Power and thermal aware scheduling for real-time computing systems

Over the past few decades, we have been enjoying tremendous benefits thanks to the revolutionary advancement of computing systems, driven mainly by the remarkable semiconductor technology scaling and the increasingly complicated processor architecture. However, the exponentially increased transistor density has directly led to exponentially increased power consumption and dramatically elevated system temperature, which not only adversely impacts the system's cost, performance and reliability, but also increases the leakage and thus the overall power consumption. Today, the power and thermal issues have posed enormous challenges and threaten to slow down the continuous evolvement of computer technology. Effective power/thermal-aware design techniques are urgently demanded, at all design abstraction levels, from the circuit-level, the logic-level, to the architectural-level and the system-level. ^ In this dissertation, we present our research efforts to employ real-time scheduling techniques to solve the resource-constrained power/thermal-aware, design-optimization problems. In our research, we developed a set of simple yet accurate system-level models to capture the processor's thermal dynamic as well as the interdependency of leakage power consumption, temperature, and supply voltage. Based on these models, we investigated the fundamental principles in power/thermal-aware scheduling, and developed real-time scheduling techniques targeting at a variety of design objectives, including peak temperature minimization, overall energy reduction, and performance maximization. ^ The novelty of this work is that we integrate the cutting-edge research on power and thermal at the circuit and architectural-level into a set of accurate yet simplified system-level models, and are able to conduct system-level analysis and design based on these models. The theoretical study in this work serves as a solid foundation for the guidance of the power/thermal-aware scheduling algorithms development in practical computing systems.^

Leakage temperature dependency aware real-time scheduling for power and thermal optimization

Catering to society's demand for high performance computing, billions of transistors are now integrated on IC chips to deliver unprecedented performances. With increasing transistor density, the power consumption/density is growing exponentially. The increasing power consumption directly translates to the high chip temperature, which not only raises the packaging/cooling costs, but also degrades the performance/reliability and life span of the computing systems. Moreover, high chip temperature also greatly increases the leakage power consumption, which is becoming more and more significant with the continuous scaling of the transistor size. As the semiconductor industry continues to evolve, power and thermal challenges have become the most critical challenges in the design of new generations of computing systems. ^ In this dissertation, we addressed the power/thermal issues from the system-level perspective. Specifically, we sought to employ real-time scheduling methods to optimize the power/thermal efficiency of the real-time computing systems, with leakage/ temperature dependency taken into consideration. In our research, we first explored the fundamental principles on how to employ dynamic voltage scaling (DVS) techniques to reduce the peak operating temperature when running a real-time application on a single core platform. We further proposed a novel real-time scheduling method, “M-Oscillations” to reduce the peak temperature when scheduling a hard real-time periodic task set. We also developed three checking methods to guarantee the feasibility of a periodic real-time schedule under peak temperature constraint. We further extended our research from single core platform to multi-core platform. We investigated the energy estimation problem on the multi-core platforms and developed a light weight and accurate method to calculate the energy consumption for a given voltage schedule on a multi-core platform. Finally, we concluded the dissertation with elaborated discussions of future extensions of our research. ^

Leakage Temperature Dependency Aware Real-Time Scheduling for Power and Thermal Optimization

Catering to society’s demand for high performance computing, billions of transistors are now integrated on IC chips to deliver unprecedented performances. With increasing transistor density, the power consumption/density is growing exponentially. The increasing power consumption directly translates to the high chip temperature, which not only raises the packaging/cooling costs, but also degrades the performance/reliability and life span of the computing systems. Moreover, high chip temperature also greatly increases the leakage power consumption, which is becoming more and more significant with the continuous scaling of the transistor size. As the semiconductor industry continues to evolve, power and thermal challenges have become the most critical challenges in the design of new generations of computing systems. In this dissertation, we addressed the power/thermal issues from the system-level perspective. Specifically, we sought to employ real-time scheduling methods to optimize the power/thermal efficiency of the real-time computing systems, with leakage/ temperature dependency taken into consideration. In our research, we first explored the fundamental principles on how to employ dynamic voltage scaling (DVS) techniques to reduce the peak operating temperature when running a real-time application on a single core platform. We further proposed a novel real-time scheduling method, “M-Oscillations” to reduce the peak temperature when scheduling a hard real-time periodic task set. We also developed three checking methods to guarantee the feasibility of a periodic real-time schedule under peak temperature constraint. We further extended our research from single core platform to multi-core platform. We investigated the energy estimation problem on the multi-core platforms and developed a light weight and accurate method to calculate the energy consumption for a given voltage schedule on a multi-core platform. Finally, we concluded the dissertation with elaborated discussions of future extensions of our research.

