Large Language Models (LLMs) are increasingly integrated into Physical-I/O limited agents, such as robots and voice assistants, which execute outputs sequentially. However, existing LLM serving systems typically employ a throughput-oriented batching mechanism, ignoring the large gap between LLM generation speed and the constrained physical I/O rates of agents, thus wasting execution slack and worsening resource contention. Besides, they treat all tokens equally and cannot anticipate the execution implications of different content, preventing scheduling aligned with agent-side behavior. To address it, we propose a new system named TimelyLLM that coordinates LLM generation with the physical behavior of agents. TimelyLLM introduces a novel segmented generation and scheduling mechanism, strategically leveraging the time gap between agent plan generation and execution to reduce contention and improve response latency under multi-agent workloads. We implement TimelyLLM on top of a widely-used LLM serving framework. We also build a dataset collection system to construct serving workloads from real-world robots, including drones, robot arms, and quadruped robots. Our evaluation demonstrates that TimelyLLM improves the time utility up to 1.52×, and reduces the overall waiting time by 84%.

Minimum-Weight Perfect Matching (MWPM) decoding is important to quantum error correction decoding because of its accuracy. However, many believe that it is difficult, if possible at all, to achieve the microsecond latency requirement posed by superconducting qubits. This work presents the first publicly known MWPM decoder, called Micro Blossom, that achieves sub-microsecond decoding latency. Micro Blossom employs a heterogeneous architecture that carefully partitions a state-of-the-art MWPM decoder between software and a programmable accelerator with parallel processing units, one of each vertex/edge of the decoding graph. On a surface code with code distance d and a circuit-level noise model with physical error rate p, Micro Blossom's accelerator employs O(d3) parallel processing units to reduce the worst-case latency from O(d12) to O(d9) and reduce the average latency from O(pd3+1) to O(p2 d2+1) when p «1.

We present Cacheback Decoding, a trainingfree and model-agnostic speculative decoding method that exploits the locality in language to accelerate Large Language Model (LLM) inference.Cacheback leverages only Least Recently Used (LRU) cache tables of token n-grams to generate draft sequences.Cacheback achieves state-of-the-art performance among comparable methods despite its minimalist design, and its simplicity allows easy integration into existing systems.Cacheback also shows potential for fast adaptation to new domains.

Current large language model (LLM) serving systems, primarily designed for text completion, are neither efficient nor adaptable for increasingly complex LLM applications due to their inflexible design. We propose a new LLM serving system architecture that serves programs instead of prompts to address this problem. These programs, called LLM Inference Programs (LIPs), allow users to customize token prediction and KV cache management at runtime and to offload parts of their application logic, such as tool execution, to the server. We describe an example of this architecture through a system named Symphony, which functions as an operating system for LIPs. Symphony exposes LLM model computations via system calls and virtualizes KV cache with a dedicated file system, while ensuring GPU efficiency with a two-level process scheduling scheme. Symphony has the potential to open the door to a more efficient and extensible ecosystem for LLM applications.

Microcontroller-based embedded systems are vulnerable to memory safety errors and must be robust and responsive because they are often used in unmanned and mission-critical scenarios. The Rust programming language offers an appealing compile-time solution for memory safety but leaves stack overflows unresolved and foils zero-latency interrupt handling. We present Hopter, a Rust-based embedded operating system (OS) that provides memory safety, system robustness, and interrupt responsiveness to embedded systems while requiring minimal application cooperation. Hopter executes Rust code under a novel finite-stack semantics that converts stack overflows into Rust panics, enabling recovery from fatal errors through stack unwinding and restart. Hopter also employs a novel mechanism called soft-locks so that the OS never disables interrupts. We compare Hopter with other well-known embedded OSes using controlled workloads and report our experience using Hopter to develop a flight control system for a miniature drone and a gateway system for Internet of Things (IoT). We demonstrate that Hopter is well-suited for resource-constrained microcontrollers and supports error recovery for real-time workloads.

Emerging large language model (LLM) applications involve diverse reasoning strategies and agentic workflows, straining the capabilities of existing serving systems built on a monolithic token generation loop. This paper introduces Pie, a programmable LLM serving system designed for flexibility and efficiency. Pie decomposes the traditional generation loop into fine-grained service handlers exposed via an API and delegates control of the generation process to user-provided programs, called inferlets. This enables applications to implement new KV cache strategies, bespoke generation logic, and seamlessly integrate computation and I/O—entirely within the application, without requiring modifications to the serving system. Pie executes inferlets using WebAssembly, benefiting from its lightweight sandboxing. Our evaluation shows Pie matches state-of-the-art performance on standard tasks (3-12% latency overhead) while significantly improving latency and throughput (1.3×-3.4× higher) on agentic workflows by enabling application-specific optimizations.

Recent advancements in robot planning using large language models (LLMs) have demonstrated significant potential, primarily due to LLMs' capabilities to understand natural language commands and generate executable plans in various languages. However, in time-sensitive and interactive applications involving mobile robots, particularly drones, the sequential token generation process inherent to LLMs introduces substantial latency, i.e., response time, during the control plan generation. In this paper, we present a system called ChatFly that tackles this latency problem using a combination of a novel programming language called MiniSpec and its runtime to reduce both the response time and generation time for the robot plan. That is, instead of asking an LLM to write a program (robotic plan) in the popular but verbose Python, ChatFly gets it to do it in MiniSpec specially designed for token efficiency and stream interpreting. Using a set of challenging drone tasks, we show that design choices made by ChatFly can reduce the average response time to 74% compared to existing works and provide a more consistent user experience, enabling responsive and intelligent LLM-based drone control.

A fault-tolerant quantum computer must decode and correct errors faster than they appear. The faster errors can be corrected, the more time the computer can do useful work. The Union-Find (UF) decoder is promising with an average time complexity slightly higher than$O(d^{3})$. We report a distributed version of the UF decoder that exploits parallel computing resources for further speedup. Using an FPGA-based implementation, we empirically show that this distributed UF decoder has a sublinear average time complexity with regard to$d$, given$O(d^{3})$parallel computing resources. The decoding time per measurement round decreases as$d$increases, a first time for a quantum error decoder. The implementation employs a scalable architecture called Helios that organizes parallel computing resources into a hybrid tree-grid structure. Using Xilinx's cycle-accurate simulator, we present cycle-accurate decoding time for$d$up to 15, with the phenomenological noise model with$p=0.1\%$. We are able to implement$d$up to 7 with a Xilinx ZC106 FPGA, for which an average decoding time is 120 ns per measurement round. Since the decoding time per measurement round of Helios decreases with$d$, Helios can decode a surface code of arbitrarily large$d$without a growing backlog.

Stack unwinding is a well-established approach for handling panics in Rust programs. However, its feasibility on resource-constrained embedded systems has been unclear due to the associated overhead and complexity. This paper presents our experience of implementing stack unwinding and panic recovery within a Rust-based soft real-time embedded operating system. We describe several novel optimizations that help achieve adequate performance for a flying drone with a CPU overhead of 2.6% and a storage overhead of 26.0% to recover from panics in application tasks and interrupt handlers.

Memory disaggregation promises transparent elasticity, high resource utilization and hardware heterogeneity in data centers by physically separating memory and compute into network-attached resource "blades". However, existing designs achieve performance at the cost of resource elasticity, restricting memory sharing to a single compute blade to avoid costly memory coherence traffic over the network.

Massive multiple-input multiple-output (MIMO) is a key technology in 5G New Radio (NR) to improve spectral efficiency. A major challenge in its realization is the huge amount of real-time computation required. All existing massive MIMO baseband processing solutions use dedicated and specialized hardware like FPGAs, which can efficiently process baseband data but are expensive, inflexible and difficult to program. In this paper, we show that a software-only system called Agora can handle the high computational demand of real-time massive MIMO baseband processing on a single many-core server. To achieve this goal, we identify the rich dimensions of parallelism in massive MIMO baseband processing, and exploit them across multiple CPU cores. We optimize Agora to best use CPU hardware and software features, including SIMD extensions to accelerate computation, cache optimizations to accelerate data movement, and kernel-bypass packet I/O. We evaluate Agora with up to 64 antennas and show that it meets the data rate and latency requirements of 5G NR.

A key challenge for frequency-division duplexing (FDD) massive multi-input multi-output (MIMO) is the large overhead in acquiring channel state information (CSI) for transmits beamforming. In this paper, we propose a scalable method called directional training to obtain downlink CSI. Directional training is motivated by two empirical results derived from massive MIMO channel measurements. First, the number of dominant angle-of-arrivals (departures) is much smaller than and nearly independent of the number of base-station antennas. Second, there is a strong correlation between uplink arrival and downlink departure angles even in FDD systems, which leads to the idea of directional training, where a small number of training symbols can be sent to estimate the dominant components of the downlink channel. Therefore, directional training measures much fewer complex coefficients than full-training-based methods, and as a result, compared with full-training, the overall channel acquisition overhead for directional training scales much slower with the number of base-station antennas. We evaluate directional training with extensive experiments with a 64-antenna base-station at two bands separated by approximately 72 MHz. Our results show that directional training-based downlink beamforming outperforms full-training systems by 150% in terms of average spectral efficiency, and loses only 5.3% average spectral efficiency from genie-aided systems.

Understanding and managing the propagation of states in operating systems has become an intractable problem due to their sheer size and complexity. Despite modularization efforts, it remains a significant barrier to many contemporary computing goals: process migration, fault isolation and tolerance, live update, software virtualization, and more. Though many previous OS research endeavors have achieved these goals through ad-hoc, tedious methods, we argue that they have missed the underlying reason why these goals are so challenging: state spill.

