Medication non-adherence is a major healthcare challenge globally, with over half of patients with chronic conditions in developed countries failing to follow their prescribed medication regimen. This can lead to poor disease outcomes, increased hospital visits, and a significant financial burden on healthcare systems [1]. These issues have driven a recent wave of research, including the development of smart adherence products [6] that can be incorporated into a patient’s daily life to monitor medication adherence. In this work, we present a radar-based system for user identification while taking medication, which extends our recent work [5]. we conducted preliminary experiments examining semi-medication-taking activities executed by 6 subjects. Our system achieved 80% accuracy in identifying who has taken the medication in a group of 3 subjects.

We quantify the classification accuracy of Neural Networks (NNs) to specific websites using only the packet size and difference in inter-packet arrival time, which are easily observable via passive attackers in the network. Our flow classification work with NNs is unique in that we do not classify traffic by application type. Rather, we observe the accuracy of various NNs classifying specific web sites using HTTP traffic over TCP. We test a diverse set of neural network structures including a fully connected network (FCN), a convolutional neural network (CNN), a long short-term memory network (LSTM), and an autoencoder network (AE). We found that CNNs consistently had the highest accuracy, typically 80-90% when using 20 million packets as training data. We suspect that individual websites generate unique traffic patterns which are discoverable using NN techniques. Our work has important privacy implications. In particular, our work supports that both packet sizes and inter-packet timing must be randomized to obtain strong web browsing privacy. Many privacy preserving techniques, such as VPNs, will require additional enhancements.

Deep learning-based recommendation systems are extensively deployed in numerous internet services, including social media, entertainment services, and search engines, to provide users with the most relevant and personalized content. Production scale deep learning models consist of large embedding tables with billions of parameters. DRAM-based recommendation systems incur a high infrastructure cost and limit the size of the deployed models. Recommendation systems based on solid-state drives (SSDs) are a promising alternative for DRAM-based systems. Systems based on SSDs can offer ample storage required for deep learning models with large embedding tables. This paper proposes SmartRec, an inference engine for deep learning-based recommendation systems that utilizes Samsung SmartSSD, an SSD with an on-board FPGA that can process data in-situ. We evaluate SmartRec with state-of-the-art recommendation models from Facebook and compare its performance and energy efficiency to a DRAM-based system on a CPU. We show SmartRec improves the energy efficiency of the recommendation inference task up to 10x in comparison to the baseline CPU implementation. In addition, we propose a novel application-specific caching system for SmartSSDs that allows the kernel on the FPGA to use its DRAM as a cache to minimize high latency SSD accesses. Finally, we demonstrate the scalability of our design by offloading the computation to multiple SmartSSDs to further improve performance.

This article proposes a novel hardware accelerator for the inference task with sparse convolutional neural networks (CNNs) by building a hardware unit to perform Image to Column ( Im2Col ) transformation of the input feature map coupled with a systolic-array-based general matrix-matrix multiplication (GEMM) unit. Our design carefully overlaps the Im2Col transformation with the GEMM computation to maximize parallelism. We propose a novel design for the Im2Col unit that uses a set of distributed local memories connected by a ring network, which improves energy efficiency and latency by streaming the input feature map only once. The systolic-array-based GEMM unit in the accelerator can be dynamically configured as multiple GEMM units with square-shaped systolic arrays or as a single GEMM unit with a tall systolic array. This dynamic reconfigurability enables effective pipelining of Im2Col and GEMM operations and attains high processing element utilization for a wide range of CNNs. Further, our accelerator is sparsity aware, improving performance and energy efficiency by effectively mapping the sparse feature maps and weights to the processing elements, skipping ineffectual operations and unnecessary data movements involving zeros. Our prototype, SPOTS, is on average 2.16 \( \times \) , 1.74 \( \times \) , and 1.63 \( \times \) faster than Gemmini, Eyeriss, and Sparse-PE, which are prior hardware accelerators for dense and sparse CNNs, respectively. SPOTS is also 78 \( \times \) and 12 \( \times \) more energy-efficient when compared to CPU and GPU implementations, respectively.

Smart solid-state drives (SmartSSDs) with onboard FPGAs are becoming mainstream, providing opportunities for near-storage computation, which is appealing for increasing the performance of data-intensive workloads such as database query processing. This paper demonstrates the performance and energy improvements by offloading the filter and aggregation operations to the FPGA on a SmartSSD with real-system experiments. We make the observation that efficiently handling null entries in the data is important. Hence, we propose a novel design to manage the metadata to handle null entries. Our real system evaluation shows that offloading the query operations to the FPGA results in 9.6× improvement in performance while consuming 10.9× less energy compared to a conventional CPU-only query processing.

Medication adherence is one of the leading factors that can make the difference between life and death, especially for patients managing chronic conditions [2]. Indeed, these issues have driven a recent wave of research, including the development of smart pill bottles that monitor when a pill is extracted. In this poster, we extend our recent work [1], where we present adaptive learning techniques for subject identification across multiple pill bottle systems. We collect inertial signals from 10 subjects taking medication pills and encode the activity signals by transforming them into 2D texture images. Then we use pre-trained Convolutional Neural Network (CNN) models for image-based classification tasks. Our approach achieved improved differentiation capacity over existing models by using deep learning models, modified through domain adaptation and transfer learning.

The COVID-19 pandemic has impacted academic life in different ways. In the mobile and pervasive computing community, there was a struggle on data collection for the evaluation of human-sensing systems. An automated and contactless solution to collect data from users at home is one way that can help in the continuation of user-centric studies. In this poster, we present a portable system for remote, in-home data collection. The system is powered by a Raspberry Pi© and input peripherals (a camera, a microphone, and a wireless receiver). Our system uses a speech interface for text-to-speech and speech-to-text conversions. The system acts as a voice-based "smart agent" that guides the user during an experiment session. We aim to use our system to collect data from a set of smart pill bottles that we previously designed for medication adherence monitoring [1] and user identification [3].

The past 40 years have seen an explosion of Radio Frequency (RF) transmitters, which motivates understanding their impacts on the natural world. The European honeybee, Apis Mellifera, has been shown to sense the Earth's magnetic field. Human Radio Frequency (RF) transmitters alter this field. For example, recent work demonstrated that human-created RF interferes with the common robin's ability to orient themselves. This work proposes an experimental design to determine if honeybees can sense RF transmissions in frequencies from 1 MHz (AM radio) to 6 GHz (WiFi). We deployed a custom-designed RF bee feeder near bee hives to test honeybees' RF sensing ability.

