This project extends the functionality of the "Spot", a next-generation wearable computer, whose development was initiated in an earlier Greenhouse program. A sensor module will be added to monitor certain aspects of the user (position, location, orientation, or biological signs). The sensor will also enable interaction with other objects in close proximity via wireless communication (Bluetooth). The Spot will be aware of the temperature of a room or the operational status of nearby equipment. It will also communicate with other devices such as a coworker's Palm Pilot. A graphical display module will also be added to improve the user's interaction with the Spot and enable multimodal interfaces to be developed. Software to support both the graphical interface and sensor module is part of the development plan.

This project will extend the functionality of a conventional PDA to make it a universal interface and controller to many of the electronic devices in the typical home or office. Rather than having a separate and distinct device interface to faxes, copiers, VCR's and the like, a user will be able to use his Palm or Handspring PDA to operate any device that is Bluetooth enabled. The user's PDA will recognize that there is another Bluetooth device in proximity, establish a link with that device, and serve as the user's interface to the device.

Moreover, the user will see only one interface for each type of device. For example, the PDA will provide the user with a single, customizable interface to use a copier, any copier. The user will interact with that familiar interface regardless of whether she is interacting with the Canon copier in her office, the Icon copier in her company's remote sales office, or the Xerox copier at her local Kinkos. The PDA will make the necessary translations for her standard copier interface to the target device.

The goal of this project is to develop a Complementary Metal Oxide Semiconductor (CMOS) optical image sensor with pixel-level analog-to-digital conversion (ADC) and pixel-level data storage. If successful, this research will lead to an alternative to current CCDs (Charge Couple Devices) that are too complex, too expensive, require too much power, and exhibit unacceptable noise levels at high resolutions for many commercial applications.

The goal of this project is to develop an optical switch on a chip. The design will incorporate Microelectromechanical (MEMs) Systems and SoC technologies and will not require the signal to be converted from optical to digital and back. This MEMS device will also monitor itself, specifically the position of the micro-mirror. The device will be able to accurately identify the position of the mirror and adjust itself appropriately to optimize signal quality.

The objective of this project is to develop the necessary CMOS SoC components for ad hoc networks in the picocell or femtocell range. These SoCs will require no onboard power (powered by the RF signal) and can be networked for distributed applications such as remote sensors. The focus of this follow-on program is to develop the on-chip antenna designs required in these applications. The project will deliver documentation, code and prototype devices. The devices will be demonstrated in two proof of concept applications.

This project will demonstrate the technical feasibility of proximity broadcasting in a high-bandwidth wireless network. The goal is to create a cost-effective pico-server that is essentially a Bluetooth radio transmitter with a range of 1 to 100 meters that can transmit multimedia content to anyone in proximity with a Bluetooth-enabled device. For example, a shopper in a store could receive coupons and sales information about products as he strolled down an aisle. Travelers could receive construction updates, weather advisories, or toll information as they drive past a transmitter on the highway.

The project will deliver a proof-of-concept prototype that uses off-the-shelf hardware and custom software. Software development will focus on solving problems inherent in proximity broadcasting such as user location detection, device interaction management, and collision detection and avoidance.

This work is focused on developing algorithms that help multimedia developers determine the minimal amount of arithmetic precision required to provide acceptable quality in video applications. These algorithms free the designer from dealing with the translation from floating point to fixed point operations, in order to minimize power requirements…a crucial objective when designing for mobile devices.

The current program focuses on implementing these algorithms at the compiler level. This project will extend an existing C++ compiler to handle lightweight arithmetic operations, and to support a semi-automated environment for exploring the proper precision trade-offs required to use the custom number formats developed in earlier work.

This project will develop a high speed CMOS analog-to-digital converter (ADC) circuit which will allow on-chip direct digitization of wideband radio frequency (RF) signals. The project features the Threshold Inverter Quantization (TIQ) techniques for faster ADC operation using standard CMOS logic circuitry preferred for SoC implementation. A major goal of the project is the design and demonstration of both 6-bit and 8-bit TIQ based flash ADC circuits.

This project focuses on the need for non-volatile storage in single chip (SoC) solutions. Because non-volatile memory is not currently available in the standard CMOS process, designers must choose either external memory, mask programmed solutions, or resort to data download. Many applications require only small amounts of permanent data storage (a few bytes to a few hundred bytes). The memory may be needed for storing trimming values for on-chip circuitry, parametric values, ID information, custom configuration parameters, etc. This project will design and develop a low-density non-volatile memory solution for standard CMOS processes.

This project addresses the growing need for better design tools for SoCs. As the number of IC's that can be integrated onto a single chip has grown, a new design challenge with respect to clocking and overall synchronization has developed in that the time-of-travel across a chip may now be longer than a single clock cycle. Future SoC designs will have to move data across on-chip networks in a way that is analogous to packets on an ethernet. The project team will develop an optimization tool which will consider technology parameters and area constraints while deriving the maximum speed of signal propagation for a bus design in that technology. The team will use the tool to develop new circuit designs and work toward circuitry that will support a complete on-chip micronet structure.