We present the third generation of Argos platforms, ArgosV3, intended for real-world applications and research. Developed from scratch specifically for many-antenna MU-MIMO, ArgosV3 is highly efficient in space, power, computation, and cost. While this new platform is highly configurable, featuring FPGA SoCs and frequency agile transceivers capable of operation from 50 MHz to 3.8 GHz, it is also highly compact and power efficient, enabling a complete 160 radio base station in less than 2 cubic feet. ArgosV3 is currently being deployed in a campus-wide multi-cell many-antenna network. For our demonstration we will show a single ArgosV3 base station serving multiple clients using a realtime LTE stack.

Like software, hardware also has bugs. These bugs, or errata, can cause unexpected behaviors, reducing performance and causing malfunctions in the entire system. As there is no way to fix an erratum after hardware is deployed, its harmful effects are often mitigated by software workarounds, usually in low-level software such as operating system. The goal of our work is to ensure system correctness and security against the harmful effects of errata and their workarounds. The first step is to systematically understand them in the wild.

In prior work, we have shown that the underdiagnosed problem of state spill remains a barrier to realizing complex systems that are easy to maintain, evolve, and run reliably. This paper shares our early experience building Theseus from scratch, an OS with the guiding principle of eliminating state spill. Theseus takes inspiration from distributed systems to rethink state management, and leverages Rust language features for maximum safety, code reuse, and efficient isolation. We intend to demonstrate Theseus as a runtime composable OS, in which entities are easily interchangeable and can evolve independently without reconfiguring or rebooting.

Drawing or dragging an object on a mobile device is annoying today because the latency is manifested spatially with an obvious gap between the touch point and the line head or dragged object. This work identifies the multiple synchronization points in the input to display path of modern mobile systems as a major source of latency, contributing about 30 ms to the overall latency.

Continuous mobile vision is limited by the inability to efficiently capture image frames and process vision features. This is largely due to the energy burden of analog readout circuitry, data traffic, and intensive computation. To promote efficiency, we shift early vision processing into the analog domain. This results in RedEye, an analog convolutional image sensor that performs layers of a convolutional neural network in the analog domain before quantization. We design RedEye to mitigate analog design complexity, using a modular column-parallel design to promote physical design reuse and algorithmic cyclic reuse. RedEye uses programmable mechanisms to admit noise for tunable energy reduction. Compared to conventional systems, RedEye reports an 85% reduction in sensor energy, 73% reduction in cloudlet-based system energy, and a 45% reduction in computation-based system energy.

In this paper, we present and study a digital-controlled method, called SoftNull, to enable full-duplex in many-antenna systems. Unlike most designs that rely on analog cancelers to suppress self-interference, SoftNull relies on digital transmit beamforming to reduce self-interference. SoftNull does not attempt to perfectly null self-interference, but instead seeks to reduce self-interference sufficiently to prevent swamping the receiver's dynamic range. Residual self-interference is then cancelled digitally by the receiver. We evaluate the performance of SoftNull using measurements from a 72-element antenna array in both indoor and outdoor environments. We find that SoftNull can significantly outperform half-duplex for small cells operating in the many-antenna regime, where the number of antennas is many more than the number of users served simultaneously.

Non-CPU devices on a modern system-on-a-chip (SoC), ranging from accelerators to I/O controllers, account for a significant portion of the chip area. It is therefore vital for system energy efficiency that idle devices can enter a low-power state while still meeting the performance expectation. This is called device runtime Power Management (PM) for which individual device drivers in commodity OSes are held responsible today. Based on the observations of existing drivers and their evolution, we consider it harmful to rely on drivers for device runtime PM. This paper identifies three pieces of information as essential to device runtime PM, and shows that they can be obtained without involving drivers, either by using a software-only approach, or more efficiently, by adding one register bit to each device. We thus suggest a structural change to the current Linux runtime PM framework, replacing the PM code in all applicable drivers with a single kernel module called the central PM agent. Experimental evaluations show that the central PM agent is just as effective as hand-tuned driver PM code. The paper also presents a tool called PowerAdvisor that simplifies driver PM efforts under the current Linux runtime PM framework. PowerAdvisor analyzes execution traces and suggests where to insert PM calls in driver source code. Despite being a best-effort tool, PowerAdvisor not only reproduces hand-tuned PM code from stock drivers, but also correctly suggests PM code never known before. Overall, our experience shows that it is promising to ultimately free driver developers from manual PM.

Many-antenna MU-MIMO faces a critical, previously unaddressed challenge: it lacks a practical control channel. At the heart of this challenge is that the potential range of MU-MIMO beamforming systems scales with up to the square of the number of base-station antennas once they have channel state information (CSI), whereas the range of traditional control channel operations remains constant since they take place before or during CSI acquisition. This range gap between no-CSI and CSI modes presents a critical challenge to the efficiency and feasibility of many-antenna base stations, as their operational range is limited to the no-CSI mode. We present a novel control channel design for many-antenna MU-MIMO, Faros, that allows the number of base-station antennas to scale up to 100s in practice. Faros leverages a combination of open-loop beamforming and coding gains to bridge the range gap between the CSI and no-CSI modes. Not only does Faros provide an elegant and efficient control channel for many-antenna MU-MIMO, but on a more fundamental level it exposes flexible, fine-grained, control over space, time, and code resources, which enables previously impossible optimizations. We implement our design on the Argos many-antenna base station and evaluate its performance in bridging the range gap, synchronization, and paging. With 108 antennas, Faros can provide over 40 dB of gain, which enables it to function reliably at over 250 meters outdoors with less than 100 μW of transmit power per antenna, 10 mW total, at 2.4 GHz.

Emerging wearable devices promise a multitude of computer vision-based applications that serve users without active engagement. However, vision algorithms are known to be resource-hungry; and modern mobile systems do not support concurrent application use of the camera. Toward supporting efficient concurrency of vision applications, we report Starfish, a split-process execution system that supports concurrent vision applications by allowing them to share computation and memory objects in a secure and efficient manner. Starfish splits the vision library from an application into a separate process, called the Core, which centrally serves all vision applications. The Core shares library call results among applications, eliminating redundant computation and memory use. Starfish supports unmodified applications and unmodified libraries without needing their source code, and guarantees correctness to the applications. In doing so, Starfish improves both the performance and energy efficiency of concurrent vision applications. Using a prototype implementation on Google Glass, we experimentally demonstrate that Starfish reduces the time spent processing repeated vision library calls by 71% - 97%. When running two to ten concurrent face recognition applications at 0.3 frames per second, Starfish reduces CPU utilization by more than 42% - 80%. Notably, this keeps CPU utilization below 13%, even as the number of applications increases. This reduces system power consumption by 19% - 58%, as Starfish maintains a power consumption at approximately 1210 mW while running the concurrent application workloads.

Context information brings new opportunities for efficient and effective applications and services on mobile devices. A wide range of research has exploited context dependency, i.e. the relations between context(s) and the outcome, to achieve significant, quantified, performance gains for a variety of applications and services. These works typically have to deal with the challenges of multiple context sources leading to a sparse training dataset, and the challenges of energy hungry context sensors. Often, they address these challenges in an application specific and ad-hoc manner. We liberate mobile application designers and researchers from these burdens by providing a methodical approach to these challenges. In particular, we 1) define and measure the context-dependency of three principal types of mobile usage (visited websites, phone calls, and app usage) in an application agnostic yet practical manner, providing insight into the performance of potential application. 2) Address the challenge of data sparseness when dealing with multiple context sources in a systematic manner. 3) Present SmartContext to address the energy challenge by automatically selecting among context sources while ensuring a minimum accuracy for each estimation. Our analysis and findings are based on one year of usage and context traces collected in real-life settings from 24 iPhone users. We present findings regarding the context dependency of three types of mobile usage from 24 users, yet our methodology and the lessons we learn can be readily extended to other types of usage as well as system resources. Our findings guide the development of context aware systems, and highlight the challenges and expectations regarding the context dependency of mobile usage.

Mobile System-on-Chips (SoC) that incorporate heterogeneous coherence domains promise high energy efficiency to a wide range of mobile applications, yet are difficult to program. To exploit the architecture, a desirable, yet missing capability is to replicate operating system (OS) services over multiple coherence domains with minimum inter-domain communication. In designing such an OS, we set three goals: to ease application development, to simplify OS engineering, and to preserve the current OS performance. To this end, we identify a shared-most OS model for multiple coherence domains: creating per-domain instances of core OS services with no shared state, while enabling other extended OS services to share state across domains. To test the model, we build K2, a prototype OS on the TI OMAP4 SoC, by reusing most of the Linux 3.4 source. K2 presents a single system image to applications with its two kernels running on top of the two coherence domains of OMAP4. The two kernels have independent instances of core OS services, such as page allocation and interrupt management, as coordinated by K2; the two kernels share most extended OS services, such as device drivers, whose state is kept coherent transparently by K2. Despite platform constraints and unoptimized code, K2 improves energy efficiency for light OS workloads by 8x-10x, while incurring less than 9% performance overhead for two device drivers shared between kernels. Our experiences with K2 show that the shared-most model is promising.

Mobile System-on-Chips (SoC) that incorporate heterogeneous coherence domains promise high energy efficiency to a wide range of mobile applications, yet are difficult to program. To exploit the architecture, a desirable, yet missing capability is to replicate operating system (OS) services over multiple coherence domains with minimum inter-domain communication. In designing such an OS, we set three goals: to ease application development, to simplify OS engineering, and to preserve the current OS performance. To this end, we identify a shared-most OS model for multiple coherence domains: creating per-domain instances of core OS services with no shared state, while enabling other extended OS services to share state across domains. To test the model, we build K2, a prototype OS on the TI OMAP4 SoC, by reusing most of the Linux 3.4 source. K2 presents a single system image to applications with its two kernels running on top of the two coherence domains of OMAP4. The two kernels have independent instances of core OS services, such as page allocator and interrupt management, as coordinated by K2; the two kernels share most extended OS services, such as device drivers, whose state is kept coherent transparently by K2. Despite platform constraints and unoptimized code, K2 improves energy efficiency for light OS workloads by 8x-10x, while incurring less than 6% performance overhead for a device driver shared between kernels. Our experiences with K2 show that the shared-most model is promising.

