance and other output metrics of interest, and export the response surfaces as a ROM. The procedure is shown in Figure 1. The ROM was then successfully implemented in an Aspen Plus® system model for a natural gas fuel cell (NGFC) power system. Additional work is in progress to improve the tool’s implementation by including the fuel/oxidant recirculation loops and to better characterize the expected approximation error of the ROM. The developed modeling approach, results of the system model integration, and ongoing efforts to improve the ROM accuracy will be presented. Figure 1. Overview of ROM Generation Procedure and Usage Figure 1

Development of MIMO–OFDM system and forward error correction techniques since 2000s

The multi-input multi-output–orthogonal frequency division multiplexing (MIMO–OFDM) system serves as one of the most promising techniques that support high data rate and high performance in diverse fading channel conditions. The paper presents an overview of MIMO–OFDM wireless technology, covering advances in physical-layer design, various space–time codes, space–frequency codes and efficient concatenated forward error correction techniques, along with interleaving to yield reduced bit error rate performance (BER). Further, the paper presents the evolution of various performance metrics that are reported since the 2000s in the literature. The evolution outcome is examined using a quadratic nonlinear fitting function to assess the interesting performance metrics. Since BER versus SNR analysis is found as the interesting performance study of the MIMO–OFDM system, the current status of the achieved BER under varying SNR is reviewed and presented.