Effective reconfigurable antenna designs to enhance performance and enable wireless powering

With the increase in traffic on the internet, there is a greater demand for wireless mobile and ubiquitous applications. These applications need antennas that are not only broadband, but can also work in different frequency spectrums. Even though there is a greater demand for such applications, it is still imperative to conserve power. Thus, there is a need to design multi-broadband antennas that do not use a lot of power. Reconfigurable antennas can work in different frequency spectrums as well as conserve power. The current designs of reconfigurable antennas work only in one band. There is a need to design reconfigurable antennas that work in different frequency spectrums. In this current era of high power consumption there is also a greater demand for wireless powering. This dissertation explores ideal designs of reconfigurable antennas that can improve performance and enable wireless powering. This dissertation also presents lab results of the multi-broadband reconfigurable antenna that was created. A detailed mathematical analyses, as well as extensive simulation results are also presented. The novel reconfigurable antenna designs can be extended to Multiple Input Multiple Output (MIMO) environments and military applications.^

A thermally wavelength-tunable photonic switch based on silicon microring resonator

Silicon photonics is a very promising technology for future low-cost high-bandwidth optical telecommunication applications down to the chip level. This is due to the high degree of integration, high optical bandwidth and large speed coupled with the development of a wide range of integrated optical functions. Silicon-based microring resonators are a key building block that can be used to realize many optical functions such as switching, multiplexing, demultiplaxing and detection of optical wave. The ability to tune the resonances of the microring resonators is highly desirable in many of their applications. In this work, the study and application of a thermally wavelength-tunable photonic switch based on silicon microring resonator is presented. Devices with 10μm diameter were systematically studied and used in the design. Its resonance wavelength was tuned by thermally induced refractive index change using a designed local micro-heater. While thermo-optic tuning has moderate speed compared with electro-optic and all-optic tuning, with silicon’s high thermo-optic coefficient, a much wider wavelength tunable range can be realized. The device design was verified and optimized by optical and thermal simulations. The fabrication and characterization of the device was also implemented. The microring resonator has a measured FSR of ∼18 nm, FWHM in the range 0.1-0.2 nm and Q around 10,000. A wide tunable range (>6.4 nm) was achieved with the switch, which enables dense wavelength division multiplexing (DWDM) with a channel space of 0.2nm. The time response of the switch was tested on the order of 10 μs with a low power consumption of ∼11.9mW/nm. The measured results are in agreement with the simulations. Important applications using the tunable photonic switch were demonstrated in this work. 1×4 and 4×4 reconfigurable photonic switch were implemented by using multiple switches with a common bus waveguide. The results suggest the feasibility of on-chip DWDM for the development of large-scale integrated photonics. Using the tunable switch for output wavelength control, a fiber laser was demonstrated with Erbium-doped fiber amplifier as the gain media. For the first time, this approach integrated on-chip silicon photonic wavelength control.