Accurately accounting for medication use is important for the efficacy and safety of patients and family members. Monitoring is also important for medication adherence. This work investigates identification of persons taking medication using a sensor-equipped pill bottle. The bottle is equipped with inertial and switch sensors in both the cap and body, making the added hardware unobtrusive, low-cost, and wireless. Our system uses inertial data to build a patient discrimination model using classification techniques. We evaluated the system using 16 subjects. Our results show that using binary Support Vector Machine (SVM), the system can discriminate one patient among 16 subjects with 94% accuracy, and has a 93% using a single sensor. Identifying the exact person in a set of 3 subjects has an accuracy higher than 91%.

In this work, we investigate the identification of persons taking medication using a sensor-equipped pill-bottle. The bottle embeds inertial sensors in both the cap and body, making the added hardware un-obtrusive, low-cost, and wireless. Our system uses inertial data to build a patient discrimination model using classification techniques. We evaluated the system using 16 subjects. Our results show that using binary Support Vector Machine (SVM), the system can discriminate one patient among 16 subjects with 94 % accuracy. Identifying the exact person in a set of 3 subjects has an accuracy higher than 91 %..

Accurate and continuous ammonia monitoring is important for laboratory animal studies and many other applications. Existing solutions are often expensive, inaccurate, or unsuitable for long-term monitoring. In this work, we propose a new ammonia monitoring approach that is low-power, automatic, accurate, and wireless.

Advanced driver assistance systems and, in particular automated driving offers an unprecedented opportunity to transform the safety, efficiency, and comfort of road travel. Developing such safety technologies requires an understanding of not just common highway and city traffic situations but also a plethora of widely different unusual events (e.g., object on the road way and pedestrian crossing highway, etc.). While each such event may be rare, in aggregate they represent a significant risk that technology must address to develop truly dependable automated driving and traffic safety technologies. By developing technology to scale road data acquisition to a large number of vehicles, this paper introduces a low-cost yet reliable solution, BigRoad, that can derive internal driver inputs (i.e., steering wheel angles, driving speed and acceleration) and external perceptions of road environments (i.e., road conditions and front-view video) using a smartphone and an IMU mounted in a vehicle. We evaluate the accuracy of collected internal and external data using over 140 real-driving trips collected in a 3-month time period. Results show that BigRoad can accurately estimate the steering wheel angle with 0.69 degree median error, and derive the vehicle speed with 0.65 km/h deviation. The system is also able to determine binary road conditions with 95% accuracy by capturing a small number of brakes. We further validate the usability of BigRoad by pushing the collected video feed and steering wheel angle to a deep neural network steering wheel angle predictor, showing the potential of massive data acquisition for training self-driving system using BigRoad.

The number of small wireless devices is rapidly increasing, making the radio channel efficiency in limited geographic areas (individual rooms or buildings) an important metric for MAC protocols. Many of these emerging devices have use-cases that are difficult to satisfy with current hardware solutions and channel access methods; for instance device mobility, small energy reserves, and requirements for low cost and small form factors. However, for most of these applications, such as health care monitoring or sensing, feedback to the radio device is unnecessary and unidirectional communication techniques are not only sufficient, but can also be advantageous. We propose an efficient, reliable technique for unidirectional communication, called Transmit Only (TO), that satisfies these requirements while maintaining packet throughput guarantees and reducing energy consumption. In this paper we will demonstrate the feasibility and performance of this kind of highly asymmetric, transmit-only protocol through theoretical, simulated, and experimental results.

Given the increasing popularity of wearable devices, this paper explores the potential to use wearables for steering and driver tracking. Such capability would enable novel classes of mobile safety applications without relying on information or sensors in the vehicle. In particular, we study how wrist-mounted inertial sensors, such as those in smart watches and fitness trackers, can track steering wheel usage and angle. In particular, tracking steering wheel usage and turning angle provide fundamental techniques to improve driving detection, enhance vehicle motion tracking by mobile devices and help identify unsafe driving. The approach relies on motion features that allow distinguishing steering from other confounding hand movements. Once steering wheel usage is detected, it further uses wrist rotation measurements to infer steering wheel turning angles. Our on-road experiments show that the technique is 99% accurate in detecting steering wheel usage and can estimate turning angles with an average error within 3.4 degrees.

This paper utilizes smartphone sensing of vehicle dynamics to determine driver phone use, which can facilitate many traffic safety applications. Our system uses embedded sensors in smartphones, i.e., accelerometers and gyroscopes, to capture differences in centripetal acceleration due to vehicle dynamics. These differences combined with angular speed can determine whether the phone is on the left or right side of the vehicle. Our low infrastructure approach is flexible with different turn sizes and driving speeds. Extensive experiments conducted with two vehicles in two different cities demonstrate that our system is robust to real driving environments. Despite noisy sensor readings from smartphones, our approach can achieve a classification accuracy of over 90 percent with a false positive rate of a few percent. We also find that by combining sensing results in a few turns, we can achieve better accuracy (e.g., 95 percent) with a lower false positive rate. In addition, we seek to exploit the electromagnetic field measurement inside a vehicle to complement vehicle dynamics for driver phone sensing under the scenarios when little vehicle dynamics is present, for example, driving straight on highways or standing at roadsides.

We investigate using smartphone WiFi signals to track human queues, which are common in many business areas such as retail stores, airports, and theme parks. Real-time monitoring of such queues would enable a wealth of new applications, such as bottleneck analysis, shift assignments, and dynamic workflow scheduling. We take a minimum infrastructure approach and thus utilize a single monitor placed close to the service area along with transmitting phones. Our strategy extracts unique features embedded in signal traces to infer the critical time points when a person reaches the head of the queue and finishes service, and from these inferences we derive a person's waiting and service times. We develop two approaches in our system, one is directly feature-driven and the second uses a simple Bayesian network. Extensive experiments conducted both in the laboratory as well as in two public facilities demonstrate that our system is robust to real-world environments. We show that in spite of noisy signal readings, our methods can measure service and waiting times to within a $10$ second resolution.