Paravirtualization is an important I/O virtualization technology since it uniquely provides all of the following benefits: the ability to share the device between multiple VMs, support for legacy devices without virtualization hardware, and high performance. However, existing paravirtualization solutions have one main limitation: they only support one I/O device class, and would require significant engineering effort to support new device classes and features. In this paper, we present Paradice, a solution that vastly simplifies I/O paravirtualization by using a common paravirtualization boundary for various I/O device classes: Unix device files. Using this boundary, the paravirtual drivers simply act as a class-agnostic indirection layer between the application and the actual device driver.

Energy accounting determines how much a software principal contributes to the total system energy consumption. It is the foundation for evaluating software and for operating system based energy management. While various energy accounting policies have been tried, there is no known way to evaluate them directly simply because it is hard to track all hardware usage by software in a heterogeneous multicore system like modern smartphones and tablets.

While computer vision algorithms and libraries have enabled and accelerated the adoption of vision processing into mobile and wearable applications, vision is a resource-hungry operation, and is thus not efficient enough to run on multiple applications simultaneously. However, we observe that many vision algorithms share identical sets of frames and features to perform their analyses, computed from the same library calls. Leveraging this observation, we design a split-process architecture to retrofit existing vision libraries to allow applications to transparently share the computational, memory, and energy overhead of vision processing.

In an 802.11ac-based MU-MIMO network comprised of multiple cells1, inter-cell interference allows only a single AP to serve its clients at the same time, significantly limiting the network capacity. In this work, we overcome this limitation by letting the APs and clients in interfering cells coordinately cancel the inter-cell interference using their antennas for beamforming. To achieve such coordinated interference cancellation in a practical way, we propose a novel two-step optimization. First, without requiring any channel knowledge, each AP and client optimizes the use of its antennas for either data communication or inter-cell interference cancellation, in order to maximize the total number of deliverable streams in the MU-MIMO network. Second, with only partial channel knowledge, each AP and client optimizes their beamforming weights after the optimal antenna usage has been identified in the first step. Our solution, CoaCa, integrates this two-step optimization into 802.11ac with small modifications and negligible overhead, allowing each AP and client to locally perform the two-step optimization. Our experimental evaluation indicates that for a MU-MIMO network with two cells, by cancelling the inter-cell interference CoaCa can convert the majority of the expected number of streams increase (50%-67%) into network capacity improvement (41%-52%).

Mobile systems are equipped with a diverse collection of I/O devices, including cameras, microphones, sensors, and modems. There exist many novel use cases for allowing an application on one mobile system to utilize I/O devices from another. This paper presents Rio, an I/O sharing solution that supports unmodified applications and exposes all the functionality of an I/O device for sharing. Rio's design is common to many classes of I/O devices, thus significantly reducing the engineering effort to support new I/O devices. Our implementation of Rio on Android consists of about 7100 total lines of code and supports four I/O classes with fewer than 500 class-specific lines of code. Rio also supports I/O sharing between mobile systems of different form factors, including smartphones and tablets. We show that Rio achieves performance close to that of local I/O for audio devices, sensors, and modem, but suffers noticeable performance degradation for camera due to network throughput limitations between the two systems, which is likely to be alleviated by emerging wireless standards.

A user nowadays owns a variety of mobile systems, including smartphones, tablets, smart glasses, and smart watches, each equipped with a plethora of I/O devices, such as cameras, speakers, microphones, sensors, and cellular modems. There are many interesting use cases in which an application running on one mobile system accesses I/O on another system, for three fundamental reasons. (i) Mobile systems can be in different physical locations or orientations. For example, one can control a smartphone's high-resolution camera from a tablet camera application to more easily capture a self-portrait. (ii) Mobile systems can serve different users. For example, one can a play music for another user if one's smartphone can access the other device's speaker. (iii) Certain mobile systems have unique I/O devices due to their distinct form factor and targeted use cases. For example, a user can make a phone call from her tablet using the modem and SIM card in her smartphone.

Modern mobile systems are equipped with a diverse collection of I/O devices, including cameras, microphones, various sensors, and cellular modem. There exist many novel use cases for allowing an application on one mobile system to utilize I/O devices from another. This video demonstrates Rio, an I/O sharing solution that supports unmodified applications and realizes many of these novel use cases. Rio's design is common to many classes of I/O devices, significantly reducing the engineering effort to support new I/O devices. Moreover, it supports all the functionalities of an I/O device for sharing. Rio also supports I/O sharing between mobile systems of different form factors, including smartphones and tablets.

Is it possible to migrate TCP/IP flows between different networks on modern mobile devices without infrastructure support or protocol changes? To answer this question, we make three research contributions: 1) We report a comprehensive characterization of IP traffic on 27 iPhone 3GS users for three months. 2) Driven by these findings, we devise two simple, effective, and easily deployable system mechanisms to support seamless flow migration without network support, and extensively evaluate their effectiveness using our field collected traces of real-life usage. Wait-n-Migrate leverages the fact that most flows are short lived. It establishes new flows on newly available networks but allows preexisting flows on the old network to terminate naturally. Resumption Agent takes advantage of the resumption functionality of modern protocols to securely resume flows without application intervention. Combined, they provide an unprecedented opportunity to immediately deploy policies that leverage multiple networks to improve the performance, efficiency, and connectivity of mobile devices. 3) We report an iPhone-based implementation of these system mechanisms and demonstrate their overhead to be negligible. Furthermore, we employ a sample switching policy, AutoSwitch, to demonstrate their performance. Through traces and field measurements, we show that AutoSwitch reduces user disruptions by an order of magnitude.

Many-antenna base stations are a rapidly growing field in wireless research. A plethora of new theoretical techniques have been recently proposed for many-antenna base stations and networks. However, without experimental validation, it is difficult or impossible to predict the practicality and performance of these techniques in real hardware, under complex, rapidly varying, real-world conditions. Indeed, there is a significant demand for a flexible many-antenna research platform which supports rapid prototyping and validation of new massive-MIMO techniques. Leveraging our experience building Argos, a 64-antenna base station prototype, we have designed and built ArgosV2, a compact, powerful, and scalable many-antenna research platform based on WARP. In addition to the physical hardware and mechanical design, we are developing a software framework, ArgosLab, which will provide synchronization and channel estimation, greatly reducing the development effort for a wide range of massive-MIMO techniques. ArgosV2 is intended to provide ultimate scalability and programmability for experimental massive-MIMO research. The modular architecture and real-time capability of ArgosV2 can support up to 100s of base station antennas and 10s of users with streaming applications. For our demonstration, we will unveil a 96-antenna base station which supports real-time streaming applications to 32 users simultaneously.

We report a first-of-its-kind smartphone software system, MoodScope, which infers the mood of its user based on how the smartphone is used. Compared to smartphone sensors that measure acceleration, light, and other physical properties, MoodScope is a "sensor" that measures the mental state of the user and provides mood as an important input to context-aware computing. We run a formative statistical mood study with smartphone-logged data collected from 32 participants over two months. Through the study, we find that by analyzing communication history and application usage patterns, we can statistically infer a user's daily mood average with an initial accuracy of 66%, which gradu-ally improves to an accuracy of 93% after a two-month personal-ized training period. Motivated by these results, we build a service, MoodScope, which analyzes usage history to act as a sensor of the user's mood. We provide a MoodScope API for developers to use our system to create mood-enabled applications. We further create and deploy a mood-sharing social application.

We present MoodScope, a software system which infers the mood of its user based on how the smartphone is used. Similar to smartphone sensors that measure acceleration, light, and other physical properties, MoodScope is a "sensor" that measures the mental state of the user and provides mood as an important input to context-aware computing. We run a formative statistical study with smartphone-logged data collected from 32 participants over two months. Through the study, we find that by analyzing communication history and application usage patterns, we can statistically infer a user's daily mood average with an accuracy of 93% after a two-month training period. Motivated by these results, we build a service, MoodScope, which analyzes usage history to act as a sensor of the user's mood.

A major hurdle to frequently performing mobile computer vision tasks is the high power consumption of image sensing. In this work, we report the first publicly known experimental and analytical characterization of CMOS image sensors. We find that modern image sensors are not energy-proportional: energy per pixel is in fact inversely proportional to frame rate and resolution of image capture, and thus image sensor systems fail to provide an important principle of energy-aware system design: trading quality for energy efficiency. We reveal two energy-proportional mechanisms, supported by current image sensors but unused by mobile systems: (i) using an optimal clock frequency reduces the power up to 50% or 30% for low-quality single frame (photo) and sequential frame (video) capturing, respectively; (ii) by entering low-power standby mode between frames, an image sensor achieves almost constant energy per pixel for video capture at low frame rates, resulting in an additional 40% power reduction. We also propose architectural modifications to the image sensor that would further improve operational efficiency. Finally, we use computer vision benchmarks to show the performance and efficiency tradeoffs that can be achieved with existing image sensors. For image registration, a key primitive for image mosaicking and depth estimation, we can achieve a 96% success rate at 3 FPS and 0.1 MP resolution. At these quality metrics, an optimal clock frequency reduces image sensor power consumption by 36% and aggressive standby mode reduces power consumption by 95%.