Enhancement of wireless networks with heterogeneous antennas

In recent years, wireless communication infrastructures have been widely deployed for both personal and business applications. IEEE 802.11 series Wireless Local Area Network (WLAN) standards attract lots of attention due to their low cost and high data rate. Wireless ad hoc networks which use IEEE 802.11 standards are one of hot spots of recent network research. Designing appropriate Media Access Control (MAC) layer protocols is one of the key issues for wireless ad hoc networks. ^ Existing wireless applications typically use omni-directional antennas. When using an omni-directional antenna, the gain of the antenna in all directions is the same. Due to the nature of the Distributed Coordination Function (DCF) mechanism of IEEE 802.11 standards, only one of the one-hop neighbors can send data at one time. Nodes other than the sender and the receiver must be either in idle or listening state, otherwise collisions could occur. The downside of the omni-directionality of antennas is that the spatial reuse ratio is low and the capacity of the network is considerably limited. ^ It is therefore obvious that the directional antenna has been introduced to improve spatial reutilization. As we know, a directional antenna has the following benefits. It can improve transport capacity by decreasing interference of a directional main lobe. It can increase coverage range due to a higher SINR (Signal Interference to Noise Ratio), i.e., with the same power consumption, better connectivity can be achieved. And the usage of power can be reduced, i.e., for the same coverage, a transmitter can reduce its power consumption. ^ To utilizing the advantages of directional antennas, we propose a relay-enabled MAC protocol. Two relay nodes are chosen to forward data when the channel condition of direct link from the sender to the receiver is poor. The two relay nodes can transfer data at the same time and a pipelined data transmission can be achieved by using directional antennas. The throughput can be improved significant when introducing the relay-enabled MAC protocol. ^ Besides the strong points, directional antennas also have some explicit drawbacks, such as the hidden terminal and deafness problems and the requirements of retaining location information for each node. Therefore, an omni-directional antenna should be used in some situations. The combination use of omni-directional and directional antennas leads to the problem of configuring heterogeneous antennas, i e., given a network topology and a traffic pattern, we need to find a tradeoff between using omni-directional and using directional antennas to obtain a better network performance over this configuration. ^ Directly and mathematically establishing the relationship between the network performance and the antenna configurations is extremely difficult, if not intractable. Therefore, in this research, we proposed several clustering-based methods to obtain approximate solutions for heterogeneous antennas configuration problem, which can improve network performance significantly. ^ Our proposed methods consist of two steps. The first step (i.e., clustering links) is to cluster the links into different groups based on the matrix-based system model. After being clustered, the links in the same group have similar neighborhood nodes and will use the same type of antenna. The second step (i.e., labeling links) is to decide the type of antenna for each group. For heterogeneous antennas, some groups of links will use directional antenna and others will adopt omni-directional antenna. Experiments are conducted to compare the proposed methods with existing methods. Experimental results demonstrate that our clustering-based methods can improve the network performance significantly. ^

VHITS: Vertical handoff initiation and target selection in a heterogeneous wireless network

Global connectivity, for anyone, at anyplace, at anytime, to provide high-speed, high-quality, and reliable communication channels for mobile devices, is now becoming a reality. The credit mainly goes to the recent technological advances in wireless communications comprised of a wide range of technologies, services, and applications to fulfill the particular needs of end-users in different deployment scenarios (Wi-Fi, WiMAX, and 3G/4G cellular systems). In such a heterogeneous wireless environment, one of the key ingredients to provide efficient ubiquitous computing with guaranteed quality and continuity of service is the design of intelligent handoff algorithms. Traditional single-metric handoff decision algorithms, such as Received Signal Strength (RSS) based, are not efficient and intelligent enough to minimize the number of unnecessary handoffs, decision delays, and call-dropping and/or blocking probabilities. This research presented a novel approach for the design and implementation of a multi-criteria vertical handoff algorithm for heterogeneous wireless networks. Several parallel Fuzzy Logic Controllers were utilized in combination with different types of ranking algorithms and metric weighting schemes to implement two major modules: the first module estimated the necessity of handoff, and the other module was developed to select the best network as the target of handoff. Simulations based on different traffic classes, utilizing various types of wireless networks were carried out by implementing a wireless test-bed inspired by the concept of Rudimentary Network Emulator (RUNE). Simulation results indicated that the proposed scheme provided better performance in terms of minimizing the unnecessary handoffs, call dropping, and call blocking and handoff blocking probabilities. When subjected to Conversational traffic and compared against the RSS-based reference algorithm, the proposed scheme, utilizing the FTOPSIS ranking algorithm, was able to reduce the average outage probability of MSs moving with high speeds by 17%, new call blocking probability by 22%, the handoff blocking probability by 16%, and the average handoff rate by 40%. The significant reduction in the resulted handoff rate provides MS with efficient power consumption, and more available battery life. These percentages indicated a higher probability of guaranteed session continuity and quality of the currently utilized service, resulting in higher user satisfaction levels.