Wireless localization using the received signal strength (RSS) can have tremendous savings over using specialized positioning infrastructures. In this work, we explore improving RSS localization performance in multipath environments by varying the transmitter's signal power and frequency. We first derive and analyze the Cramér-Rao Lower Bound (CRLB) of RSS-based localization based on the frequency dependent path loss propagation model that considers the transmitter's signal power and frequency. The derived CRLB shows the feasibility of improving localization performance by applying frequency and power level selection for RSS-based localization. Using this analysis, we develop two new selection metrics based on the observed standard deviations of RSS as well as residuals. We then show a set of selection methods that attempt to select the combinations of power and frequencies which minimize the localization error in a representative class of localization algorithms. Our simulation results confirm the proposed selection methods can improve the localization accuracy under CRLB. Additionally, using active RFID tags, we experimentally characterize the effect of using multiple signal powers and frequencies on a wide spectrum of RSS-based algorithms. We found that the performance of all the algorithms improves when leveraging on multiple power levels and frequencies, although different algorithms present different sensitivity in terms of localization accuracy under different selection methods.

We investigate using smartphone WiFi signals to track human queues, which are common in many business areas such as retail stores, airports, and theme parks. Real-time monitoring of such queues would enable a wealth of new applications, such as bottleneck analysis, shift assignments, and dynamic workflow scheduling. We take a minimum infrastructure approach and thus utilize a single monitor placed close to the service area along with transmitting phones. Our strategy extracts unique features embedded in the signal traces to infer the critical time points when a person reaches the head of the queue and finishes service, and from these inferences we derive a person's waiting and service times. We develop a feature driven approach in our system. Extensive experiments conducted both in the laboratory demonstrate that our system is robust to queues with different waiting time. We show that in spite of noisy signal readings, our methods can measure important time periods in queue (e.g., service and waiting times) to within a $10$ second resolution.

This paper utilizes smartphone sensing of vehicle dynamics to determine driver phone use, which can facilitate many traffic safety applications. Our system uses embedded sensors in smartphones, i.e., accelerometers and gyroscopes, to capture differences in centripetal acceleration due to vehicle dynamics. These differences combined with angular speed can determine whether the phone is on the left or right side of the vehicle. Our low infrastructure approach is flexible with different turn sizes and driving speeds. Extensive experiments conducted with two vehicles in two different cities demonstrate that our system is robust to real driving environments. Despite noisy sensor readings from smartphones, our approach can achieve a classification accuracy of over $90\%$ with a false positive rate of a few percent. We also find that by combining sensing results in a few turns, we can achieve better accuracy (e.g., $95\%$) with a lower false positive rate.

This paper presents the design and evaluation of a novel distributed shared hosting approach, DMap, for managing dynamic identifier to locator mappings in the global Internet. DMap is the foundation for a fast global name resolution service necessary to enable emerging Internet services such as seamless mobility support, content delivery and cloud computing. Our approach distributes identifier to locator mappings among Autonomous Systems (ASs) by directly applying K>1 consistent hash functions on the identifier to produce network addresses of the AS gateway routers at which the mapping will be stored. This direct mapping technique leverages the reach ability information of the underlying routing mechanism that is already available at the network layer, and achieves low lookup latencies through a single overlay hop without additional maintenance overheads. The proposed DMap technique is described in detail and specific design problems such as address space fragmentation, reducing latency through replication, taking advantage of spatial locality, as well as coping with inconsistent entries are addressed. Evaluation results are presented from a large-scale discrete event simulation of the Internet with ~26,000 ASs using real-world traffic traces from the DIMES repository. The results show that the proposed method evenly balances storage load across the global network while achieving lookup latencies with a mean value of ~50 ms and 95th percentile value of ~100 ms, considered adequate for support of dynamic mobility across the global Internet.

In the emerging paradigm of pervasive computing, applications change their behaviors in response to their environmental context, which is provided by the smart objects in the Internet of Things (IoT). Due to the inherent heterogeneity of physical world objects, realizing the IoT requires service layers to fill the gap between the low level interfaces of networked objects and the applications which use them. In this paper, we show that the MobilityFirst Future Internet Architecture is an ideal platform for realizing pervasive computing in an IoT framework. In particular, MobilityFirst's identity based routing, overloaded identities, content caching and in-network compute plane are excellent building blocks for IoT applications. We then present a detailed example of a location based service built using MobilityFirst.

This work addresses the fundamental problem of distinguishing between a driver and passenger using a mobile phone, which is the critical input to enable numerous safety and interface enhancements. Our detection system leverages the existing car stereo infrastructure, in particular, the speakers and Bluetooth network. Our acoustic approach has the phone send a series of customized high frequency beeps via the car stereo. The beeps are spaced in time across the left, right, and if available, front and rear speakers. After sampling the beeps, we use a sequential change-point detection scheme to time their arrival, and then use a differential approach to estimate the phone's distance from the car's center. From these differences a passenger or driver classification can be made. To validate our approach, we experimented with two kinds of phones and in two different cars. We found that our customized beeps were imperceptible to most users, yet still playable and recordable in both cars. Our customized beeps were also robust to background sounds such as music and wind, and we found the signal processing did not require excessive computational resources. In spite of the cars' heavy multipath environment, our approach had a classification accuracy of over 90 percent, and around 95 percent with some calibrations. We also found, we have a low false positive rate, on the order of a few percent.

This work addresses the fundamental problem of distinguishing between a driver and passenger using a mobile phone, which is the critical input to enable numerous safety and interface enhancements. Our detection system leverages the existing car stereo infrastructure, in particular the speakers and Bluetooth network. Our acoustic approach has the phone send a series of customized high frequency beeps via the car stereo. The beeps are spaced in time across the left, right, and if available, front and rear speakers. After sampling the beeps, we use a sequential change-point detection scheme to time their arrival, and then use a differential approach to estimate the phone's distance from the car's center. From these differences a passenger or driver classification can be made. To validate our approach, we experimented with two kinds of phones and in two different cars. We found that our customized beeps were imperceptible to most users, yet still playable and recordable in both cars. Our customized beeps were also robust to background sounds such as music and wind, and we found the signal processing did not require excessive computational resources. In spite of the cars' heavy multi-path environment, our approach had a classification accuracy of over 90%, and around 95% with some calibrations. We also found we have a low false positive rate, on the order of a few percent.

