Proposed CRCD/EI Model for SP-COM

     At ASU, the signal processing and communications curriculum (Fig. 1 – right hand side) includes four undergraduate courses, i.e., the required junior-level signals and systems (EEE 303) and random signal analysis (EEE 350), and the senior electives digital signal processing (EEE 407) and communication systems (EEE 455). We have obtained approval from the department (see chair’s letter) to develop a new senior 3-credit elective course entitled “Introduction to Signal Processing for Communications Research” (EEE498).  

Digital communications algorithms associated with the proposed    CRCD modules. The pertinent proposal sections describing the algorithms and CRCD modules are also cited. A planned CRCD research experiences includes capstone UG projects that will use a comprehensive MATLAB implementation of this system.

      EEE 498 will be a modular course that will consist of four comprehensive research modules (CRMs). These interrelated CRMs are: source coding, channel coding, time-varying signaling, and multi-carrier modulation (Fig. 1 – left hand side).   All these modules fit into the digital communication transceiver framework shown in Fig. 2.  More compact junior-senior modules  (JSM) will also be formed and injected in the junior/senior SP-COM curriculum. Finally demonstration modules (DM) will be developed to provide outreach (WISE) and freshmen and sophomore summer research experiences.   All of the created modules will be self-contained, i.e., they will have adequate tutorial exposition, prerequisite reading assignments, a well defined task, and a laboratory.  Each CRM will have six lectures and one computer research laboratory. A capstone final project will also be part the course.  The course will be developed and taught by the four Co-PIs.  The four research laboratory exercises will be developed based on our award winning Java-DSP software [Spa02,Ko03] technology (http://jdsp.asu.edu). 

      The CRCD model in Figure 1 presents an approach where SP-COM research is immersed across the undergraduate signal processing and communications curriculum.  The large 6-lecture 1 lab CRMs, the 1  lecture/1 lab modules JSMs, and the J-DSP demonstration modules (DMs) draw their content from SP-COM research that consists of our own research activities [Spa03,Spa94,Fou03, Bel02a, Pain00, Ram02, Dum98,Dum99a,Pap02b,Tep01], and from relevant research presented by others in refereed publications. 

        One CRM will be developed by each of the PIs; the four CRMs will form the core content that goes in the new pre-approved senior level elective EEE498 called “Introduction to Signal Processing for Communications Research.”  JSMs consist of one streaming video lecture and one J-DSP web-based lab and are essentially a subset of the CRMs. JSMs will be injected in our four undergraduate courses (303, 350, 407, 455).  Finally, DMs are formed for the summer camps that provide SP-COM research experiences to freshmen and sophomores, as well as for summer outreach activities. The demo modules will consist mostly of demonstrations of research concepts and a small hands-on exercise. Pre-requisite structures are soft. The only hard pre-requisites for EE students enrolling in EEE 407, 455, and 498 are the signals and systems, EEE 303, and the random signal analysis (EEE 350). EEE 303 and 350 are required in the EE curriculum for all students.  For students from non- EE disciplines that desire to enroll in EEE 498 without taking 303 and 350, we will provide an opportunity to complete pre-requisite material through a series of streaming video lectures that have 303 and 350 content.  This pre-requisite streaming content has been developed for our on-line distance learning program.   Details on the curriculum and the modules are given in subsequent sections.

  Curriculum Development

      We will integrate SP-COM research across the curriculum in a manner that promotes the objectives of this CRCD. We will have the following curriculum development activities

   A freshman/sophomore SP/COM summer research experience: We start with a freshman-sophomore experience using summer SP-COM camps and research demo modules (DMs). We will stimulate interest in the SP-COM area using examples such as a cellular phone paradigm. We will provide a laboratory experience where students experience a computer simulation of such a system with capabilities to vary parameters and observe results. Specifics on the DMs will be presented later. The summer camps will take place immediately after the spring semester and an extensive campaign will be launched during the pre-registration period to promote these sessions among freshman and sophomore students.  We will also coordinate these sessions and run them in parallel with summer camps for high-school prospects that are organized by the Women in Science and Engineering (WISE) (see letter).