A hurdle to frequently performing mobile computer vision tasks is the high energy cost of image sensing. In particular, modern image sensors are not energy proportional; for low resolution and low frame rate capture, the image sensor consumes almost the same amount of energy as it does at high resolutions and high frame rates. We reveal two system-level energy proportional mechanisms: (i) using an optimal pixel clock frequency; (ii) entering low power standby mode between frames. These techniques can be implemented by the image sensor driver with minimal hardware adjustment. Further improvements can be made by designing sensors with heterogeneous hardware architectures. With energy proportionality, computer vision frameworks can be optimized for power consumption, continuously requesting low resolution frames with low energy while only occasionally using high energy to request high resolution frames. This will in turn enable low power continuous mobile vision applications.

No abstract available.

Many emerging mobile applications and services are based on smartphones. We have performed a four-month field study of the adoption and usage of smartphone-based services by 14 novice teenage users. From the field study, we present the application usage and usage characteristics of our participants. We show that their usage is highly mobile, location-dependent, and serves multiple social purposes. Furthermore, we report qualitative lessons regarding the evaluation of smartphone-based services. In particular, we highlight the cases that an accurate evaluation would require a long-term and/or field study instead of a short or lab-based study, and the cases where studying a particular application independently is insufficient and a holistic study, i.e., involving the whole device, is necessary. We further present guidelines on effectively shortening the length of a study. These lessons are supported in part by five identified contributing factors to usage evolution.

It gives us great pleasure to introduce this special section on outstanding papers from the ACM MobiSys 2012 conference, for which we were honored to be the program committee cochairs. The conference was held in June 2012 and included 32 high-quality papers selected from 182 submissions, a record high for both paper submissions and paper acceptances. The conference's call-for-papers informed authors that a few outstanding papers would be considered for a fast-tracked special section in the IEEE Transactions on Mobile Computing (TMC). The selection process for the papers included in this special section was a byproduct of the selection process for the MobiSys 2012 Best Paper Award. Immediately after the MobiSys 2012 program committee meeting, we asked the committee members to nominate accepted papers for the Best Paper Award. Based on the responses we received, five papers with the most nominations were selected as candidates for the award. Their nominations were kept confidential until the opening of the conference. We then asked the program committee and external reviewer panel members to volunteer for the Best Paper Award committee to determine the final best paper and select three out of the five nominated papers for this special section. Based on the responses and conflicts of interest, four members of the program committee and one member of the external reviewer panel formed the Best Paper Award committee. At the conference, this committee selected the best paper and the three papers to be fast-tracked for this special section. The paper review process for the selected papers followed the standard procedure for fast-tracked TMC special sections. The authors were asked to submit a journal version with an explanation of how they addressed the reviews and feedback received at the conference. Note that each of these three papers received five or more conference reviews and had gone through a shepherding process to address reviewer comments for the final, camera-ready version. As required by the TMC fast-tracking procedure, we invited at least two reviewers for each submission. At least one reviewer was a MobiSys program committee member who reviewed the corresponding conference submission and another was not a member of program committee. In the end, the review process produced three very highquality papers that we are proud to present to the readers of TMC.

To accomplish frequent, simple tasks with high efficiency, it is necessary to leverage low-power, microcontroller-like processors that are increasingly available on mobile systems. However, existing solutions require developers to directly program the low-power processors and carefully manage inter-processor communication. We present Reflex, a suite of compiler and runtime techniques that significantly lower the barrier for developers to leverage such low-power processors. The heart of Reflex is a software Distributed Shared Memory (DSM) that enables shared memory objects with release consistency among code running on loosely coupled processors. In order to achieve high energy efficiency without sacrificing performance much, the Reflex DSM leverages (i) extreme architectural asymmetry between low-power processors and powerful central processors, (ii) aggressive compile-time optimization, and (iii) a minimalist runtime that supports efficient message passing and event-driven execution. We report a complete realization of Reflex that runs on a TI OMAP4430-based development platform as well as on a custom tri-processor mobile platform. Using smartphone sensing applications reported in recent literature, we show that Reflex supports a programming style very close to contemporary smartphone programming. Compared to message passing, the Reflex DSM greatly reduces efforts in programming heterogeneous smartphones, eliminating up to 38% of the source lines of application code. Compared to running the same applications on existing smartphones, Reflex reduces the average system power consumption by up to 81%.

The current paper establishes empirical patterns associated with mobile internet use on smartphones and explores user differences in these behaviors. We apply a naturalistic and longitudinal logs-based approach to collect real usage data from 24 iPhone users in the wild. These data are used to describe smartphone usage and analyze revisitation patterns of web browsers, native applications, and physical locations where phones are used. Among our findings are that web page revisitation through browsers occurred very infrequently (approximately 25% of URLs are revisited by each user), bookmarks were used sparingly, physical traversing patterns mirrored virtual (internet) traversing patterns and users systematically differed in their web use. We characterize these differences and suggest ways to support users with enhanced design of smartphone technologies and content.

The ability to account system resource usage by software is the key to the design and optimization of modern computer systems. For example, scheduling and memory management are two classic operating system (OS) functions based on the ability to account the CPU and memory usage by process. Energy has become an important system resource due to electricity and thermal concerns. This is particularly true for mobile systems that are battery-powered and require compact form factors. Knowing the energy contribution by a process, or per-process energy accounting, is the foundation for OS energy management and optimization [11, 9], incentive mechanisms for emerging applications in participatory sensing and cooperative communication, detecting rogue applications [8], and software optimization for energy [6].

The WiFi radio in smartphones consumes a significant portion of energy when active. To reduce the energy consumption, the Power Saving Mode was standardized in IEEE 802.11 and two major implementations, Static PSM and Dynamic PSM, have been widely used in mobile devices. Unfortunately, both PSMs have inherent drawbacks: Static PSM is energy efficient but imposes considerable extra delays on data transfers; Dynamic PSM incurs little extra delay but misses energy saving opportunities. In this paper, we first analyze a one-week trace from 10 users and show that more than 80% of all traffic are Web 2.0 flows, which are of very small sizes and short durations. Targeting these short but dominant flows, we propose a system called Percy, to achieve the best of both worlds (Static and Dynamic PSM), i.e., to maximize the energy saving while minimizing the delay of flow completion time. Percy works by deploying a web proxy at the AP and suitably configuring the PSM parameters, and is designed to work with unchanged clients running Dynamic PSM, and unchanged APs and Internet servers. We evaluate our system via trace-driven testbed experiments. Our results show that Percy saves 40-70% energy compared to Dynamic PSM configurations of Nokia, iPhone and Android, while imposing low extra delay that can hardly be perceived by users.

Previous studies have found that smartphone users differ by orders of magnitude. We explore this variability to understand how users install and use native applications in ecologically-valid environments. A quasi-experimental approach is applied to compare how users in different socio-economic status (SES) groups adopt new smartphone technology along with how applications are installed and used. We present a longitudinal study of 34 iPhone 3GS users. 24 of these participants were chosen from two carefully selected SES groups who were otherwise similar and balanced. Usage data collected through an in-device programmable logger, as well as several structured interviews, identify similarities, differences, and trends, and highlight systematic differences in smartphone usage. A group of 10 lower SES participants were later recruited and confirm the influence of SES diversity on device usage. Among our findings are that a large number of applications were uninstalled, lower SES groups spent more money on applications and installed more applications overall, and the lowest SES group perceived the usability of their iPhones poorly in comparison to the other groups. We further discuss the primary reasons behind this low score, and suggest design implications to better support users across SES brackets.

Multi-user multiple-input multiple-output theory predicts manyfold capacity gains by leveraging many antennas on wireless base stations to serve multiple clients simultaneously through multi-user beamforming (MUBF). However, realizing a base station with a large number antennas is non-trivial, and has yet to be achieved in the real-world. We present the design, realization, and evaluation of Argos, the first reported base station architecture that is capable of serving many terminals simultaneously through MUBF with a large number of antennas (M >> 10). Designed for extreme flexibility and scalability, Argos exploits hierarchical and modular design principles, properly partitions baseband processing, and holistically considers real-time requirements of MUBF. Argos employs a novel, completely distributed, beamforming technique, as well as an internal calibration procedure to enable implicit beamforming with channel estimation cost independent of the number of base station antennas. We report an Argos prototype with 64 antennas and capable of serving 15 clients simultaneously. We experimentally demonstrate that by scaling from 1 to 64 antennas the prototype can achieve up to 6.7 fold capacity gains while using a mere 1/64th of the transmission power.

Mobile browser is known to be slow because of the bottleneck in resource loading. Client-only solutions to improve resource loading are attractive because they are immediately deployable, scalable, and secure. We present the first publicly known treatment of client-only solutions to understand how much they can improve mobile browser speed without infrastructure support. Leveraging an unprecedented set of web usage data collected from 24 iPhone users continuously over one year, we examine the three fundamental, orthogonal approaches a client-only solution can take: caching, prefetching, and speculative loading. Speculative loading, as is firstly proposed and studied in this work, predicts and speculatively loads the subresources needed to open a webpage once its URL is given. We show that while caching and prefetching are highly limited for mobile browsing, speculative loading can be significantly more effective. Empirically, we show that client-only solutions can improve the browser speed by about 1.4 second on average for websites visited by the 24 iPhone users. We also report the design, realization, and evaluation of speculative loading in a WebKit-based browser called Tempo. On average, Tempo can reduce browser delay by 1 second (~20%).

The present report is an empirical analysis of smartphone personalisation. We collected data from two groups of users to measure how they adapt the content, interface and physical appearance of their devices. This user-driven personalisation is measured with a simple heuristic approach to quantify the behaviour. Using these scores, we explore how users differ from each other in how they personalise their smartphones with a focus on gender differences, usability and device usage in the wild. Among our findings are that not all users personalise their smartphones, females and males personalise their iPhones differently, and those who personalised their phones more tended to rate it as more usable. The users who personalised more also used their device for greater periods of time on a broader range of applications. For instance, individuals who adapted their iPhones to a greater degree also accessed the Web more often and spent more time browsing once it was accessed. We conclude with a discussion of possible factors underlying the large user diversity of smartphone personalisation found in this research.