In this paper, we consider the problem of tracking fine-grained speeds variations of vehicles using signal strength traces from GSM enabled phones. Existing speed estimation techniques using mobile phone signals can provide longer-term speed averages but cannot track short-term speed variations. Understanding short-term speed variations, however, is important in a variety of traffic engineering applications-for example, it may help distinguish slow speeds due to traffic lights from traffic congestion when collecting real time traffic information. Using mobile phones in such applications is particularly attractive because it can be readily obtained from a large number of vehicles. Our approach is founded on the observation that the large-scale path loss and shadow fading components of signal strength readings (signal profile) obtained from the mobile phone on any given road segment appear similar over multiple trips along the same road segment except for distortions along the time axis due to speed variations. We therefore propose a speed tracking technique that uses a Derivative Dynamic Time Warping (DDTW) algorithm to realign a given signal profile with a known training profile from the same road. The speed tracking technique then translates the warping path (i.e., the degree of stretching and compressing needed for alignment) into an estimated speed trace. Using 6.4 hours of GSM signal strength traces collected from a vehicle, we show that our algorithm can estimate vehicular speed with a median error of ± 5mph compared to using a GPS and can capture significant speed variations on road segments with a precision of 68% and a recall of 84%.

In contrast to traditional sensor networks, the "Internet of Things" focuses on interactions between humans and physical objects rather than on sensing and reporting low level information. While several middle-ware systems have been created to simplify the task of managing and aggregating data from multiple sensor networks that use different hardware and software, management of the data is not sufficient to build an Internet of Things.

In this paper, we propose a management framework for protecting large computer systems against operator mistakes. By detecting and confining mistakes to isolated portions of the managed system, our framework facilitates correct operation even by inexperienced operators. We built a prototype management system called Barricade based on our framework. We evaluate Barricade by deploying it for two different systems, a prototype Internet service and an enterprise computer infrastructure, and conducting experiments with 20 volunteer operators. Our results are very promising. For example, we show that Barricade can detect and contain 39 out of the 43 mistakes that we observed in 49 live operator experiments performed with our Internet service.

This paper introduces an algorithm that estimates the speed of a mobile phone by matching time-series signal strength data to a known signal strength trace from the same road. Knowing a mobile phone's speed is useful, for example, to estimate traffic congestion or other transportation performancemetrics. The proposed algorithmcan be implemented in the carrier's infrastructure with Network Measurement Reports obtained by a base station or on a mobile phone with signal strength readings obtained by the handset and depending on implementation choices, promises lower energy consumption than Global Positioning System (GPS) receivers. We evaluate the effectiveness of our algorithm on highway and arterial roads using GSM signal strength traces obtained from several phones over a one month period. The results show that the Correlation algorithm is significantly more accurate than existing techniques based on handoffs or phone localization.

Wireless localization using the received signal strength (RSS) can have tremendous savings over using specialized positioning infrastructures. In this work, we explore improving RSS localization performance in multipath environments by varying the transmitter's signal power and frequency. Using a theoretical analysis, we first show how selection of different signal powers and frequencies can improve localization accuracy for the least squares algorithm. We next develop a set of selection methods that attempt to select the combinations of power and frequencies which minimize the localization error. Our selection methods are based on the observed standard deviations of RSS as well as algorithm specific residuals. Using active RFID tags, we experimentally characterize the effect of using multiple signal powers and frequencies on a wide spectrum of RSS-based algorithms. We found that the performance of all the algorithms improves when leveraging on multiple power levels and frequencies, although different algorithms present different sensitivity in terms of localization accuracy under different selection methods.

We explore the problem of detecting whether a device has moved within a room. Our approach relies on comparing summaries of received signal strength measurements over time, which we call descriptors. We consider descriptors based on the differences in the mean, standard deviation, and histogram comparison. In close to 1000 mobility events we conducted, our approach delivers perfect recall and near perfect precision for detecting mobility at a granularity of a few seconds. It is robust to the movement of dummy objects near the transmitter as well as people moving within the room. The detection is successful because true mobility causes fast fading, while environmental mobility causes shadow fading, which exhibit considerable difference in signal distributions. The ability to produce good detection accuracy throughout the experiments also demonstrates that our approach can be applied to varying room environments and radio technologies, thus enabling novel security, health care, and inventory control applications.

We show that signal strength variability can be reduced by employing multiple low-cost antennas at fixed locations. We further explore the impact of this reduction on wireless localization by analyzing a representative set of algorithms ranging from fingerprint matching, to statistical maximum likelihood estimation, to threshold bounding of signal fingerprints, and to multilateration. Using an indoor wireless testbed, we provide experimental evaluation of the localization performance under multiple antennas. We found that in nearly all cases the performance of localization algorithms improved when using multiple antennas. Specifically, the median and the 90th percentile error can be reduced up to 70 percent. Additionally, we found that multiple antennas improve the localization stability significantly, up to 100 percent improvement, when there are small-scale three-dimensional movements of a mobile device around a given location.

Particle filters have been found to be effective in tracking mobile targets in indoor environments. One frequently encountered problem in these settings occurs when the target's movement pattern changes unexpectedly; such as when the target turns around, enters a room from a corridor or turns left or right at an intersection. If the particle filter makes an incorrect prediction, it might not be able to recover using the normal techniques of prediction, weight update and resampling. We propose an approach to automatically restart the particle filter by sampling the latest trusted observation when the particle cloud diverges too much from the observations. The restart decision is based on Kullback-Leibler divergence between the probability surfaces associated with the current observation and the particle cloud. Through an experimental study we show that the restart algorithm allows the successful early recovery of stranded particle filters, in our scenarios providing a 36% average improvement in localization accuracy.

Particle filters are frequently used to track mobile targets in indoor environments. However, standard particle filters encounter problems tracking targets facing decisions involving divergent choices such as intersections of corridors. The target either turns to the right or left, intermediate values are not possible. The available observations might not be (at least initially) sufficient to decide which choice was taken by the target. If the prediction model takes the wrong decision, the model will diverge very quickly from the real target location. In this paper we present a modified particle filter which tracks multiple hypotheses about the decisions made by the target. Whenever the target faces a decision, the particle cloud is split by a predefined, possibly probabilistic, hypothesis modifier. The resulting particle clouds have their own prediction model but they share the weight update and resampling step. This separation lasts until the observations can conclusively identify one of the hypotheses as the correct one, or until the hypotheses converge. Our approach uses measurement of wireless media signal strengths to provide the input necessary for the localization using the GRAIL system. We validate our model through experiments covering several movement and decision scenarios typical in indoor environments.

We are considering the problem of dynamic localization of human targets in an indoor environment, such as an office building, where GPS signals are not receivable. Previous work has shown that static localization is possible through the measurement of the wireless signal strengths. Dynamic localization (tracking) can be achieved by performing periodic static localizations and filling in the gaps through an appropriate filtering technique. We are using a sampling-importance-resampling particle filter which is a probabilistic reasoning technique for this purpose.