   A junior research experience by incorporating modules in the required 303 and 350 courses: Both courses are offered every semester. Towards the end of each semester, we will inject select one small self-contained JSM in each course. A signaling module will be developed by Dr. Papandreou-Suppappola and injected in 303 because of the inherent association of this course with basis functions and Fourier transforms.  A channel coding module will be developed by Dr. Duman and injected in 350 because of the association of the topic with probability theory. The JSM will contain at least one  streaming web lecture. The Co-PI in charge (see Fig. 1) of the JSM introduces the essentials of the research module, prepares and disseminates notes, requires the students to visit a designated web site with the prerequisite material.  A reading assignment and a canned computer laboratory experiment using our own J-DSP web-based simulation environment is assigned.  The assignment will be part of the homework structure of the courses and will be graded. 

   A senior level experience by injecting SP-COM research modules in 455 and 407: The experience and structure provided is again similar in scope and length with those given in 303 and 350 and both 407 and 455 are also offered every semester.  Towards the end of each semester Dr. Spanias will develop and assign a self-contained  JSM on source coding because of the association of the topic with DSP content. Dr Tepedelenlioglu will develop and assign a small self-contained JSM on multicarrier systems because of the association of 455 with modulation systems.  Again the JSM structures and scheduling of activities (assignments, usage of J-DSP simulation, etc) are similar to those described in 303 and 350.  The assignment will be part of the homework structure of the courses and will be graded. 

   A dedicated new course (498) that will be developed especially to promote the goals of this CRCD: TThe new and approved course will be co-developed and co-taught by the four Co-PIs and will consist of four large SP-COM modules (source coding (Spanias), signaling (Papandreou-Suppappola), Channel coding (Duman), and multi-carrier systems (Tepedelenlioglu)). The interrelated course topics fit well in the digital transceiver framework of Fig. 2 and the syllabus of the course is attached in the single page documents of this proposal.  The course will be offered once per semester,  Each CRM has 6 lectures and 1 lab. We will also permit graduate students to attend it as with all the senior level courses in our institution.  The course will contain a formally proposed semester long capstone research project. Students will be required to turn in a final report and give a research presentation that will be evaluated.  Although the formal pre-requisites will be 303 and 350 for this course, the Co-PIs will allow students from other information technology related disciplines to attend it, subject to approval and provided that they complete pre-requisite linear systems and probability theory content embedded in other web-based modules that have been previously created to support the ASU distance learning program.  Finally, we note that we will develop two modalities for the course.  The default modality will be synchronous real-time delivery of carefully prepared lectures in power-point format.   Although the presentations will be delivered by all four participating Co-PIs, the content of all four modules will be coordinated and the course will be administered by the PI. The second modality, which is planned for later, will consist of streaming video presentations of the same lectures (video over power point) and will target a distance learning audience. The labs will be based on J-DSP.

   Preparation of J-DSP online CRCD laboratories in all the above modules and courses: All of the CRCD curriculum courses and modules involve at least one self contained computer laboratory experience.  We intend to use our very own web-based simulation environment for these laboratories. The J-DSP concept and software were originally developed at ASU to provide online DSP laboratory experiences to distance learners [Spa00]. J-DSP was rated as one of three top educational resources for 2003 by the NEEDS committee and was recognized as such at the IEEE Frontiers in education conference.  The J-DSP editor [Spa01] is an object-oriented DSP simulation environment built from the ground up for education.  All DSP signal manipulation functions appear in J-DSP as blocks that are brought into the simulation environment by a drag-and-drop process. Signal flow is established by linking the blocks. Students can access J-DSP on the web, perform computer laboratory exercises, and create electronic lab reports that they submit using special servelets [Spa00] integrated with J-DSP (see demos on J-DSP web site jdsp.asu.edu). We emphasize that J-DSP is not simply a “number-crunching” tool but an integrated simulation/animation environment for total web delivery of simulations, animations, and streaming video lectures. It is both a tool for students performing on-line labs and for instructors [Spa01] that create web lectures with coordinated simulations. Therefore it is distinctly different than typical "number-crunching" environments (such as Simulink®, SPW®, etc.). Additional Java functionality will be developed and embedded in J-DSP to support the sistinct SP-COM functions that support the modules. This task will be carried by the research assistants (supervised by the PIs).