Displays based on organic light-emitting diode (OLED) technology are appearing on many mobile devices. Unlike liquid crystal displays (LCD), OLED displays consume dramatically different power for showing different colors. In particular, OLED displays are inefficient for showing bright colors. This has made them undesirable for mobile devices because much of the web content is of bright colors. To tackle this problem, we present the motivational studies, design, and realization of Chameleon, a color adaptive web browser that renders webpages with power-optimized color schemes under user-supplied constraints. Driven by the findings from our motivational studies, Chameleon provides end users with important options, offloads tasks that are not absolutely needed in real time, and accomplishes real-time tasks by carefully enhancing the codebase of a browser engine. According to measurements with OLED smartphones, Chameleon is able to reduce average system power consumption for web browsing by 41 percent and is able to reduce display power consumption by 64 percent without introducing any noticeable delay.

Emerging organic light-emitting diode (OLED)-based displays obviate external lighting, and consume drastically different power when displaying different colors, due to their emissive nature. This creates a pressing need for OLED display power models for system energy management, optimization as well as energy-efficient GUI design, given the display content or even the graphical-user interface (GUI) code. In this work, we study this opportunity using commercial QVGA OLED displays and user studies. We first present a comprehensive treatment of power modeling of OLED displays, providing models that estimate power consumption based on pixel, image, and code, respectively. These models feature various tradeoffs between computation efficiency and accuracy so that they can be employed in different layers of a mobile system. We validate the proposed models using a commercial QVGA OLED module and a mobile device with a QVGA OLED display. Then, based on the models, we propose techniques that adapt GUIs based on existing mechanisms as well as arbitrarily under usability constraints. Our measurement and user studies show that more than 75 percent display power reduction can be achieved with user acceptance.

High-speed wireless network interfaces are among the most power-hungry components on mobile systems. This is particularly true for multiple-input-multiple-output (MIMO) network interfaces which use multiple RF chains simultaneously. In this paper, we present a novel power management solution for MIMO network interfaces on mobile systems, called antenna management. The key idea is to adaptively disable a subset of antennas and their RF chains to reduce circuit power consumption, when the capacity improvement of using a large number of antennas is small. Antenna management judiciously determines the number of active antennas to minimize energy per bit while satisfying the data rate requirement. This work provides both theoretical framework and system design of antenna management. We first present an algorithm that efficiently solves the problem of minimizing energy per bit and, then offer its 802.11n-compliant system designs. We employ both Matlab-based simulation and prototype-based experiment to validate the energy efficiency benefit of antenna management. The results show that antenna management can achieve 21% one-end energy per bit reduction to the front end of the MIMO network interface, compared to a static MIMO configuration that keeps all antennas active.

In this work, we offer a novel system, MicroStation (μStation) that allows ubiquitous data broadcasting applications using the FM radio on mobile devices such as smartphones. μStation includes two key modules to enable data broadcasting based on existing mobile FM radio hardware. Channel Selector assigns different FM channels to neighboring μStation broadcasters to avoid collision and guides μStation listeners to find their broadcasting of interest. Data Codec realizes bit-level communication between mobile devices through existing FM radio hardware. We describe an implementation of μStation on the Nokia N900 smartphone, and provide low-level APIs and services to support application development. We also demonstrate two representative applications: Facebook-FM and Sync-Flash. These applications demonstrate the capability of μStation to readily enable a new class of ubiquitous data broadcasting applications on mobile devices.

In this work, we report the first study of an important realization of directional communication, beamforming, on mobile devices. We first demonstrate that beamforming is already feasible on mobile devices in terms of form factor, device mobility and power efficiency. Surprisingly, we show that by making an increasingly profitable tradeoff between transmit and circuit power, beamforming with state-of-the-art integrated CMOS implementations can be more power-efficient than its single antenna counterpart. We then investigate the optimal way of using beamforming in terms of device power efficiency, by allowing a dynamic number of active antennas. We propose a simple yet effective solution, BeamAdapt, which allows each mobile client in a network to individually identify the optimal number of active antennas with guaranteed convergence and close-to-optimal performance. We finally report a WARP-based prototype of BeamAdapt and experimentally demonstrate its effectiveness in realistic environments, and then complement the prototype-based experiments with Qualnet-based simulation of a large-scale network. Our results show that BeamAdapt with four antennas can reduce the power consumption of mobile clients by more than half compared to a single antenna, while maintaining a required network throughput.

Displays based on organic light-emitting diode (OLED) technology are appearing on many mobile devices. Unlike liquid crystal displays (LCD), OLED displays consume dramatically different power for showing different colors. In particular, OLED displays are inefficient for showing bright colors. This has made them undesirable for mobile devices because much of the web content is of bright colors.

System energy models are important for energy optimization and management in mobile systems. However, existing system energy models are built in a lab setting with the help from a second computer. Not only are they labor-intensive; but also they do not adequately account for the great diversity in the hardware and usage of mobile systems. Moreover, existing system energy models are intended for energy estimation for time intervals of one second or longer; they do not provide the required rate for fine-grain use such as per-application energy accounting.

No abstract available.

No abstract available.

In this paper, we present a comprehensive study on sensor-assisted video encoding (SaVE) schemes for video capturing on mobile devices in real-world environments. Our purpose is to reduce the computational complexity of video encoding by leveraging sensors that are increasingly available on mobile devices, e.g., accelerometers and digital compasses. Motion estimation is a key component of video encoding. In this paper, SaVE calculates the rotational movement of a camera (on mobile devices) and then infers the global motion in the camera imager. SaVE subsequently employs the estimated global motion as predictors to simplify motion estimation algorithms for state-of-the-art H.264/AVC video coding. We have constructed a prototype of SaVE and evaluated its performance with a pair of accelerometers, a digital compass, and their combination. Our experimental results show that SaVE can significantly reduce the computations of motion estimation while achieving equal or better video quality. Our results also show that SaVE has a strong noise-resistant capability. Therefore, it can be practically employed in real-world environments.

Context information brings new opportunities for efficient and effective system resource management of mobile devices. In this work, we focus on the use of context information to achieve energy-efficient, ubiquitous wireless connectivity. Our field-collected data show that the energy cost of network interfaces poses a great challenge to ubiquitous connectivity, despite decent availability of cellular networks. We propose to leverage the complementary strengths of Wi-Fi and cellular interfaces by automatically selecting the most efficient one based on context information. We formulate the selection of wireless interfaces as a statistical decision problem. The challenge is to accurately estimate Wi-Fi network conditions without powering up the network interface. We explore the use of different context information, including time, history, cellular network conditions, and device motion, to statistically estimate Wi-Fi network conditions with negligible overhead. We evaluate several context-based algorithms for the estimation and prediction of current and future network conditions. Simulations using field-collected traces show that our network estimation algorithms can improve the average battery lifetime of a commercial mobile phone for an ECG reporting application by 40 percent, very close to the estimated theoretical upper bound of 42 percent. Furthermore, our most effective algorithm can predict Wi-Fi availability for one and ten hours into the future with 95 and 90 percent accuracy, respectively.

Wireless access is known to be power-hungry for mobile devices. A key reason is that devices radiate power in all directions and much of this power will not reach the destination. To address this waste, we present BeamSwitch, a multi-antenna system designed to realize directional communication efficiently. Unlike power-hungry and expensive beamforming, BeamSwitch requires only one transceiver. We provide an 802.11-compliant design and prototype of BeamSwitch. Our measurements show that with three passive directional antennas, BeamSwitch reduces the power consumption of a commercial 802.11 adapter by up to 20% and provide better quality under diverse propagation environments and extreme rotation.

We report a first-of-its-kind realization of directional transmission for smartphone-like mobile devices using multiple passive directional antennas, supported by only one RF chain. The key is a multi-antenna system (MiDAS) and its antenna selection methods that judiciously select the right antenna for transmission. It is grounded by two measurement-driven studies regarding 1) how smartphones rotate during wireless usage in the field and 2) how orientation and rotation impact the performance of directional antennas under various propagation environments.

Mobile phones are truly personal devices loaded with personal data such as photos, contacts, and call history. Yet it is often necessary or desirable to share our phones with others. This is especially true as mobile phones are integrating features conventionally provided by other dedicated devices, from MP3 players to games consoles. Yet existing phones assume a single user and provide little protection for private data and applications when a phone is shared. That is, when we lend our phones to others, we give away complete access. In this work, we present xShare, a protection solution to address this problem. xShare allows phone owners to rapidly specify what they want to share and place the phone into a restricted mode where only the data and applications intended for sharing can be accessed. We first present two formative user studies and derive the design requirements of xShare. We then offer the design of xShare based on file-level access control. We describe the implementation of xShare on Windows Mobile and report a comprehensive evaluation, including performance measurements, usability, and a one-month field trial.

Emerging organic light-emitting diode (OLED)-based displays obviate external lighting; and consume drastically different power when displaying different colors, due to their emissive nature. This creates a pressing need for OLED display power models for system energy management, optimization as well as energy-efficient GUI design, given the display content or even the graphical user interface (GUI) code. In this work, we present a comprehensive treatment of power modeling of OLED displays, providing models that estimate power consumption based on pixel, image, and code, respectively. These models feature various tradeoffs between computation efficiency and accuracy so that they can be employed in different layers of a mobile system. We validate the proposed models using a commercial QVGA OLED module. For example, our statistical learning-based image-level model reduces computation by 1600 times while keeping the error below 10%, compared to the more accurate pixel-level model.