This work investigates the lower bounds of wireless localization accuracy using signal strength on commodity hardware. Our work relies on trace-driven analysis using an extensive indoor experimental infrastructure. First, we report the best experimental accuracy, twice the best prior reported accuracy for any localization system. We experimentally show that adding more and more resources (e.g., training points or landmarks) beyond a certain limit, can degrade the localization performance for lateration-based algorithms, and that it could only be improved further by "cleaning" the data. However, matching algorithms are more robust to poor quality RSS measurements. We next compare with a theoretical lower bound using standard Cramer Rao Bound (CRB) analysis for unbiased estimators, which is frequently used to provide bounds on localization precision. Because many localization algorithms are based on different mathematical foundations, we apply a diverse set of existing algorithms to our packet traces and found that the variance of the localization errors from these algorithms are smaller than the variance bound established by the CRB. Finally, we found that there exists a wide discrepancy from what free- space models predict in the signal to distance function even in an environment with limited shadowing and multipath, thereby imposing a fundamental limit on the achievable localization accuracy indoors.

Operator mistakes are a significant source of unavailability in Internet services. In our previous work, we proposed operator action validation as an approach for detecting mistakes while hiding them from the service and its users. Previous validation strategies have limitations, however, including the need for instances of correct behavior for comparison. In this paper, we propose a novel model-based validation strategy that addresses these limitations and complements our previous techniques. Model-based validation calls for service engineers to define models of Internet services that can be used to differentiate between correct and incorrect configurations and behaviors. These models are then used to guide the specification of validation assertions that check the correctness of operator actions before they are exposed. We have implemented a prototype model-based validation system for two services, the Web crawler of a commercial search engine (Ask.com) and an academic yet realistic online auction service. Experimentation with model-based validation demonstrates that it is highly effective at detecting and hiding both activated and latent mistakes.

We present the DECODE technique to determine whether a set of transmitters are comoving, i.e., moving together in close proximity. Comovement information can find use in applications ranging from inventory tracking to social network sensing and to optimizing mobile device localization. The positioning errors from indoor RSS-based localization systems tend to be too large, making it difficult to detect whether two devices are moving together based on the interdevice distances. DECODE achieves accurate comovement detection by exploiting the correlations in positioning errors over time. DECODE can not only be implemented in the position space but also in the signal space where a correlation in shadow fading due to objects blocking the path between the transmitter and receiver exists. This technique requires no change in or cooperation from the tracked devices other than sporadic transmission of packets. Using experiments from an office environment, we show that DECODE can achieve near-perfect comovement detection at walking speed mobility using correlation coefficients computed over approximately 60-second time intervals. We further show that DECODE is generic and could accomplish detection for mixed mobile transmitters of different technologies (IEEE 802.11b/g and IEEE 802.15.4), and our results are not very sensitive to the frequency at which transmitters communicate.

Recently, it has been noted that localization algorithms that use signal strength are susceptible to noncryptographic attacks, which consequently threatens their viability for sensor applications. In this work, we examine several localization algorithms and evaluate their robustness to attacks where an adversary attenuates or amplifies the signal strength at one or more landmarks. We study both point-based and area-based methods that employ received signal strength for localization, and propose several performance metrics that quantify the estimator's precision, bias, and error, including Hölder metrics, which quantify the variability in position space for a given variability in signal strength space. We then conduct a trace-driven evaluation of a set of representative algorithms, where we measured their performance as we applied attacks on real data from two different buildings. We found the median error degraded gracefully, with a linear response as a function of the attack strength. We also found that area-based algorithms experienced a decrease and a spatial-shift in the returned area under attack, implying that precision increases though bias is introduced for these schemes. Additionally, we observed similar values for the average Hölder metric across most of the algorithms, thereby providing strong experimental evidence that nearly all the algorithms have similar average responses to signal strength attacks with the exception of the Bayesian Networks algorithm.

We present the DECODE technique to determine from a remote receiver whether a set of transmitters are co-moving, i.e., moving together in close proximity. Co-movement information can find use in applications ranging from inventory tracking, to social network sensing, and to optimizing mobile device localization. DECODE detects co-moving transmitters by identifying correlations in communication signal strength due to shadow fading. Unlike localization systems, it can operate using measurements from only a single receiver. It requires no changes in or cooperation from the tracked devices other than sporadic transmission of packets. Using experiments from an office environment, we show that DECODE can achieve near perfect co-movement detection at walking-speed mobility using correlation coefficients computed over approximately 60-second time intervals.

We show that signal strength variability can be reduced by employing multiple low-cost antennas at fixed locations. We further explore the impact of this reduction on wireless localization by analyzing a representative set of algorithms ranging from fingerprint matching, to statistical maximum likelihood estimation, and to multilateration. We provide experimental evaluation using an indoor wireless testbed of the localization performance under multiple antennas. We found that in nearly all cases the performance of localization algorithms improved when using multiple antennas. Specifically, the median and the 90th percentile error can be reduced up to 70%. Additionally, we found that multiple antennas improve the localization stability significantly, up to 100% improvement, when there are small scale 3-dimensional movements of a mobile device around a given location.

We present two parallel algorithms and their Unified Parallel C implementations for Bayesian indoor positioning systems. Our approaches are founded on Markov Chain Monte Carlo simulations. We evaluated two basic partitioning schemes: inter-chain partitioning which distributes entire Markov chains to different processors, and intra-chain which distributes a single chain across processors. Evaluations on a 16-node symmetric multiprocessor, a 4-node cluster comprising of quad processors, and a 16 single- processor-node cluster, suggest that for short chains intra- chain scales well on the first two platforms with speedups of up to 12. On the other hand, inter-chain gives speedups of 12 only for very long chains, sometimes of up to 60,000 iterations, on all three platforms. We used the LogGP model to analyze our algorithms and predict their performance. Model predictions for inter-chain are within 5% of the actual execution time, while for intra-chain they are 7%-25% less due to load imbalance not captured in the model.

