Information Electronics combines the most recent innovations in Information Processing with the actual realization of System-level Electronics. System Electronics requires the design and development of Analog, Digital, and Software aspects of Embedded Systems. We have an overall activity in the physical layer development of Networking for the Embedded System Information Super Highway. A system level view of electronics development typically uses a top down design approach, which favors the use of digital and software blocks as opposed to analog blocks. However, a top down design only approach often fails to meet the system objectives in our research projects, which by definition, are often "bleeding edge" or "grand challenge" embedded system scenarios. The difficulty of a top down approach is the assumption that lowest level blocks specified, can in fact, be designed and built. As digital blocks replace analog blocks, the remaining analog blocks become nearly impossible to build, considering the usual system develop constraints (performance, size, power, cost, design time, etc.).

     Thus, a key sustaining capability of Information Electronics is to be able to translate system level requirements into achievable, next generation, analog circuit performance capabilities.

     Embedded System Networking involves a surprisingly large number of analog Input/Output (I/O) issues. A good system testbed that manifests these issues is a Wireless Sensor Node in an embedded network. Such a node contains a variety of Codecs (coder/decoder blocks) to bridge the Analog and Digital domains. There are audio and video codecs for multi-media content, RF codecs for wireless communication, and of course, the variety of sensor/actuator modality codecs (position, force, environment, etc.). Therefore, a major need for a Wireless Sensor Node is a highly re-configurable codec block, or a programmable Universal Codec.

     Combining several codecs along with processors, RF, and baseband analog circuits encompasses many of the issues surrounding System on a Chip (SOC) development methodologies. Software engineering methodologies encompass many of the steps that are used to develop SOC hardware/software for a Wireless Sensor Node. Not only are software engineering methods used to develop the network software and processor firmware executed by the node in its network, but software engineering methods are critical for translating requirements specifications with/into Hardware Description Language (HDL) constructs to accelerate the implementation and re-use of physical hardware.

     Even though there are many skills that are needed in our research activities, the paramount skill remains as the ability to remove the constraints on the overall system objectives that are produced by the analog I/O issues. Thus, we often reformulate the overall system descriptions and algorithms to be consistent with the realities and limitations of high performance, mixed signal circuits. In order to be intimately familiar with mixed signal realities, we place a major emphasis on placing our design ASICs into actual systems for evaluation and cataloguing. 

     In addition, our students often participate in strategic internship programs that are aligned with their research training and writing. Given the sheer amount of effort involved in SOC methods for embedded system networking, we work with a variety of groups at other universities, government laboratories, and industrial R & D teams. A common goal among these groups is the advancement of mixed signal development in an SOC framework. Nevertheless, although such a concern typically enforces a strong hardware focus, we strive to keep abreast of the latest innovations in information theories, statistical signal processing, and software engineering. For example, building a miniature embedded system, such as a hearing aid to selectively filter out someone's voice (for example, an overbearing mother-in-law), requires creative innovation in information processing, cutting edge circuit design, and people-skills. Interestingly, many difficult sensing tasks share this common aspect of dual innovation in Information and Electronics.

-BACK-