Energy efficiency has become increasingly important to mobile systems on which wireless interfaces are among the largest power consumers. While existing physical layer power optimization mostly focuses on improving the transmission efficiency, our recent work has showed that wireless interfaces can spend most of its time and energy in very short idle periods between transmitting two packets [9]. In this work, we present a physical layer optimization method, drowsy transmission, which explicitly considers the power cost of such idle periods in physical layer power optimization through joint power control/rate selection and power management. We provide a control theoretical formulation of the optimization problem and present a dynamic programming based solution and its approximation that is close form and practical. We further offer an on-line learning technique to cope with unknown channel and traffic. Using a power model from a commercial wireless network interface card, we demonstrate that drowsy transmission can reduce the energy per bit by 70% and 40% in comparison to power control/rate selection-based optimization and optimization with disjoint power control/rate selection and power management, respectively. Moreover, the achieved energy per bit is very close to the theoretical lower bound. Our evaluation shows that the proposed on-line learning technique can assess the channel and approach the performance under pre-known channel in as short as 200 ms. We also show that our optimization introduces negligible packet delays.

Emerging organic light-emitting diode (OLED)-based displays have drastically different power consumption when displaying different colors, due to their emissive nature. They bring a new opportunity for power saving by transforming GUI colors. In this work, we study this opportunity using a commercial-off-the-shelf QVGA OLED module and user studies. We present techniques that adapt GUIs based on existing mechanisms as well as arbitrarily under usability constraints. Our measurement and user studies show that more than 75% display power reduction can be achieved with user acceptance.

We present the theoretical foundation, implementation, and experimental evaluation of a novel power-saving mechanism for wireless transmission from multiple-input multiple-output (MIMO) transceivers, called RF chain management. RF chain management seeks to minimize the energy per bit for MIMO transmission, via adaptively choosing the optimal RF chain configuration, and satisfies the minimum data rate requirement at the same time. Our simulation shows that up to 45% and averagely 23% energy per bit reduction can be achieved. We have also built a prototype based on the WARP platform, and our experimental results have proved the feasibility of RF chain management in real systems and under realistic channels.

We report a series of user studies that evaluate the feasibility and usability of light-weight user authentication with a single tri-axis accelerometer. We base our investigation on uWave, a state-of-the-art recognition system for user-created free-space manipulation, or gestures. Our user studies address two types of user authentication: non-critical authentication (or identification) for a user to retrieve privacy-insensitive data; and critical authentication for protecting privacy-sensitive data. For non-critical authentication, our evaluation shows that uWave achieves high recognition accuracy (98%) and its usability is comparable with text ID-based authentication. Our results also highlight the importance of constraints for users to select their gestures. For critical authentication, the evaluation shows uWave achieves state-of-the-art resilience to attacks with 3% false positives and 3% false negatives, or 3% equal error rate. We also show that the equal error rate increases to 10% if the attackers see the users performing their gestures. This shows the limitation of gesture-based authentication and highlights the need for visual concealment.

Motion estimation is a key component of modern video encoding and is very compute-intensive. We present a novel Sensor-assisted Video Encoding (SaVE) method to reduce the computational complexity of motion estimation in H.264/AVC encoders, leveraging accelerometers and digital compasses that are increasingly available on mobile devices. Using these sensors, SaVE calculates the rotational movement of a camera and then infers the global motion in the camera image sensor; it subsequently employs the estimated global motion to simplify the state-of-the-art motion estimation algorithms, UMHS and EPZS used in H.264/AVC encoders. We have constructed a prototype of SaVE and report extensive evaluation of it. Our experimental results show that SaVE can reduce the computations of UMHS and EPZS algorithms by up to 27% and 18%, respectively, while achieving the same or better video quality.

Loaded with personal data, e.g. photos, contacts, and call history, mobile phones are truly personal devices. Yet it is often necessary or desirable to share our phones with others. This is especially true as mobile phones are integrating features conventionally provided by other dedicated devices, from MP3 players to games consoles. Unfortunately, when we lend our phones to others, we give away complete access because existing phones assume a single user and provide little protection for private data and applications. In this work, we present xShare, a protection solution to address this problem. xShare allows phone owners to rapidly specify what they want to share and place the phone into a restricted mode where only the data and applications intended for sharing can be accessed.

The proliferation of accelerometers on consumer electronics has brought an opportunity for interaction based on gestures or physical manipulation of the devices. We present uWave, an efficient recognition algorithm for such interaction using a single three-axis accelerometer. Unlike statistical methods, uWave requires a single training sample for each gesture pattern and allows users to employ personalized gestures and physical manipulations. We evaluate uWave using a large gesture library with over 4000 samples collected from eight users over an elongated period of time for a gesture vocabulary with eight gesture patterns identified by a Nokia research. It shows that uWave achieves 98.6% accuracy, competitive with statistical methods that require significantly more training samples. Our evaluation data set is the largest and most extensive in published studies, to the best of our knowledge. We also present applications of uWave in gesture-based user authentication and interaction with three-dimensional mobile user interfaces using user created gestures.

Consumer electronics and mobile devices intended for pervasive applications are often subject to shaking that makes their screen difficult to read. To address this usability challenge, we present NoShake, a system for screen content stabilization. NoShake utilizes a single accelerometer, now present in numerous consumer electronics and mobile devices. The core of NoShake is a physics inspired model that dynamically compensates for the device shaking by shifting the screen content opposite the direction of the shake. The model is efficient, parametric, and can be fine tuned based on shaking pattern detection. We implement a prototype of NoShake on an Apple iPhone and conduct user studies in a number of scenarios, which highlight the strengths as well as limitations of NoShake in coping with shaking devices.

Nanowire crossbar is one of the most promising circuit solutions for nanoelectronics. However, it is still unclear whether or how they can be competitive in implementing logic circuits, as compared to their MOSFET counterparts. We analyze nanowire crossbars in area, speed, and power, in comparison with their MOSFET counterparts. We show nanowire crossbars do not scale well in terms of logic density and speed. To achieve performance close to their MOSFET counterparts, crossbar circuits need faster field-effect transistors (FETs) to compensate the high resistance of nanowires. Motivated by the analysis and comparative study, we propose a crossbar cells design based on judicious use of silicon nanowires. The crossbar cell is compatible with the conventional MOSFET fabrication and design methodologies, in particular, standard cell-based integrated circuit design. We evaluate logic circuits synthesized with crossbar cells and MOSFET cells based on the MCNC91 benchmark. The results show that crossbar cells can provide a density advantage of more than four times over the traditional MOSFET circuits with the same process technology, while achieving close performance and consuming less than one third power.

Nanowire crossbar is one of the most promising circuit solutions for nanoelectronics. We show nanowire crossbars do not scale well in terms of logic density and speed. We consequently propose a crossbar cell design based on judicious use of silicon nanowire crossbars with microscale pitches and small dimensions. The crossbar cell is compatible with the conventional MOSFET fabrication and standard cell-based integration. We evaluate logic circuits using crossbar cells and show that they can improve density by more than fourfold over the traditional MOSFET circuits with the same process technology, while achieving close performance and over threefold power reduction.

Accelerometers have appeared on many camcorders, cameras and mobile phones. We present algorithms that estimate camera movement from accelerometer readings and apply the estimation to significantly improve the compute-intense motion estimation in video encoding. We have implemented a working prototype that simultaneously captures video and three-axis acceleration data. The video is then compressed with a reference MPEG-2 encoder, modified to incorporate the accelerometer readings to assist motion estimation. Our experimental data shows a two to three times speed improvement for the entire encoding process, in comparison with full search.

Wireless interfaces are major power consumers on mobile systems. Considerable research has improved the energy efficiency of elongated idle periods or created more elongated idle periods in wireless interfaces, often requiring cooperation from applications or the network infrastructure. With increasing wireless mobile data, it has become critical to improve the energy efficiency of active wireless interfaces. In this work, we present micro power management (μPM), a solution inspired by the mismatch between the high performance of state-of-the-art 802.11 interfaces and the modest data rate requirements by many popular network applications. μPM enables an 802.11 interface to enter unreachable power-saving modes even between MAC frames, without noticeable impact on the traffic flow. To control data loss, μPM leverages the retransmission mechanism in 802.11 and controls frame delay to adapt to demanded network throughput with minimal cooperation from the access point. Based on a theoretical framework, we employ simulation to systematically investigate an effective and efficient implementation of μPM. We have built a prototype μPM on an open-access wireless hardware platform. Measurements show that more than 30% power reduction for the wireless transceiver can be achieved with μPM for various applications without perceptible quality degradation.

Energy efficiency has become a very important and challenging issue for resource-constrained mobile computers. In this article, we propose a novel dynamic software management (DSOM) framework to improve battery utilization. We have designed and implemented a DSOM module in user space, independent of the operating system (OS), which explores quality-of-service (QoS) adaptation to reduce system energy and employs a priority-based preemption policy for multiple applications to avoid competition for limited energy resources. Software energy macromodels for mobile applications are employed to predict energy demand at each QoS level, so that the DSOM module is able to select the best possible trade-off between energy conservation and application QoS; it also honors the priority desired by the user. Our experimental results for some mobile applications (video player, speech recognizer, voice-over-IP) show that this approach can meet user-specified task-oriented goals and significantly improve battery utilization.

Radio Frequency Identification (RFID) technology has the potential to dramatically improve numerous industrial practices. However, it still faces many challenges, including security and reliability, which may limit its use in many application scenarios. While security has received considerable attention, reliability has escaped much of the research scrutiny. In this work, we investigate the reliability challenges in RFID-based tracking applications, where objects (e.g., pallets, packages, and people) tagged with low-cost passive RFID tags pass by the RFID reader's read zone. Our experiments show that the reliability of tag identification is affected by several factors, including the inter-tag distance, the distance between the tag and antenna, the orientation of the tag with respect to the antenna, and the location of the tag on the object. We demonstrate that RFID system reliability can be significantly improved with the application of simple redundancy techniques.