   Capstone projects and development of a comprehensive MATLAB research simulation of a digital transceiver system:  In addition to the self-contained compact modular J-DSP laboratory exercises above that are intended for hands-on exposition to distinct SP-COM concepts, we will prepare a comprehensive computer simulation of a digital transceiver system (Fig. 2) to support capstone projects in 498.   This simulation will permit the students to run the more computationally intensive research tasks, test their own code, run and manipulate long and complex data structures and bit streams, evaluate precisely computational complexity, assess requirements for real-time and power consumption, computer bit error rates, etc.

   CRCD Activities for Recruitment and Retention of Minorities: We will set up summer camps specifically tailored to recruit students from underrepresented groups by collaborating with on-campus ASU minority recruitment organizations.  Dr. Papandreou-Suppappola has worked closely and systematically with the ASU Women in Applied Science and Engineering (WISE:   http://www.eas.asu.edu/~wise/) and the Office of Minority Engineering Programs (OMEP). She will work closely with WISE and OMEP reps that run and attend minority conferences.  WISE and OMEP have committed to include CRCD J-DSP-based demonstration modules in their high school orientation events to stimulate interest of high school students from underrepresented groups  in engineering (see letter).  We plan to develop and/or adapt CRCD DM especially to expose high school students to a variety of SP-COM experiments. We will use the cellular telephone paradigm and additional demonstrations that were developed as part of our previous CCLI program. We will coordinate publicity with WISE and OMEP that have databases and publicity content that targets native Indian, Hispanic, and African American high school students in the Phoenix metro. We will assess and document this activity carefully by running specific questionnaires and instruments and by monitoring with WISE/OMEP applicants to ASU engineering programs that participated in these activities.

   Details on additional CRCD Research Activities in EEE 498

      The new course EEE 498 will have a capstone CRCD project that will be proposed formally by a team of 4 students.  Teams will include three undergraduate students and one graduate student from the SP-COM group designated by the PIs. The role of the graduate student will be to provide information and training to get the students started in the process.  Portions of the SP-COM work done by the UG students will be associated with a thesis component of the graduate student.

     The objectives of the 498 course will not only include exposure to particular research topics but also a process where undergraduates will be exposed to basic elements, practices, logistics of research, such as:

·         Research proposal and report development

      The projects PIs will embed in the course research proposal and report writing with specifications similar to those of NSF

·         Working on team oriented research tasks with collaborators and supervisors

      The Co-PIs will define capstone projects in EEE 498 that will require students to work in teams with graduate students. 

·         Promotion of ethics of research

      We will have formal lectures to educate undergraduate students and promote practices that are consistent with the notion of research promoted by NSF and IEEE 

·         Exposure to research simulation and productivity tools

      We will embed students in several productivity tools that include software environments, web-based tools, statistical tools, search tools

·         Research literature survey practices using web and library resources

      Students will be exposed to successful literature survey models.  The students will be required to conduct and document research in a paper in the SP-COM area

·         Reporting formats of research results with standard IEEE/ACM specifications

      We will promote research formats similar to those of IEEE and ACM and institute a process where papers are subject to response to reviewers and rebuttals as in refereed publications

·         Research presentations to peers, supervisors, and potential sponsors

      We will form formal structure where teams have to present their research in an annual report

   Pedagogical Foundations for SP-COM research transitions to the curriculum (Ledlow, PIs)

      Pedagogical foundations and strategies for the transition of research to the curriculum will be formed with the assistance of Dr. Susan Ledlow who is a budgeted instructional specialist.  Current research reviewed by the National Research Council and documented in the book [Bran 2000] indicates that: to develop competence in an area of inquiry students must: a) have a deep foundation of factual knowledge, i.e. the necessary SP-COM theory, b) understand facts and ideas in the context of a conceptual framework, e.g. in this case the target wireless application, c) organize knowledge in ways that facilitate retrieval and application. Similarly, the Boyer Commission on Educating Undergraduates (2000) notes that many students graduate having accumulated the necessary number of courses, but still lacking a coherent body of knowledge or any inkling as to how one concept might relate to others. This CRCD project recognizes the importance of incorporating the most current research on learning when developing new curricula that include modules that will be used by “old faculty.” Therefore connections of the injected research modules with the existing theory and topics covered in the affected EE courses will be a priority in this CRCD.   In addition, our challenge in developing the innovations in the new SP-COM curriculum is to move away from the transmission of information in the traditional lecture format to the facilitation of discovery- to create “significant learning experiences” [Fink 03].  The organized hands on research activities will be organized in a manner that stimulates critical thinking and by promoting a “what if” type of approach.  We will emphasize and employ active learning (especially inquiry-based learning) as a means of helping students learn for understanding [Chic 87] [Bran 00] [Light01]. The curriculum will focus on the development of higher order thinking skills [Shul99, Bean96. Bloo57; Paul95] and is built around authentic tasks. The modules and planned hands-on activities will be developed in a manner that engages UG students both independently and collaboratively in the discovery of SP-COM principles.  