Accurately positioning nodes in wireless and sensor networks is important because the location of sensors is a critical input to many higher-level networking tasks. However, the localization infrastructure can be subjected to non-cryptographic attacks, such as signal attenuation and amplification, that cannot be addressed by traditional security services. We propose several attack detection schemes for wireless localization systems. We first formulate a theoretical foundation for the attack detection problem using statistical significance testing. Next, we define test metrics for two broad localization approaches: multilateration and signal strength. We then derived both mathematical models and analytic solutions for attack detection for any system that utilizes those approaches. We also studied additional test statistics that are specific to a diverse set of algorithms. Our trace-driven experimental results provide strong evidence of the effectiveness of our attack detection schemes with high detection rates and low false positive rates across both an 802.11 (WiFi) network as well as an 802.15.4 (ZigBee) network in two real office buildings. Surprisingly, we found that of the several methods we describe, all provide qualitatively similar detection rates which indicate that the different localization systems all contain similar attack detection capability.

We present a new method for localization in wireless networks based on the measurement of the received signal strength (RSS) from multiple access points in an indoor setting. Our approach starts by learning a smoothed RSS surface for each of the access points from a set of training data. We then extract the isometric lines of the RSS surface (iso-RSS lines) for each access point. To perform the localization, the user measures the incoming signal strength of each access point and identifies the corresponding iso-RSS line. Ideally, the exact location would be the common intersection point of these lines. However, noise and measurement imperfections make the lines not intersect in a single point. We search for the smallest rectangular area which is intersected by all the selected iso-RSS lines. This rectangle is interpreted as the most likely location of the user; the area of the rectangle is an estimate of the localization error. We describe an efficient method for finding the minimal intersection area based on recursive grid partitioning. Through experiments over multiple indoor data sets we show that our approach provides a better localization accuracy than existing localization algorithms.

Operator mistakes have been identified as a significant source of unavailability in Internet services. In this paper, we propose a new language, A, for service engineers to write assertions about expected behaviors, proper configurations, and proper structural characteristics. This formalized specification of correct behavior can be used to bolster system understanding, as well as help to flag operator mistakes in a distributed system. Operator mistakes can be caused by anything from static misconfiguration to physical placement of wires and machines. This language, along with its associated runtime system, seeks to be flexible and robust enough to deal with the wide array of operator mistakes while maintaining a simple interface for designers or programmers.

We investigate the impact of landmark placement on localization performance using a combination of analytic and experimental analysis. For our analysis, we have derived an upper bound for the localization error of the linear least squares algorithm. This bound reflects the placement of landmarks as well as measurement errors at the landmarks. We next develop a novel algorithm, maxL minE, that using our analysis, finds a pattern for landmark placement that minimizes the maximum localization error. To show our results are applicable to a variety of localization algorithms, we then conducted a series of localization experiments using both an 802.11 (WiFi) network as well as an 802.15.4 (ZigBee) network in a real building environment. We use both received signal strength (RSS) and time-of-arrival (ToA) as ranging modalities. Our experimental results show that our landmark placement algorithm is generic because the resulting placements improve localization performance across a diverse set of algorithms, networks, and ranging modalities

In this work we show how to reduce the computational cost of using Bayesian networks for localization. We investigate a range of Monte Carlo sampling strategies, including Gibbs and Metropolis. We found that for our Gibbs samplers, most of the time is spent in slice sampling. Moreover, our results show that although uniform sampling over the entire domain suffers occasional rejections, it has a much lower overall computational cost than approaches that carefully avoid rejections. The key reason for this efficiency is the flatness of the full conditionals in our localization networks. Our sampling technique is also attractive because it does not require extensive tuning to achieve good performance, unlike the Metropolis samplers. We demonstrate that our whole domain sampling technique converges accurately with low latency. On commodity hardware our sampler localizes up to 10 points in less than half a second, which is over 10 times faster than a common general-purpose Bayesian sampler. Our sampler also scales well, localizing 51 objects with no location information in the training set in less than 6 seconds. Finally, we present an analytic model that describes the number of evaluations per variable using slice sampling. The model allows us to analytically determine how flat a distribution should be so that whole domain sampling is computationally more efficient when compared to other methods

No abstract available.

In this paper we present a simple ray-sector model of signal strength for indoor 802.11 networks. Signal strength is an important parameter for a variety of important wireless networking tasks, such as localization and topology control. A sufficiently accurate, yet generic, method of generating signal strength maps is needed in order to accurately simulate, design and evaluate these systems. Our ray-sector model constructs signal maps by adding signal bias with sectors defined by rings and randomized rays, using a traditional log-linear decay model as a baseline. We show our model generates distortions similar to measured radio maps by quantitatively comparing the behavior of micro-benchmarks using maps from two buildings and those generated by our model. Finally, we demonstrate the utility of our model for higher-level applications by showing it accurately predicts the performance for two dissimilar localization algorithms

Using existing wireless communication networks as a localization infrastructure promises enormous cost and deployment savings over specific localization infrastructures. In this work we investigate a Bayesian network approach that uses a combination of radio signal strength (RSS) to distance estimation along with angle-of-arrival (AoA) information. We characterize the resulting localization accuracy using data collected outdoors using different radios, indoor data, and simulated data. We show how the localization performance degrades in indoor environments and analyze the different sources of errors that cause this performance degradation as compared to outdoor settings. We found our network is quite sensitive to variations in the distance to signal strength, and the additional angle information had only a small impact on localization accuracy.

Online services are rapidly becoming the supporting infrastructure for numerous users' work and leisure, placing higher demands on their availability and correct functioning. Increasingly, these services are comprised of complex conglomerates of distributed hardware and software components. Added to this complexity, these services evolve quite frequently accumulating considerable heterogeneity within them, while allowing little time for their in-depth understanding by service personnel. Thus, it is not surprising that mistakes by service operators are common, and have been deemed to be the primary cause of service downtime.

In this paper, we propose a two-phase methodology for systematically evaluating the performability (performance and availability) of cluster-based Internet services. In the first phase, evaluators use a fault-injection infrastructure to characterize the service's behavior in the presence of faults. In the second phase, evaluators use an analytical model to combine an expected fault load with measurements from the first phase to assess the service's performability. Using this model, evaluators can study the service's sensitivity to different design decisions, fault rates, and other environmental factors. To demonstrate our methodology, we study the performability of a multitier Internet service. In particular, we evaluate the performance and availability of three soft state maintenance strategies for an online bookstore service in the presence of seven classes of faults. Among other interesting results, we clearly isolate the effect of different faults, showing that the tier of Web servers is responsible for an often dominant fraction of the service unavailability. Our results also demonstrate that storing the soft state in a database achieves better performability than storing it in main memory (even when the state is efficiently replicated) when we weight performance and availability equally. Based on our results, we conclude that service designers may want an unbalanced system in which they heavily load highly available components and leave more spare capacity for components that are likely to fail more often.

