Emmanuel Agu

The effects of loss and latency on user performance in unreal tournament 2003

The growth in the popularity of interactive network games has increased the importance of a better understanding of the effects of packet loss and latency on user performance. While previous work on network games has studied user tolerance for high latencies and has studied the effects of latency on user performance in real-time strategy games, to the best of our knowledge, there has been no systematic study of the effects of loss and latency on user performance. In this paper we study user performance for Unreal Tournament 2003 (UT2003), a popular first person shooter game, under varying amounts of packet loss and latency. First, we deduced typical real world values of packet loss and latency experienced on the Internet by monitoring numerous operational UT2003 game servers. We then used these deduced values of loss and latency in a controlled networked environment that emulated various conditions of loss and latency, allowing us to monitor UT2003 at the network, application and user levels. We designed maps that isolated the fundamental first person shooter interaction components of movement and shooting, and conducted numerous user studies under controlled network conditions. We find that typical ranges of packet loss have no impact on user performance or on the quality of game play. The levels of latency typical for most UT2003 Internet servers, while sometimes unpleasant, do not significantly affect the outcome of the game. Since most first person shooter games typically consist of generic player actions similar to those that we tested, we believe that these results have broader implications.

The effect of latency on user performance in Warcraft III

Latency on the Internet is a well-known problem for interactive applications. With the increase in interactive network games comes the increased importance of understanding the effects of latency on user performance. Classes of network games such as First Person Shooters (FPS) and Real Time Strategy (RTS) differ in their user interaction model and hence susceptibility to latency. While previous work has measured the effects of latency on FPS games, there has been no systematic investigation of the effects of latency on RTS games. In this work, we design and conduct user studies that measure the impact of latency on user performance in Warcraft III, a popular RTS game. As a foundation for the research, we separated typical Warcraft III user interactions into the basic components of explore, build and combat, and analyzed each individually. We find modest statistical correlations between user performance and latency for exploration, but very weak correlations for building and combat. Overall, the effect of even very high latency, while noticeable to users, has a negligible effect on the outcome of the game. We attribute this somewhat surprising result to the nature of RTS game-play that clearly favors strategy over the real-time aspects.

Mobile extensions to RSVP

RSVP (resource reservation protocol) is a receiver oriented resource reservation setup protocol targeted for integrated services packet networks (ISPNs). RSVP has a number of desirable attributes which make it a leading candidate for Internet standardization. These attributes include flexibility, robustness, and scalability. However, RSVP does not adequately address the issue of resource reservation for mobile hosts. This paper highlights the limitations of RSVP in mobile environments and proposes extensions to obviate these deficiencies. The basic characteristics of the proposed augmentations include: (1) quiescent resource reservations, (2) virtual receivers, (3) intelligent pre-allocation of resources based on analysis of user mobility patterns, and (4) predictive look-ahead dynamic dormant multicast trees.

Smartphone-Based Wound Assessment System for Patients With Diabetes

Diabetic foot ulcers represent a significant health issue. Currently, clinicians and nurses mainly base their wound assessment on visual examination of wound size and healing status, while the patients themselves seldom have an opportunity to play an active role. Hence, a more quantitative and cost-effective examination method that enables the patients and their caregivers to take a more active role in daily wound care potentially can accelerate wound healing, save travel cost and reduce healthcare expenses. Considering the prevalence of smartphones with a high-resolution digital camera, assessing wounds by analyzing images of chronic foot ulcers is an attractive option. In this paper, we propose a novel wound image analysis system implemented solely on the Android smartphone. The wound image is captured by the camera on the smartphone with the assistance of an image capture box. After that, the smartphone performs wound segmentation by applying the accelerated mean-shift algorithm. Specifically, the outline of the foot is determined based on skin color, and the wound boundary is found using a simple connected region detection method. Within the wound boundary, the healing status is next assessed based on red-yellow-black color evaluation model. Moreover, the healing status is quantitatively assessed, based on trend analysis of time records for a given patient. Experimental results on wound images collected in UMASS-Memorial Health Center Wound Clinic (Worcester, MA) following an Institutional Review Board approved protocol show that our system can be efficiently used to analyze the wound healing status with promising accuracy.

Characteristics of Mobile Web Content

The World Wide Web is no longer tethered to our desktops and laptops. The Web has gone mobile, providing instant access to information anywhere and anytime. The mobile Web can be considered a shadow of the World Wide Web, implemented using specialized markup languages and design techniques adapted for comparatively limited mobile phones and PDAs. Despite the growing importance and usage of the mobile Web, surprising little is known about it. This paper presents the results of a study of mobile Web content conducted in May and June of 2006. The study examines the content of over one-million mobile Web pages from around the world using a search-assisted crawling methodology to locate and study pages for three of the most popular mobile Web formats - WML 1.0, WML 2.0/XHTML mobile profile (XHTML-MP) and compact HTML (C-HTML). The objective is to study the relative characteristics of these mobile Web content formats, as well as compare them with a similar sampling of non-mobile (HTML) content. We found that WML is the dominant mobile Web content type, although regional differences do exist. We found that all three mobile content types studied were on the same order of magnitude for average page characteristics such as number of links (under 10) and number of images (around 1), but pages in the newest format, XHTML-MP, are 50% larger on average than those in WML. Not surprisingly, all of these characteristics are much smaller than for HTML content pages gathered with the same methodology. In terms of specific features, only 7 % of pages used WML cards, but 50% of XHTML-MP servers dynamically adapted the content served based on the user agent. Finally, we found less than 4% of mobile pages contained ad objects, which is much less than for HTML pages.