Involving Undergraduates in Ongoing SP-COM Research Projects

      Even though of our undergraduate students will have some experiences in research through some of the modules that will be embedded in the required curriculum courses 303 and 350, other mechanisms for immersing students in ongoing SP-COM research will be used as well.  Involving undergraduate in open-ended active research outside of the aforementioned curriculum structure is indeed a difficult proposition.  It will require extensive planning and a process where tasks are broken into small and manageable sub-tasks.  Most of the ongoing research projects in SP-COM involve some form of a software simulation which is often sufficiently modularized to allow the novice to obtain at least a structured simulation experience where parameters of a certain algorithm are varied and the performance changes can be observed.  We have several such computer simulations (mostly with MATLAB)), where complete transceiver systems can be simulated in a manner that allows the students to explore the performance of the communication systems by varying pertinent algorithmic parameters under different channel noise and fading conditions etc. 

  Summer SP-COM Research Activities for Qualified Undergraduate Students

      In addition, to the freshman/sophomore summer camp research experience (Fig. 1), we plan summer research activities for the more qualified senior level students.   Since faculty and graduate students tend to have minimal loads in terms of teaching and taking classes respectively, several research programs tend to become very active in the summer. Hence it makes sense to provide opportunities to maintain the research activities of qualified (those that have done reasonably well in EEE 498) undergraduate students. We will  

-  budget summer salaries for at least 5 undergraduates to be involved in SP-COM research

-    work with our industry collaborators to create summer internships

-    work with the ASU Office of Corporate Leaders that specializes in internships to place some of these students in industry where communications and signal processing research is being conducted

 Details of the Body of Research and Planned Modules for this CRCD

2.7.1 Source Coding of Speech (Spanias)  

      Advanced source coding algorithms for speech signals are now part of most trendy communications and multimedia applications.  Source coding of speech or otherwise known as speech coding [Spa94, Spa03c] is an area that has received a lot of attention both in the communications and in the signal processing research communities.  Early applications were military (secure communications) and while defense applications continue to use extensively such algorithms, speech coding has found applications in large scale and volume in cellular telephony and internet telephony (voice-over-IP).  The purpose of speech coding algorithms is to represent speech signals with a minimum number of bits while maintaining the perceptual quality and intelligibility of the signal.   The research frontier in the area is to develop robust low-complexity algorithms that maintain high quality signal at bit rates below 4 Kbits/s.  The speech coding algorithm paradigm lends itself well to education and the PI of this project has used the model extensively to introduce applications in a DSP class [Att03].  There is a large body of research that was generated at ASU in [Spa04,Spa03c,Ahm98,Ahm99,Ahm01,Att02,Pai04] as well as a number of standards for cellular communications that are continuously evolving [Chu03] (also see [Spa03c] and references therein.   Most of the speech coders are based on source-system configurations.

      Early models for speech coding represent speech with a time-varying source-system mechanism that captures the interaction of the vocal chords and lungs with the vocal tract.  The source-system model represents voiced speech (e.g., vowels) with a filter excited by periodic impulses and unvoiced speech (e.g., consonants) by a filter excited by a noise-like waveform.  Early systems produced synthetic speech and were eventually replaced by the modern source-system configuration known as code excited (CELP algorithms) algorithm.  Variants of the CELP algorithm are embedded in most cellular telephony standards (e.g. GSM, CDMA, ITU etc). The CELP is a rich educational and research paradigm. It involves mathematical concepts such as optimization and pattern matching; statistical concepts such as maximum likelihood estimation and linear prediction; principles of acoustics such as perceptual relevance and filtering; computational complexity; issues relating to joint source/channel coding  