Practical dynamic thermal management on Intel desktop computer

Fueled by increasing human appetite for high computing performance, semiconductor technology has now marched into the deep sub-micron era. As transistor size keeps shrinking, more and more transistors are integrated into a single chip. This has increased tremendously the power consumption and heat generation of IC chips. The rapidly growing heat dissipation greatly increases the packaging/cooling costs, and adversely affects the performance and reliability of a computing system. In addition, it also reduces the processor's life span and may even crash the entire computing system. Therefore, dynamic thermal management (DTM) is becoming a critical problem in modern computer system design. Extensive theoretical research has been conducted to study the DTM problem. However, most of them are based on theoretically idealized assumptions or simplified models. While these models and assumptions help to greatly simplify a complex problem and make it theoretically manageable, practical computer systems and applications must deal with many practical factors and details beyond these models or assumptions. The goal of our research was to develop a test platform that can be used to validate theoretical results on DTM under well-controlled conditions, to identify the limitations of existing theoretical results, and also to develop new and practical DTM techniques. This dissertation details the background and our research efforts in this endeavor. Specifically, in our research, we first developed a customized test platform based on an Intel desktop. We then tested a number of related theoretical works and examined their limitations under the practical hardware environment. With these limitations in mind, we developed a new reactive thermal management algorithm for single-core computing systems to optimize the throughput under a peak temperature constraint. We further extended our research to a multicore platform and developed an effective proactive DTM technique for throughput maximization on multicore processor based on task migration and dynamic voltage frequency scaling technique. The significance of our research lies in the fact that our research complements the current extensive theoretical research in dealing with increasingly critical thermal problems and enabling the continuous evolution of high performance computing systems.

Field programmable gate array based target detection and gesture recognition

The move from Standard Definition (SD) to High Definition (HD) represents a six times increases in data, which needs to be processed. With expanding resolutions and evolving compression, there is a need for high performance with flexible architectures to allow for quick upgrade ability. The technology advances in image display resolutions, advanced compression techniques, and video intelligence. Software implementation of these systems can attain accuracy with tradeoffs among processing performance (to achieve specified frame rates, working on large image data sets), power and cost constraints. There is a need for new architectures to be in pace with the fast innovations in video and imaging. It contains dedicated hardware implementation of the pixel and frame rate processes on Field Programmable Gate Array (FPGA) to achieve the real-time performance. ^ The following outlines the contributions of the dissertation. (1) We develop a target detection system by applying a novel running average mean threshold (RAMT) approach to globalize the threshold required for background subtraction. This approach adapts the threshold automatically to different environments (indoor and outdoor) and different targets (humans and vehicles). For low power consumption and better performance, we design the complete system on FPGA. (2) We introduce a safe distance factor and develop an algorithm for occlusion occurrence detection during target tracking. A novel mean-threshold is calculated by motion-position analysis. (3) A new strategy for gesture recognition is developed using Combinational Neural Networks (CNN) based on a tree structure. Analysis of the method is done on American Sign Language (ASL) gestures. We introduce novel point of interests approach to reduce the feature vector size and gradient threshold approach for accurate classification. (4) We design a gesture recognition system using a hardware/ software co-simulation neural network for high speed and low memory storage requirements provided by the FPGA. We develop an innovative maximum distant algorithm which uses only 0.39% of the image as the feature vector to train and test the system design. Database set gestures involved in different applications may vary. Therefore, it is highly essential to keep the feature vector as low as possible while maintaining the same accuracy and performance^

Deployment of a Hybrid Multicast Switch in Energy-Aware Data Center Network: A Case of Fat-Tree Topology

Recently, energy efficiency or green IT has become a hot issue for many IT infrastructures as they attempt to utilize energy-efficient strategies in their enterprise IT systems in order to minimize operational costs. Networking devices are shared resources connecting important IT infrastructures, especially in a data center network they are always operated 24/7 which consume a huge amount of energy, and it has been obviously shown that this energy consumption is largely independent of the traffic through the devices. As a result, power consumption in networking devices is becoming more and more a critical problem, which is of interest for both research community and general public. Multicast benefits group communications in saving link bandwidth and improving application throughput, both of which are important for green data center. In this paper, we study the deployment strategy of multicast switches in hybrid mode in energy-aware data center network: a case of famous fat-tree topology. The objective is to find the best location to deploy multicast switch not only to achieve optimal bandwidth utilization but also to minimize power consumption. We show that it is possible to easily achieve nearly 50% of energy consumption after applying our proposed algorithm.