In this work, we explore active tapes, a novel sensor network architecture. An active tape is a sequence of sensor nodes and related units (such as batteries) organized around a bus. A bus of programmable sensor nodes is, in effect, a programmable linear array, thus the term active tape. A bus provides a simple mechanism to share resources. In a sensor network context, the primary sharing concerns center around energy, sensing and networking. This paper serves as an introduction and initial exploration into the design space of active tapes.

We characterize the fundamental limits of localization using signal strength in indoor environments. Signal strength approaches are attractive because they are widely applicable to wireless sensor networks and do not require additional localization hardware. We show that although a broad spectrum of algorithms can trade accuracy for precision, none has a significant advantage in localization performance. We found that using commodity 802.11 technology over a range of algorithms, approaches and environments, one can expect a median localization error of 10 ft and 97th percentile of 30 ft. We present strong evidence that these limitations are fundamental and that they are unlikely to transcend without a fundamentally more complex environmental models or additional localization infrastructure.

We present a distributed algorithm for maximizing 1-hop broadcast coverage in dense wireless sensor networks. Our strategy is built upon an analytic model that predicts the optimal range for maximizing 1-hop broadcast coverage given information like network density and node sending rate. The algorithm allows each node to set the maximizing radio range using only the locally observed sending rate and node density. The algorithm is thus critically dependent on the empirical determination of these parameters. Our algorithm can observe the parameters using only message eavesdropping and thus does not require extra protocol messages. Using simulation, we show that in spite of many simplifications in the model and incomplete density information in a live network, our algorithm converges fairly quickly and provides good coverage for both uniform and non-uniform networks across a wide range of conditions. We also demonstrate the utility of our algorithm for higher layer protocols by showing that it significantly improves the reception rate for a flooding application as well as the performance of a localization protocol.

We characterize the fundamental limits of localization using signal strength in indoor environments. Signal strength approaches are attractive because they are widely applicable to wireless sensor networks and do not require additional localization hardware. We show that although a broad spectrum of algorithms can trade accuracy for precision, none has a significant advantage in localization performance. We found that using commodity 802.11 technology over a range of algorithms, approaches and environments, one can expect a median localization error of 10ft and 97th percentile of 30ft. We present strong evidence that these limitations are fundamental and that they are unlikely to be transcended without fundamentally more complex environmental models or additional localization infrastructure.

In this paper, we evaluate the performance, availability, and combined performability of four soft state maintenance strategies in two multitier Internet services, an online book store and an auction service. To take soft state and service latency into account, we propose an extension of our previous quantification methodology, and novel availability and performability metrics. Our results demonstrate that storing the soft state in a database can achieve better performability than storing it in main memory, even when the state is efficiently replicated. Strategies that offload the handling of soft state from the database increase the load on other tiers and, consequently, increase the impact of faults in these tiers on service availability. Based on these results, we conclude that service designers need to provision the cluster and balance the load with availability and cost, as well as performance, in mind.

We present a new approach that uses compiler-directed fault-injection for coverage testing of recovery code in Internet services, to evaluate their robustness to operating system and I/O hardware faults. We define a set of program-fault coverage metrics that enable quantification of Java catch blocks exercised during fault-injection experiments. We use compiler analyses to instrument application code in two ways: to direct fault injection to occur at appropriate points during execution, and to measure the resulting coverage. As a proof of concept for these ideas, we have applied our techniques manually to Muffin, a proxy server; we obtained a high degree of coverage of catch blocks, with on average 85% of the expected faults per catch being experienced as caught exceptions.

We consider the impact of different communication architectures on the performability (performance plus availability) of cluster-based servers. In particular, we use a combination of fault-injection experiments and analytic modeling to evaluate the performability of two popular communication protocols, TCP and VIA, as the intra-cluster communication substrate of a sophisticated Web server. Our analysis leads to several interesting conclusions, the most surprising of which is, under the same fault load, VIA-based servers deliver greater availability than TCP-based servers. If we assume higher fault rates for VIA-based servers because the underlying technology is more immature and programming model more complex, we find that packet errors or application faults would have to occur at approximately 4 times the rate in TCP-based servers before their performabilities equalize. We use our results from the study to suggest that high-performance and robust communication layers for highly available cluster-based servers should preserve message boundaries, as opposed to using byte streams, use single-copy transfers, pre-allocate channel resources, and report errors in manner consistent with the network fabric's fault model.

We introduce PlanetP, content addressable publish/subscribe service for unstructured peer-to-peer (P2P) communities. PlanetP supports content addressing by providing: (1) a gossiping layer used to globally replicate a membership directory and an extremely compact content index; and (2) a completely distributed content search and ranking algorithm that help users find the most relevant information. PlanetP is a simple, yet powerful system for sharing information. PlanetP is simple because each peer must only perform a periodic, randomized, point-to-point message exchange with other peers. PlanetP is powerful because it maintains a globally content-ranked view of the shared data. Using simulation and a prototype implementation, we show that PlanetP achieves ranking accuracy that is comparable to a centralized solution and scales easily to several thousand peers while remaining resilient to rapid membership changes.

Cluster-based servers can substantially increase performance when nodes cooperate to globally manage resources. However, in this paper we show that cooperation results in a substantial availability loss, in the absence of high-availability mechanisms. Specifically, we show that a sophisticated cluster-based Web server, which gains a factor of 3 in performance through cooperation, increases service unavailability by a factor of 10 over a non-cooperative version. We then show how to augment this Web server with software components embodying a small set of high-availability techniques to regain the lost availability. Among other interesting observations, we show that the application of multiple high-availability techniques, each implemented independently in its own subsystem, can lead to inconsistent recovery actions. We also show that a novel technique called Fault Model Enforcement can be used to resolve such inconsistencies. Augmenting the server with these techniques led to a final expected availability of close to 99.99%.

No abstract available.

We consider the problem of increasing the availability of shared data in peer-to-peer systems. In particular, we conservatively estimate the amount of excess storage required to achieve a practical availability of 99.9% by studying a decentralized algorithm that only depends on a modest amount of loosely synchronized global state. Our algorithm uses randomized decisions extensively together with a novel application of an erasure code to tolerate autonomous peer actions as well as staleness in the loosely synchronized global state. We study the behavior of this algorithm in three distinct environments modeled on previously reported measurements. We show that while peers act autonomously, the community as a whole will reach a stable configuration. We also show that space is used fairly and efficiently, delivering three times availability at a cost of six times the storage footprint of the data collection when the average peer availability is only 24%.