2.7.1.1 CRM for Source Coding to be Embedded in EEE498

      For the CRM the students will be exposed to a series of six lectures on speech coding research and students will be asked to carry a comprehensive survey and make an attempt to define the state of the art. The large module will provide background information, an introduction to linear prediction and vector quantization as used in CELP, and examples of successful CELP coders.   Students will be asked to carry a literature survey and develop a report. Students will be asked to explain some of the current methods for source coding and identify and report a series of research problems that the research community regards as unsolved or challenging. A simulation task then will be carried in the form of a computer lab to provide intuition on parametric representations of speech.  A lab exercise will be developed for a simulation in J-DSP.  The J-DSP lab exercise will take students through specific tasks that include:  developing a complete speech analysis synthesis simulation by integrating (calling) the functions developed for speech. Students will experiment with different channel conditions and their effect on the coder. Students will also be able to visualize Java code for CELP and develop knowledge on issues associated with computational complexity

2.7.1.2 JSM for Source Coding to be Embedded in EEE407

      JSM in EEE 407 will be injected to show how a filter and a frequency response are used to parameterize speech to reduce the bit rate.  The module will include notes, a streaming lecture on the CRCD web site, demos, and a reading assignment.  The JSM will be part of the homework structure of EEE 407.  A research task in a SM for EEE 407 would be to explore how the algorithm would behave in the presence of noise and how the performance of the coder changes for different bit rates.  Exposure to code could also make students aware of issues of computational complexity and concepts associated with fixed-point real-time implementation.  A J-DSP lab will also be developed and assigned.

2.7.1.3 DM for Source Coding in Freshman/Sophomore Summer Camps

      The demo module will be developed based on our Java technology. Again the DM will be self contained with a package of notes, a purely qualitative lecture on source coding, and some additional web resources.  The J-DSP demo will demonstrate the operation of an algorithm in the context of a cellular phone and high light the importance of software technologies and relevant math theory in cellular communications. Students will also receive promotional material for the JSM and CRM with a hope that they will seek more information in the SP-COM later in the curriculum.

2.7.1.4  A Sample Capstone Research Project in Source Coding

      We make available the MATLAB transceiver infrastructure and we require students to test different algorithms for source coding and evaluate and document their performance in the presence of channel noise, fading, and acoustic background noise.   A literature review of relevant algorithms is required.

  2.7.2 Research in Channel Coding (Physical Layer Communications (Duman)

      Our objective is to expose our undergraduate students to research in various aspects of physical layer digital communications. Specifically, we have identified channel coding (in particular, turbo codes), equalization algorithms (for wireless and recording channels), and wireless communications (specifically, space-time coding techniques) as the areas of interest. These topics align very well with our ongoing funded projects ([Dum01, Dum00]) and ongoing research efforts.  We first plan to introduce the basics of digital communications in two lectures, discuss various channel models, signaling techniques, and identify specific issues to motivate the proposed research activities. The basic block diagram is shown in Figure below. The three topics are then covered in the remainder of the course module (approximately 2 lectures will be allocated for each). Channel coding is the process of adding controlled redundancy to the transmitted information sequence to protect the data against various impairments (e.g., noise, interference, fading) introduced by the channel. Specifically, turbo coding is a very powerful and practical channel coding technique which provides near optimal performance over noisy channels. The research problems involved are the invention of various code design principles, study of their bit error rate performance over different channels, development of new and efficient iterative decoding algorithms (see, for example, [Dum98, Dum99a, Dum99b, Ste00, Bel02]. The students will get exposure to this research by getting exposed to carefully designed smaller tasks managed by graduate students involved heavily on the project. As for specific applications, we have identified turbo codes for recording channels, and for wireless (fading) channels.  Many practical channels of interest (e.g., magnetic recording channels) introduce what is called intersymbol interference (ISI) where the received signal is corrupted by interference from adjacent symbols transmitted. The ISI, if not taken care of properly, degrades the error rate performance of the system considerably. Therefore, various channel equalization techniques are employed in practice to remove the ISI in practical communication systems. Borrowing from our experience (see, for example,  [Dum01a, Dum01b, Zha03b, Zha03c]) with signaling for magnetic recording channels, we will expose our undergraduate students to different channel equalization algorithms and the performance improvement offered by them.  Signaling for wireless channels is an important area of research in physical layer communications. Specifically, diversity techniques that are used to combat the multipath fading inherent in wireless links are very important technique. Among the different diversity techniques, space-time coding combines the spatial diversity, and channel coding to improve the bit error rate performance of the wireless communication systems (see [Ste01, Ghr03, Ste01a, Ste03b, Bah02, Bah03a, Bah03b]). The students will conduct simulations of various space-time block and trellis coding systems using JDSP software to get exposure to the types of improvements offered by diversity techniques, specifically, by space-time coding for signaling over multipath fading (wireless) channels.