Delay-sensitive service request scheduling for cloud computing

Cloud computing realizes the long-held dream of converting computing capability into a type of utility. It has the potential to fundamentally change the landscape of the IT industry and our way of life. However, as cloud computing expanding substantially in both scale and scope, ensuring its sustainable growth is a critical problem. Service providers have long been suffering from high operational costs. Especially the costs associated with the skyrocketing power consumption of large data centers. In the meantime, while efficient power/energy utilization is indispensable for the sustainable growth of cloud computing, service providers must also satisfy a user's quality of service (QoS) requirements. This problem becomes even more challenging considering the increasingly stringent power/energy and QoS constraints, as well as other factors such as the highly dynamic, heterogeneous, and distributed nature of the computing infrastructures, etc. ^ In this dissertation, we study the problem of delay-sensitive cloud service scheduling for the sustainable development of cloud computing. We first focus our research on the development of scheduling methods for delay-sensitive cloud services on a single server with the goal of maximizing a service provider's profit. We then extend our study to scheduling cloud services in distributed environments. In particular, we develop a queue-based model and derive efficient request dispatching and processing decisions in a multi-electricity-market environment to improve the profits for service providers. We next study a problem of multi-tier service scheduling. By carefully assigning sub deadlines to the service tiers, our approach can significantly improve resource usage efficiencies with statistically guaranteed QoS. Finally, we study the power conscious resource provision problem for service requests with different QoS requirements. By properly sharing computing resources among different requests, our method statistically guarantees all QoS requirements with a minimized number of powered-on servers and thus the power consumptions. The significance of our research is that it is one part of the integrated effort from both industry and academia to ensure the sustainable growth of cloud computing as it continues to evolve and change our society profoundly.^

Effective Reconfigurable Antenna Designs to Enhance Performance and Enable Wireless Powering

With the increase in traffic on the internet, there is a greater demand for wireless mobile and ubiquitous applications. These applications need antennas that are not only broadband, but can also work in different frequency spectrums. Even though there is a greater demand for such applications, it is still imperative to conserve power. Thus, there is a need to design multi-broadband antennas that do not use a lot of power. Reconfigurable antennas can work in different frequency spectrums as well as conserve power. The current designs of reconfigurable antennas work only in one band. There is a need to design reconfigurable antennas that work in different frequency spectrums. In this current era of high power consumption there is also a greater demand for wireless powering. This dissertation explores ideal designs of reconfigurable antennas that can improve performance and enable wireless powering. This dissertation also presents lab results of the multi-broadband reconfigurable antenna that was created. A detailed mathematical analyses, as well as extensive simulation results are also presented. The novel reconfigurable antenna designs can be extended to Multiple Input Multiple Output (MIMO) environments and military applications.

A Thermally Wavelength-tunable Photonic Switch Based on Silicon Microring Resonator