To address the productivity bottlenecks in power analysis and optimization of modern systems, we propose to treat power as asignal and leverage the rich set of signal processing techniques. We first investigate the power signal properties of digital systems and analyze their limitations. We then study signal processing techniques to detect temporal and structuralcorrelations of power signals. Finally, we employ these techniquesto accelerate the simulation of an architecture-level power simulator. Our experiments with the SPEC2000 benchmark suite show that it is possible to accelerate power simulation by 100X without introducing significant errors at various resolution levels.

Mobile phone users have to deal with limited battery lifetime through a reciprocal process we call human-battery interaction (HBI). We conducted three user studies in order to understand HBI and discover the problems in existing mobile phone designs. The studies include a large-scale international survey, a one-month field data collection including quantitative battery logging and qualitative inquiries from ten mobile phone users, and structured interviews with twenty additional mobile phone users. We evaluated various aspects of HBI, including charging behavior, battery indicators, user interfaces for power-saving settings, user knowledge, and user reaction. We find that mobile phone users can be categorized into two types regarding HBI and often have inadequate knowledge regarding phone power characteristics. We provide qualitative and quantitative evidence that problems in state-of-the-art user interfaces has led to under-utilized power-saving settings, under-utilized battery energy, and dissatisfied users. Our findings provide insights into improving mobile phone design for users to effectively deal with the limited battery lifetime. Our work is the first to systematically address HBI on mobile phones and is complementary to the extensive research on energy-efficient design for a longer battery lifetime.

Ubiquitous connectivity on mobile devices will enable numerous new applications in healthcare and multimedia. We set out to check how close we are towards ubiquitous connectivity in our daily life. The findings from our recent field-collected data from an urban university population show that while network availability is decent, the energy cost of network interfaces poses a great challenge. Based on our findings, we propose to leverage the complementary strength of Wi-Fi and cellular networks by choosing wireless interfaces for data transfers based on network condition estimation. We show that an ideal selection policy can more than double the battery lifetime of a commercial mobile phone, and the improvement varies with data transfer patterns and Wi-Fi availability.

Customized multimedia content delivery has become one of the most desirable applications to mobile device users. However its intense usage of wireless and user interfaces poses a great challenge to device usability and battery lifetime. In this paper, we describe our design and implementation of an energy-efficient multimedia messaging system to address this challenge. We construct a hierarchical system for users to access multimedia content, leveraging widely available short message service (SMS), an embedded system-based new interfacing device, and the Internet capability of mobile devices. Being the first of its type, the new system not only reduces energy overhead but also improves the usability of the service.

Most methods for hardware power estimation operate at the register-transfer level (RTL) or lower levels of design abstraction. Since cycle-accurate functional descriptions (CAFDs) are being widely adopted in integrated circuit (IC) design flows, power estimation can potentially benefit from the inherent increase in the efficiency of a cycle-based functional simulation. However, attempts at power estimation for functional descriptions have suffered from a poor accuracy because the design decisions performed during their synthesis lead to an unavoidable large uncertainty in any power estimate that is based solely on the functional description. The authors propose a methodology for a CAFD power estimation that combines the accuracy achieved by structural RTL power estimation with the efficiency of cycle-accurate functional simulation. This goal is achieved by viewing a CAFD as an abstraction of a specific known RTL implementation that is synthesized from it. The authors identify correlations between a CAFD and its RTL implementation and "back-annotate" information into the CAFD solely for power estimation. The resulting RTL-aware CAFD contains a layer of code that instantiates virtual placeholders for RTL components and maps values of CAFD variables into the RTL components' inputs/outputs, thus enabling efficient and accurate power estimation. Power estimation is performed in the proposed methodology by simply cosimulating the RTL-aware CAFD with a simulatable power-model library that contains power macromodels for each RTL component. Techniques to further improve the speed of CAFD power estimation through the use of control-state-based adaptive power sampling are presented. The authors have implemented and evaluated the proposed techniques in the context of a commercial C-based hardware design flow. Experiments with a number of large industrial designs (up to 1 000 000 gates) demonstrate that the proposed methodology achieves an accuracy very close to RTL power estimation with two-three orders of magnitude speedup in estimation times

Mobile computers, such as cell phones and personal digital assistants (PDAs), have dramatically increased in sophistication. At the same time, the desire of consumers for portability limits batters size. As a result, many researchers have targeted hardware and software energy optimization. However, most of these techniques focus on compute-intensive applications rather than interactive applications, which are dominant in mobile computers. These systems frequently use graphical user interfaces (GUIs) to handle human-computer interaction. This paper is the first to explore how GUIs can be designed to improve system energy efficiency. We investigate how GUI design approaches should be changed to improve system. Energy efficiency and provide specific suggestions to mobile computer designers to enable them to develop more energy-efficient systems. We demonstrate that energy-efficient GUI (E/sup 2/GUI) design techniques can improve the average system energy of three benchmarks (text-viewer, personnel viewer, and calculator) by 26.9, 45.2, and 16.4 percent, respectively. Average performance is simultaneously improved by 23.7, 34.6, and 19.3 percent, respectively.

Power has become a major concern for mobile computing systems such as laptops and handhelds, on which a significant fraction of software usage is interactive instead of compute-intensive. For interactive systems, an analysis shows that more than 90 percent of system energy and time is spent waiting for user input. Such idle periods provide vast opportunities for dynamic power management (DPM) and voltage scaling (DVS) techniques to reduce system energy. In this work, we propose to utilize user interface information to predict user delays based on human-computer interaction history and theories from the field of psychology. We show that such a delay prediction can be combined with DPM/DVS for aggressive power optimization. We verify the effectiveness of our methodologies with usage traces collected on a personal digital assistant (PDA) and a system power model based on accurate measurements. Experiments show that using predicted user delays for DPM/DVS achieves an average of 21.9 percent system energy reduction with little sacrifice in user productivity or satisfaction.

Power has become a critical concern for battery-driven computing systems, on which many applications that are run are interactive. System-level voltage scaling techniques, such as dynamic voltage scaling (DVS) and adaptive body biasing (ABB), have been shown to reduce energy consumption effectively. Previous works on DVS and ABB exploit low CPU utilization of the processor to drive voltage scaling. This has become inadequate for modern interactive applications involving high CPU usage. In this work, we target computer responsiveness during voltage scaling to exploit more opportunities for energy reduction. Instead of CPU utilization, we use the user-perceived latency, the delay between user input and computer response, to drive voltage scaling. Considering the tradeoff between energy consumption and computer responsiveness during voltage scaling not only reduces energy consumption effectively, but also ensures good computer responsiveness for interactive applications. Experimental results show that for the 70nm technology, during the execution of seven commonly-used interactive applications, the energy consumption of the processor using user-perceived latency driven DVS is reduced by an average of 37.3%, and the user-perceived latency by an average of 18.3%, compared to CPU utilization driven DVS. If both DVS and ABB are performed simultaneously based on the user-perceived latency, then the energy consumption is reduced by another 38.9% compared to when DVS is performed alone, while maintaining a similar computer responsiveness level. We have implemented user-perceived latency driven voltage scaling under Linux with X Window system. However, the methodology is extensible to other operating systems as well.

Although computing hardware has become increasingly more powerful, computer responsiveness is still an important issue due to multi-tasking and software bloat. We propose a holistic approach for improving computer responsiveness through user focus-aware resource management for CPU, memory, disk I/O, and graphics processing. Previous approaches only address one or two of these problems simultaneously. To the best of our knowledge, our work is the first to address disk I/O scheduling for better responsiveness. It also offers better solutions for the other problems. We also exploit the user-perceived latency to perform dynamic voltage scaling of the CPU to reduce power consumption at run-time without sacrificing responsiveness. We implemented our approach in the Linux/X Window system (henceforth referred to as Linux/X) on an IBM Thinkpad R32 laptop with mobile Pentium 4-M processor, which has two performance levels with different frequency/supply voltage settings: high (30.0W at 1.8GHz/1.3V) and low (20.8W at 1.2GHz/1.2V). Experimental results show that user focus-aware resource management achieves a significant improvement in computer responsiveness and some in energy efficiency. For example, for TuxRacer, a video racing game, with GpsDrive, a navigation system, running in the background, it provides a reduction of 42.0% in user-perceived latency and 7.5% in energy consumption with respect to the Linux/X system.

Handhelds, such as smart-phones and Pocket PCs, have the potential to become the computing, storage, and connectivity hub, or Digital Hub, for pervasive computing. However, their current interfacing paradigms fall short of achieving this goal. To meet this challenge, we present the system and hardware design for a Bluetooth-based personal-area network (PAN) of low-power wireless interfacing devices. The network consists of a wrist-watch, single-hand single-tap multi-finger keypad, smart speech portal, and GPS receiver. These devices serve a handheld in a synergistic fashion, collectively providing the user with immediate and more natural access to computing power and enabling more and better services.

Energy efficiency has become a critical issue for battery-driven computers. Significant work has been devoted to improving it through better software and hardware. However, the human factors and user interfaces have often been ignored. Realizing their extreme importance, we devote this work to a comprehensive treatment of their role in determining and improving energy efficiency. We analyze the minimal energy requirements and overheads imposed by known human sensory/speed limits. We then characterize energy efficiency for state-of-the-art interfaces available on two commercial handheld computers. Based on the characterization, we offer a comparative study for them.Even with a perfect user interface, computers will still spend most of their time and energy waiting for user responses due to an increasingly large speed gap between users and computers in their interactions. Such a speed gap leads to a bottleneck in system energy efficiency. We propose a low-power low-cost cache device, to which the host computer can outsource simple tasks, as an interface solution to overcome the bottleneck. We discuss the design and prototype implementation of a low-power wireless wrist-watch for use as a cache device for interfacing.With this work, we wish to engender more interest in the mobile system design community in investigating the impact of user interfaces on system energy efficiency and to harvest the opportunities thus exposed.