2.7.2.1 CRM for Channel Coding to be Embedded in EEE498

     The students will be exposed to channel coding, equalization and space-time coding research by means of six lectures, computer exercises using JDSP, and interactive class activities. The graduate students involved in the related research projects will work with the undergraduate students taking the course to carry out specific tasks. In particular, the students will conduct simulation based studies to understand the benefits of various channel coding techniques, including, convolutional codes and turbo codes. Different channel models (e.g., additive white Gaussian noise channel and fading channel) will be used, and the bit error rates offered by different coding schemes will be evaluated. Furthermore, the students will be exposed to Intersymbol interference and wireless channels, equalization and space-time coding techniques via simulations. In addition to the simulation based study, the students will be required to present a comprehensive survey of the coding/equalization techniques in order to gain thorough theoretical understanding of the subject.  

2.7.2.2 JSM for Channel Coding to be Embedded in EEE350

      In EEE 350, the students will use Monte Carlo simulation techniques to see “probability at work.” Specifically, they will conduct Monte Carlo simulations using JDSP for the simplified digital communication systems. They will estimate the bit error probabilities with no channel coding, with simple convolutional codes and some turbo codes over additive white Gaussian noise channels using the existing JDSP functionality.  

2.7.2.3 DM for Channel Coding in Freshman/Sophomore Summer Camps

      A JDSP exercise will illustrate the potential and use of channel coding techniques for different levels of noise. Along with a complete channel coding system simulation, we will also demonstrate the workings of a typical decoding algorithm by graphical illustrations.

  2.7.2.4  A Sample Capstone Research Project in Channel Coding

      A thorough study of turbo codes and turbo coded modulation schemes will be conducted for wireless channels both theoretically and via simulations. For the theoretical component the students will conduct a survey, and for the simulation study they will use the JDSP developed at ASU.  

2.7.3        Time-Varying Signaling (Papandreou-Suppappola)

       Time-varying signals such as linear and nonlinear chirps have spectral characteristics that vary  with time [Pap02c,Pap98], and are naturally found in many applications that are of interest to undergraduate students such as music. Examples of such signals include speech, radar, sonar and seismic acoustic waves, biomedical signals like ECG, biological signals such as bat and dolphin echolocation sounds; they also include the impulse response of a fast varying wireless communication channel. As the Fourier transform cannot provide multiple frequencies that occur at the same time, the appropriate tools for analyzing time-varying signals are time-frequency representations [Pap02c]. Recently, time-frequency methods have been introduced to provide advanced methodologies in order to improve wireless technology. For example, linear chirps have been used recently for wireless communications due to their inherent immunity against Doppler and multipath fading. In our signal processing for  wireless communications research, we applied time-varying signals for modulation in multi-user  systems  [She03,Mac03,Gup01,She03b, Mac03b,Gup03]. Specifically, we developed novel signaling schemes using linear and nonlinear chirps for code division multiple access (CDMA) techniques that significantly reduce multiple access interference over traditional modulation methods.  We have also used chirps to obtain multipath diversity and to estimate the complex wireless channel in order to increase bit error rate performance. We plan to transfer our state-of-the art research into our proposed curriculum innovations with the following modules. 

2.7.3.1  CRM for Time-Varying Signaling to be Embedded in EEE498

      We plan to prepare six lectures together with homework problems,  J-DSP-based exercises and interactive class activities. Our lectures, accompanied by power point notes, will follow the block diagram in Figure 2 and will cover: introduction of modulation techniques for single and multiple user wireless systems, introduction of time-varying signals and time-frequency  representation processing tools, time-varying modulation techniques and their use in multi user systems, estimation of time-varying signaling parameters, diversity techniques and the use of time-varying signals for multipath diversity, and channel estimation using time-varying pilot signals.  The students will be required to read relevant research papers as well as to find their own resources for reading material in collaboration with assigned graduated students. The learning objectives of this module are for the students to learn the basic concepts of modulation in wireless communication systems and to understand the role of advanced time-frequency processing techniques in increasing system performance. We will design various assessment methodologies to ensure that the students achieve their objectives, and based on the feedback from the students, the module will be continuously maintained.