PowerSpy: fine-grained software energy profiling for mobile devices

Battery power capacity has shown very little growth, especially when compared with the exponential growths of CPU power, memory and disk space. Hence, battery power is frequently the most constraining resource on a mobile device. As a foundation for optimizing application energy usage on mobile devices, it is increasingly important to profile system-wide energy usage in order to accurately determine where the energy is going. Previous work on profiling energy usage has either required external hardware multimeters, provided coarse grain results or required modifications to the operating system or/and profiled application. We present PowerSpy, which tracks and reports the battery energy consumed by the different threads of a monitored application, the operating system, other applications in a multi-threaded environment along with I/O devices. Using PowerSpy, we are able to measure the power consumption of five diverse applications including a Web browser, VRML graphics browser, compiler and video player, all without requiring modification to the applications source code.

Physically-based real-time diffraction using spherical harmonics

Diffraction, interference, dispersive refraction and scattering are four wavelength-dependent mechanisms that produce iridescent colors. Wavelength-dependent functions need to be sampled at discrete wavelengths in the visible spectrum, which increases the computational intensity of rendering iridescence. Furthermore, diffraction requires careful sampling since its response function varies at a higher frequency variation with sharper peaks than interference or dispersive refraction. Consequently, rendering physically accurate diffraction has previously either been approximated using simplified color curves, or been limited to offline rendering techniques such as ray tracing. We propose a technique for real-time rendering of physically accurate diffraction on programmable hardware. Our technique adaptively samples the diffraction BRDF and precomputes it to Spherical Harmonic (SH) basis that preserves the peak intensity of the reflected light. While previous work on diffraction used low dynamic range lights, we preserve the full dynamic range of the incident illumination and the diffractive response over the entire hemisphere of incoming light directions. We defer conversion from a wavelength representation to a tone mapped RGB triplet until display.

Remote execution for 3D graphics on mobile devices

Mobile devices have limited processing power and wireless networks have limited bandwidth. A modern photorealistic graphics application is resource-hungry, consumes large amounts of CPU cycles, memory and network bandwidth if networked. Moreover running them on mobile devices may also diminish their battery power in the process. The majority of graphics computations are floating point operations, which can run significantly slower on mobile devices that do not have floating point units or 3D graphics accelerators. Proposed solutions such as input mesh simplification are lossy and reduce photorealism. Remote execution, wherein part or entire rendering process is offloaded to a powerful surrogate server, is an attractive solution. We propose pipeline-splitting, a paradigm whereby 15 sub-stages of the graphics pipeline are isolated and instrumented with networking code such that they can run on either a mobile client or a surrogate server. To validate our concepts, we instrument Mesa3D, a popular implementation of the OpenGL graphics to support pipeline-splitting, creating Remote Mesa (RMesa). We explore various mappings of the graphics pipeline to the client and server while monitoring key performance metrics such as total rendering time, power consumption on the client and network usage and establish conditions under which remote execution is an optimal solution. Our results show that even with the incurred roundtrip delay, our remote execution framework can improve rendering performance by up to 10 times when rendering a moderate-sized graphics mesh file.

Malicious WiFi networks: A first look

WiFi networks have become ubiquitously deployed. Anecdotal evidence suggests that wireless networks are less secure than wired networks. However, very little quantitative data exists to characterize the insecurity levels of WiFi networks. In this paper, we take a first look at city-level WiFi security statistics for eighteen cities within the United States. For the purposes of this study, we define unsecure WiFi networks as those that do not use WEP, WPA2 or any WiFi security standard. We found that on average, 45 percent of WiFis studied were insecure, with Miami, FL having 81 percent of WiFi networks unsecure. We also found that Texas had several cities with a high number of blacklisted IP addresses. In the end, we found no strong correlation between WiFi insecurity rates and blacklisting rates, which we attribute to inadequate education of the inhabitants of cities on how to report malicious activity so that perpetuators can be blacklisted. Finally, we found that the percentage of secure WiFi networks in cities with municipal WiFi networks was comparable to that of cities without municipal WiFi networks.

Wound image analysis system for diabetics

Diabetic foot ulcers represent a significant health issue, and daily wound care is necessary for wound healing to occur. The goal of this research is to create a smart phone based wound image analysis system for people with diabetes to track the healing process of chronic ulcers and wounds. This system has been implemented on an Android smart phone in collaboration with a PC (or embedded PC). The wound image is captured by the smart phone camera and transmitted to the PC via Wi-Fi for image processing. The PC converts the JPEG image to bitmap format, then performs boundary segmentation on the wound in the image. The segmentation is done with a particular implementation of the level set algorithm, the distance regularized level set evolution (DRLSE) method, which eliminates the need for re-initialization of level set function. Next, an assessment of the wound healing is performed with color segmentation within the boundaries of the wound image, by applying the K-Mean color clustering algorithm based on the red-yellow-black (RYB) evaluation model. Finally, the results are re-formatted to JPEG, transmitted back to the smart phone and displayed. To accelerate the wound image segmentation, we have implemented the DRLSE method on the GPU and CPU cooperative hardware platform in data-parallel mode, which has greatly improved the computational efficiency. Processing wound images acquired from UMASS Medical Center has demonstrated that the wound image analysis system provides accurate wounds area determination and color segmentation. For all wound images of size around 640 x 480, with complicated wound boundaries, the wound analysis consumed max 3s, which is 5 times faster than the same algorithm running on the CPU alone.