We demonstrate a framework for improving the availability of cluster based Internet services. Our approach models Internet services as a collection of interconnected components, each possessing well defined interfaces and failure semantics. Such a decomposition allows designers to engineer high availability based on an understanding of the interconnections and isolated fault behavior of each component, as opposed to ad-hoc methods. In this work, we focus on using the entire commodity workstation as a component because it possesses natural, fault-isolated interfaces. We define a failure event as a reboot because not only is a workstation unavailable during a reboot, but also because reboots are symptomatic of a larger class of failures, such as configuration and operator errors. Our observations of 3 distinct clusters show that the time between reboots is best modeled by a Weibull distribution with shape parameters of less than 1, implying that a workstation becomes more reliable the longer it has been operating. Leveraging this observed property, we design an allocation strategy which withholds recently rebooted workstations from active service, validating their stability before allowing them to return to service. We show via simulation that this policy leads to a 70-30 rule-of-thumb: For a constant utilization, approximately 70% of the workstation failures can be masked from end clients with 30% extra capacity added to the cluster, provided reboots are not strongly correlated. We also found our technique is most sensitive to the burstiness of reboots as opposed to absolute lengths of workstation uptimes.

This work quantifies how persistent increases in processor speed compared to I/O speed reduce the performance gap between specialized, high performance messaging layers and general purpose protocols such as TCP/IP and UDP/IP. The comparison is important because specialized layers sacrifice considerable system connectivity and robustness to obtain increased performance. We first quantify the scaling effects on small messages by measuring the LogP performance of two Active Message II layers, one running over a specialized VIA layer and the other over stock UDP as we scale the CPU and I/O components. We then predict future LogP performance by mapping the LogP model's network parameters, particularly overhead into architectural components. Our projections show that the performance benefit afforded by specialized messaging for small messages will erode to a factor of 2 in the next 5 years. Our models further show that the performance differential between the two approaches will continue to erode without a radical restructuring of the I/O system. For long messages, we quantify the variable per-page instruction budget that a zero-copy messaging approach has for page table manipulations if it is to outperform a single-copy approach. Finally we conclude with an examination of future I/O advances that would result in substantial improvements to messaging performance.

We present a study of the architectural requirements and scalability of the NAS Parallel Benchmarks.Through direct measurements and simulations, we identify the factors which affect the scalability of benchmark codes on two relevant and distinct platforms; a cluster of workstations and a ccNUMA SGI Origin 2000.We find that the benefit of increased global cache size is pronounced in certain applications and often offsets the communication cost.By constructing the working set profile of the benchmarks, we are able to visualize the improvement of computational efficiency under constant-problem-size scaling.We also find that, while the Origin MPI has better point-to-point performance, the cluster MPI layer is more scalable with communication load.However, communication performance within the applications is often much lower than what would be achieved by microbenchmarks.We show that the communication protocols used by MPI runtime library are influential to the communication performance in applications, and that the benchmark codes have a wide spectrum of communication requirements.

This paper examines NFS sensitivity to performance characteristics of emerging networks.We adopt an unusual method of inserting controlled delays into live systems to measure sensitivity to basic network parameters.We develop a simple queuing model of an NFS server and show that it reasonably characterizes our two live systems running the SPECsfs benchmark.Using the techniques in this work, we can infer the structure of servers from published SPEC results.Our results show that NFS servers are most sensitive to processor overhead; it can be the limiting factor with even a modest number of disks.Continued reductions in processor overhead will be necessary to realize performance gains from future multigigabit networks.NFS can tolerate network latency in the regime of newer LANs and IP switches.Due to NFS's historic high mix of small metadata operations, NFS is quite insensitive to network bandwidth.Finally, we find that the protocol enhancements in NFS version 3 tolerate high latencies better than version 2 of the protocol.

In this paper, we study how to minimize the latency of a message through a network that consists of a number of store-and-forward stages. This research is especially relevant for today's low overhead communication systems that employ dedicated processing elements for protocol processing. We develop an abstract pipeline model that reveals a crucial performance tradeoff involving the effects of the overhead of the bottleneck stage and the bandwidth of the remaining stages. We exploit this tradeoff to develop a suite of fragmentation algorithms designed to minimize message latency. We also provide an experimental methodology that enables the construction of customized pipeline algorithms that can adapt to the specific system characteristics and application workloads. By applying this methodology to the Myrinet-GAM system, we have improved its latency by up to 51%. Our theoretical framework is also applicable to pipelined systems beyond the context of high speed networks.

This work provides a systematic study of the impact of communication performance on parallel applications in a high performance network of workstations. We develop an experimental system in which the communication latency, overhead, and bandwidth can be independently varied to observe the effects on a wide range of applications. Our results indicate that current efforts to improve cluster communication performance to that of tightly integrated parallel machines results in significantly improved application performance. We show that applications demonstrate strong sensitivity to overhead, slowing down by a factor of 60 on 32 processors when overhead is increased from 3 to 103 µs. Applications in this study are also sensitive to per-message bandwidth, but are surprisingly tolerant of increased latency and lower per-byte bandwidth. Finally, most applications demonstrate a highly linear dependence to both overhead and per-message bandwidth, indicating that further improvements in communication performance will continue to improve application performance.

In this paper, we analyze four parallel sorting algorithms (bitonic, column, radix, and sample sort) with the LogP model. LogP characterizes the performance of modern parallel machines with a small set of parameters: the communication latency (L), overhead (o), bandwidth (g), and the number of processors (P). We develop implementations of these algorithms in Split-C, a parallel extension to C, and compare the performance predicted by LogP to actual performance on a CM-5 of 32 to 512 processors for a range of problem sizes. We evaluate the robustness of the algorithms by varying the distribution and ordering of the key values. We also briefly examine the sensitivity of the algorithms to the communication parameters. We show that the LogP model is a valuable guide in the development of parallel algorithms and a good predictor of implementation performance. The model encourages the use of data layouts which minimize communication and balanced communication schedules which avoid contention. With an empirical model of local processor performance, LogP predictions closely match observed execution times on uniformly distributed keys across a broad range of problem and machine sizes. We find that communication performance is oblivious to the distribution of the key values, whereas the local processor performance is not; some communication phases are sensitive to the ordering of keys due to contention. Finally, our analysis shows that overhead is the most critical communication parameter in the sorting algorithms.