2.7.3.2 JSM for Time-Varying Signaling to be Embedded in EEE303

      This will include a subset of the CRM module with streaming video and pre-packaged exercises as an assignment. It will be aimed at introducing to junior undergraduate students time-varying signals and some processing tools based on the Fourier transform such as the spectrogram. The exercise will guide the students through the use of linear chirps for modulation in a multi-user mobile system scenario, and the students will have to test various parameters and determine the corresponding  system performance.

2.7.3.3 DM for Time-Varying Signaling in Freshman/Sophomore Summer Camps

      This will be a simple J-DSP demonstration that will guide students through a mobile communication system implementation where linear chirp-based modulation is used in a multi-user environment.  The DM module will also be used as part of our WISE outreach activities.

2.7.3.4  A Sample Capstone Research Project in Time-Varying Signaling

      Matlab software for the multi-user system with chirp-based modulation will be provided to the students, but they will be required to provide a literature review on multiple access systems. The students will be asked to vary the number of users and the chirp parameters such as the frequency modulation rate and compare the difference in bit error rate performance.

2.7.4 OFDM for Wireless Communications (Tepedelenlioglu)

      Orthogonal Frequency Division Multiplexing (OFDM) makes use of the Fast Fourier transform (FFT) and is used in fast telephone modems, e.g., in the Digital Subscriber Line (DSL) service or wireless local area networks (WLAN). This is the technology of choice for several standards (IEEE 801.11 for WLANs and DSL for internet). Our research has mainly focused on improving diversity and reducing the effects of carrier frequency offset [Ma00, Tep02a, Tep02b, Tep04a,Tep04b,Wan00]. The basic idea of OFDM is that “convolution in the time-domain is multiplication in the frequency domain”, which enables low-complexity equalization of the (convolutional) multipath channel. Hence, the basic principle behind OFDM is very accessible to anyone that has taken a signals and systems course. We now elaborate on our modules that distill these research ideas and incorporate them in the curriculum.

2.7.4.1 CRM for OFDM to be Embedded in EEE498

      This module will focus on why OFDM works as a frequency-domain communication technique, and draw from the basic notions of signals and systems. The lectures will begin by motivating OFDM as a technology by giving what every-day technologies it is used in (e.g., wireless card of a laptop). Then the system will be explained using basic notions taught in signals and systems, including convolution, Fourier Transform and frequency-response. We will develop MATLAB software that can illustrate how an OFDM system works, and generate bit error rate curves. The success of the course will be measured by supplementary homework problems and a final quiz.

2.7.4.2 JSM for OFDM to be Embedded in EEE455

       The OFDM simulation setup will be used in EEE455, our undergraduate course in probability, for illustrating the notion of bit error rate in a realistic system setup. This will be explained in the context of Monte Carlo simulations of the bit error rate, and give students a chance to see how calculation of probabilities using software can have practical applications.

2.7.4.3 DM for OFDM in Freshman/Sophomore Summer Camps

      This will be a simple graphical software demonstration of the workings of a digital communication system. The notion of transmitter, receiver, and channel will be qualitative defined. The performance over channels of different quality will be shown graphically, and illustrated using transmitted images and voice to show the deterioration of  the quality over different channels

2.7.4.4  A Sample Capstone Research Project in OFDM

       This project will involve a literature survey of some of the impairments of OFDM system due to rapid changes in the channel, and presence of frequency offsets. These impairments will be observed through MATLAB simulations and quantified in terms of bit error rate. The goal of the project is to make sure that the students understand how different non-ideal conditions lead to degradation in performance, and have them think about the mitigation of these conditions.  