Silicon photonics is a very promising technology for future low-cost high-bandwidth optical telecommunication applications down to the chip level. This is due to the high degree of integration, high optical bandwidth and large speed coupled with the development of a wide range of integrated optical functions. Silicon-based microring resonators are a key building block that can be used to realize many optical functions such as switching, multiplexing, demultiplaxing and detection of optical wave. The ability to tune the resonances of the microring resonators is highly desirable in many of their applications. In this work, the study and application of a thermally wavelength-tunable photonic switch based on silicon microring resonator is presented. Devices with 10µm diameter were systematically studied and used in the design. Its resonance wavelength was tuned by thermally induced refractive index change using a designed local micro-heater. While thermo-optic tuning has moderate speed compared with electro-optic and all-optic tuning, with silicon’s high thermo-optic coefficient, a much wider wavelength tunable range can be realized. The device design was verified and optimized by optical and thermal simulations. The fabrication and characterization of the device was also implemented. The microring resonator has a measured FSR of ~18 nm, FWHM in the range 0.1-0.2 nm and Q around 10,000. A wide tunable range (>6.4 nm) was achieved with the switch, which enables dense wavelength division multiplexing (DWDM) with a channel space of 0.2nm. The time response of the switch was tested on the order of 10 us with a low power consumption of ~11.9mW/nm. The measured results are in agreement with the simulations. Important applications using the tunable photonic switch were demonstrated in this work. 1×4 and 4×4 reconfigurable photonic switch were implemented by using multiple switches with a common bus waveguide. The results suggest the feasibility of on-chip DWDM for the development of large-scale integrated photonics. Using the tunable switch for output wavelength control, a fiber laser was demonstrated with Erbium-doped fiber amplifier as the gain media. For the first time, this approach integrated on-chip silicon photonic wavelength control.

Delay-Sensitive Service Request Scheduling for Cloud Computing

Cloud computing realizes the long-held dream of converting computing capability into a type of utility. It has the potential to fundamentally change the landscape of the IT industry and our way of life. However, as cloud computing expanding substantially in both scale and scope, ensuring its sustainable growth is a critical problem. Service providers have long been suffering from high operational costs. Especially the costs associated with the skyrocketing power consumption of large data centers. In the meantime, while efficient power/energy utilization is indispensable for the sustainable growth of cloud computing, service providers must also satisfy a user's quality of service (QoS) requirements. This problem becomes even more challenging considering the increasingly stringent power/energy and QoS constraints, as well as other factors such as the highly dynamic, heterogeneous, and distributed nature of the computing infrastructures, etc. In this dissertation, we study the problem of delay-sensitive cloud service scheduling for the sustainable development of cloud computing. We first focus our research on the development of scheduling methods for delay-sensitive cloud services on a single server with the goal of maximizing a service provider's profit. We then extend our study to scheduling cloud services in distributed environments. In particular, we develop a queue-based model and derive efficient request dispatching and processing decisions in a multi-electricity-market environment to improve the profits for service providers. We next study a problem of multi-tier service scheduling. By carefully assigning sub deadlines to the service tiers, our approach can significantly improve resource usage efficiencies with statistically guaranteed QoS. Finally, we study the power conscious resource provision problem for service requests with different QoS requirements. By properly sharing computing resources among different requests, our method statistically guarantees all QoS requirements with a minimized number of powered-on servers and thus the power consumptions. The significance of our research is that it is one part of the integrated effort from both industry and academia to ensure the sustainable growth of cloud computing as it continues to evolve and change our society profoundly.

Essays on durable goods consumption and firm innovation

This dissertation comprises three individual chapters. Chapter Two examines how free riding across neighbors influenced the diffusion of color television sets in rural China. Chapter Three tests for asymmetric information between a firm’s management and other investors concerning its patent output. Chapter Four discusses how knowledge stocks influence a patenting firm’s later diversification. Chapter Two documents the existence of a type of network effects—free riding across neighbors—in the consumption of color television sets in rural China, which reduces the propensity of non-owners to purchase. I construct a model of the timing of the purchase of a durable good in the presence of free riding, and test its key implications using household survey data in rural China. Chapter Three tests for asymmetric information between a firm’s management and other investors about its patent output by examining insider trading patterns and stock price changes in R&D intensive firms. It demonstrates that management has considerable information about its patent output beyond what is known to investors. It also shows that the predictive power of insider trading patterns on patent output comes from purchases rather than sales. Chapter Four discusses two sequential channels through which knowledge stocks may influence a firm’s later diversification. One is that firms with more knowledge are more likely to enter a new industry. The other is that firms’ businesses have a better chance of surviving, conditional on being formed. By examining U.S. public patenting firms in manufacturing sectors for 1984-1996, I find that knowledge stocks predict the likelihood of new industry entry when controlling for firm size. However, this predictive power is weakened when diversification effects are included. On the other hand, a survival study of newly established segments shows that initial knowledge stocks have significant positive effects on segment survival, whereas diversification effects are insignificant.