We propose an algorithm for efficient threshold network synthesis of arbitrary multioutput Boolean functions. Many nanotechnologies, such as resonant tunneling diodes, quantum cellular automata, and single electron tunneling, are capable of implementing threshold logic efficiently. The main purpose of this work is to bridge the current wide gap between research on nanoscale devices and research on synthesis methodologies for generating optimized networks utilizing these devices. While functionally-correct threshold gates and circuits based on nanotechnologies have been successfully demonstrated, there exists no methodology or design automation tool for general multilevel threshold network synthesis. We have built the first such tool, threshold logic synthesizer (TELS), on top of an existing Boolean logic synthesis tool. Experiments with 56 multioutput benchmarks indicate that, compared to traditional logic synthesis, up to 80.0% and 70.6% reduction in gate count and interconnect count, respectively, is possible with the average being 22.7% and 12.6%, respectively. Furthermore, the synthesized networks are well-balanced structurally. The novelty of this work lies in the introduction of the first comprehensive synthesis methodology and tool for general multilevel threshold logic design.

Interconnects (wires, buffers, clock distribution networks, multiplexers, and busses) consume a significant fraction of total circuit power. In this paper, we demonstrate the importance of optimizing on-chip interconnects for power during high-level synthesis. We present a methodology to integrate interconnect power optimization into high-level synthesis. It not only reduces datapath unit power consumption in the resultant register-transfer level architecture, but also optimizes interconnects for power. We take into account physical design information and coupling capacitance to estimate interconnect power consumption accurately for deep submicron technologies. We show that there is significant spurious (i.e., unnecessary) switching activity in the interconnects and propose techniques to reduce it. Compared with interconnect-unaware power-optimized circuits, interconnect power can be reduced by 53.1% on average, while overall power is reduced by an average of 26.8%, with negligible area overhead. Compared with area-optimized circuits, the interconnect power reduction is 72.9% and overall power reduction is 56.0%, with 44.4% area overhead. The power reductions are obtained solely through switched capacitance reduction (no voltage scaling is assumed).

We propose an algorithm for efficient threshold network synthesis of arbitrary multi-output Boolean functions. The main purpose of this work is to bridge the wide gap that currently exists between research on the development of nanoscale devices and research on the development of synthesis methodologies to generate optimized networks utilizing these devices. Many nanotechnologies, such as resonant tunneling diodes (RTD) and quantum cellular automata (QCA), are capable of implementing threshold logic. While functionally correct threshold gates have been successfully demonstrated, there exists no methodology or design automation tool, threshold logic synthesizer (TELS), on top of an existing Boolean logic synthesis tool. Experiments with about 60 multi-output benchmarks were performed, though the results of only 10 of them are reported in this paper because of space restriction. They indicate that up to 77% reduction in gate count is possible when utilizing threshold logic, with an average reduction being 52%, compared to traditional logic synthesis. Furthermore, the synthesized networks are well-balanced, and hence delay-optimized.

Cycle-accurate functional descriptions (CAFD) are being widely adopted in integrated circuit (IC) design flows. Power estimation can potentially benefit from the inherent increase in simulation efficiency of cycle-based functional simulation. Currently, most approaches to hardware power estimation operate at the register-transfer level (RTL), or lower levels of design abstraction. Attempts at power estimation for functional descriptions have suffered from poor accuracy because the design decisions performed during their synthesis lead to an unavoidable, large uncertainty in any power estimate that is based solely on the functional description. We propose a methodology for CAFD power estimation that combines the accuracy achieved by power estimation at the structural RTL with the efficiency of cycle-accurate functional simulation. We achieve this goal by viewing a CAFD as an abstraction of a specific, known RTL implementation that is synthesized from it. We identify correlations between a CAFD and its RTL implementation, and "back-annotate" information into the CAFD solely for the purpose of power estimation. The resulting RTL-aware CAFD contains a layer of code that instantiates virtual placeholders for RTL components, and maps values of CAFD variables into the RTL components' inputs/outputs, thus enabling efficient and accurate power estimation. Power estimation is performed in our methodology by simply co-simulating the RTL-aware CAFD with a simulatable power model library that contains power macro-models for each RTL component. We present techniques to further improve the speed of CAFD power estimation, through the use of control state-based adaptive power sampling. We have implemented and evaluated the proposed techniques in the context of a commercial C-based hardware design flow. Experiments with a number of large industrial designs (up to 1 million gates) demonstrate that the proposed methodology achieves accuracy very close to RTL power estimation with two-to-three orders of magnitude speedup in estimation times.

One important way to reduce power consumption is to reduce the spurious switching activity in a circuit or circuit component, i.e., activity that is not required by its specified functionality. Given a scheduled behavior and functional unit binding, we show that spurious switching activity can be reduced through proper register binding using retentive multiplexers. Retentive multiplexers can preserve their previous select signal values in the control steps in which the select signals are don't cares. A functional unit, in which spurious switching activity is completely eliminated, is called perfectly power managed. We present a general sufficient condition for register binding to ensure a set of functional units to be perfectly power managed. This condition not only applies to data-flow intensive behaviors, but also to control-flow intensive behaviors. It leads to a straightforward power-managed (PM) register-binding algorithm, which uses this condition to preserve the previous values in the input registers of a functional unit during the states in which the unit is idle. The proposed algorithm is general and independent of the functional unit binding and scheduling algorithms. Hence, it can be easily incorporated into existing high-level synthesis systems. For the benchmarks we experimented with, an average 40.7% power reduction was achieved by our method at the cost of 6.9% average area overhead, compared to power-optimized register-transfer level circuits, which did not use PM register binding.

A significant fraction of the software and resource usage of a modern handheld computer is devoted to its graphical user interface (GUI). Moreover, GUIs are direct users of the display and also determine how users interact with software. Given that displays consume a significant fraction of system energy, it is very important to optimize GUIs for energy consumption. This work presents the first GUI energy characterization methodology. Energy consumption is characterized for three popular GUI platforms (Windows, X Window system, and Qt) from the hardware, software, and application perspectives. Based on this characterization, insights are offered for improving GUI platforms, and designing GUIs in an energy-efficient and aware fashion. Such a characterization also provides a firm basis for further research on GUI energy optimization.

In this paper, we describe a comprehensive high-level synthesis system for control-flow intensive as well as data-dominated behaviors. We propose a new control-data flow graph model to preserve the parallelism inherent in the application, as well as to facilitate high-level synthesis. Our algorithm, which is based on an iterative improvement strategy, performs clock selection, scheduling, module selection, resource allocation and assignment simultaneously to fully derive the benefits of design space exploration at the behavior level. The system can be used to optimize area, power or energy, by selecting the cost function accordingly. Experimental results show that for energy-optimized designs, energy is reduced by up to 79.4% (an average of 42.2%), with an average of 24.8% area overhead, compared to area-optimized designs. For power-optimized designs, power is reduced by up to 70.8% (an average of 56.7%), with an average of 25.2% area overhead, compared to area-optimized designs. No Vdd scaling is performed to obtain the above results.

Interconnects (wires, buffers, clock distribution networks, multiplexers and busses) consume a significant fraction of total circuit power. In this work, we demonstrate the importance of optimizing on-chip interconnects for power during high-level synthesis. We present a methodology to integrate interconnect power optimization into high-level synthesis. Our binding algorithm not only reduces power consumption in functional units and registers in the resultant register-transfer level (RTL) architecture, but also optimizes interconnects for power. We take physical design information into account for this purpose. To estimate interconnect power consumption accurately for deep sub-micron (DSM) technologies, wire coupling capacitance is taken into consideration. We observed that there is significant spurious (i.e., unnecessary) switching activity in the interconnects and propose techniques to reduce it. Compared to interconnect-unaware power-optimized circuits, our experimental results show that interconnect power can be reduced by 53.1% on an average, while reducing overall power by an average of 26.8% with 0.5% area overhead. Compared to area-optimized circuits, the interconnect power reduction is 72.9% and overall power reduction is 56.0% with 44.4% area overhead.

A circuit or circuit component that does not contain any spurious switching activity, i.e., activity that is not required by its specified functionality, is called perfectly power managed (PPM). We present a general sufficient condition for register binding to ensure that a given set of functional units is PPM. This condition not only applies to data-flow intensive (DFI) behaviors but also to control-flow intensive (CFI) behaviors. It leads to a straightforward power-managed (PM) register binding algorithm. The proposed algorithm is independent of the functional unit binding and scheduling algorithms. Hence, it can be easily incorporated into existing high-level synthesis systems. For the benchmarks we experimented with, an average 45.9% power reduction was achieved by our method at the cost of 7.7% average area overhead, compared to power-optimized register-transfer level (RTL) circuits which did not use PM register binding.

High performance Mandarin voice dialer is much more difficult than its English counterpart to achieve, especially on inexpensive hardware as ASIC. One way to improve its performance is to incorporate rejecters into the system. In our study, an MLP based postprocessor, an a posteriori probability estimator, is applied after HMM Viterbi recognition. Poor utterances, which are recognized by HMMs but have low a posteriori probability, are then rejected. Rejecting 4.9% of all the testing utterances, the MLP rejector boosts the HMM-based system's single digit accuracy from 97.1% to 99.6% for the Mandarin voice dialer, a ten-syllable speaker independent task. The performance is better than those of rejection based on linear discrimination, anti-digit models or likelihood ratio.

A dynamic neural network architecture based on the time-delay neural network and the convolutional neural network is originated. The dynamic network achieves much better performance than those of MLP and TDNN when dealing with syllable recognition. Such performance is also comparable to that of the more popular HMM method.