CRCD Assessment Components (Haag, Ledlow, all PIs)

The hypothesis is that students will be at the following position after they complete the modules

-    have an appreciation of the state-of-the-art and research frontier in the areas covered

-    students can ask intelligent questions after observing a research presentation in an area covered in EEE 498

-    students are introduced to various techniques and elements associated with the notion of doing research (proposal writing. research reporting, task scheduling, deliverables)

-    students are able to perform an appropriate literature review in the field that they were exposed

-    students appreciate research and are excited by the prospect of doing open ended work

-    students are interested in attending graduate schools and performing thesis research

For every module and for the entire CRCD we will devise a qualitative and quantitative assessment process.   Assessment will be done through electronic web tools, pre- and post- quizzes, presentations, one-to-one interviews, and ordinary in-class testing.

  Electronic Web-based assessment:  We will adapt and make use of web-based tools and methods that we have developed for a previous CCLI project [Spa03a] and we will perform both concept specific and general assessment of the knowledge that the students gained.

  Pre-and-post assessment quizzes: Concept specific assessment will consist of specific technical questions that will assess whether the students understood some of the basics covered in the modules. A pre-test will establish a reference before exposure to the module and a post test will examine their knowledge after they have been exposed to the module content.  Processing of the responses will be carried to identify statistically significant differences. These practices proved very successful in signals and systems classes [Buc03, Kat01] and other areas [Eva01,CIAT]

  General assessment will examine whether the students gained knowledge on general methodologies for conducting research. They will assess the tools used, accessibility of content, and the overall experience.

  One-to-one interviews [Ale02.Per03] will be used to assess the learning of each of the project participants.  We will ask specific questions to assess the level of learning, we will ask the students to pose questions on research and judge whether they have acquired a body of knowledge that enables them to ask intelligent research- oriented questions; we will assess whether they begun develop an intuition on emerging applications

Lab- Specific Assessment: This will be specific to each lab.  The cognitive goals specific to each lab will necessitate the use of their own instruments.  Once lab objectives have been clearly defined, an assessment activity to test the outcome of that objective will be chosen.  After completed labs are assigned we will assess as follows: a) On-campus Assessment and distance learning participants assessment (later in the 3^rd year when it becomes available); b) Industrial Assessors: Motorola (see letter) and General Dynamics contacts will help assemble industry evaluations, and c) Presentation and publication: Improvement using reviews from our papers and the questions by the audience in our conference presentations will be used to improve our labs.

Student and peer evaluations: we will ask students to both evaluate the course, evaluate the instructors, evaluate the content of each module, evaluate the performance of their peers

 Identify from student graduate school applications whether we have been able to motivate students to attend graduate school:  Two of the PIs are member of the graduate committee that has access to graduate application data.  Also all the PIs are in a team that evaluates applicants for the ASU graduate program and hence have access to data as well.  As for students that apply to other programs, we will device a process to monitor how many apply. Usually most students with graduate school aspirations in the SP-COM area ask SP-COM faculty to serve as references on their graduate school applications.

  Other components of the skill building and assessment tasks, Annual CRCD Workshop

      We will organize an annual CRCD workshop where students present formally their CRCD capstone projects, research papers that indicate our latest CRCD research results. The audience will consist of the faculty involved in this CRCD, the combined industry/university oversight committee, undergraduate and graduate students relating to this project and SP-COM in general, and other invited industry and university representatives. Through this conference, we will institute and assess the following

  -    how students present their work in a formal setting (criteria are quality of materials, validity of data and results generated, poise and confidence, ability to respond to critical questions)

-    a process where the CRCD UG student audience will be required to pose questions to the presenting students and the Co-PIs will assess the ability of the CRCD students to pose relevant questions

-    a competition will be designed to award the student with the best presentation.  The selected student will be awarded a free registration to a major research conference in the SP-COM field

-    the currency of topics as perceived by the oversight committee.  The oversight committee will be asked to assess formally whether the modules reflect knowledge considered as current in the SP-COM field

Experience and Capabilities of Principal Investigators and Collaborators

       This culturally and gender-diverse team is strong and consists of an IEEE Fellow and IEEE field series award winner, three NSF CAREER award winners, and two budgeted senior personnel faculty collaborators that will guide the assessment and pedagogy tasks. The lead PI, Andreas Spanias, will be the director of this CRCD and will overlook the entire project milestones. Andreas Spanias has expertise and publications in speech and audio coding and will be responsible for developing modules associated with source coding.  He will also be in change of immersing JSMs in the DSP class (EEE 407), will coordinate the new EEE498, organize the CRCD workshop, and coordinate all J-DSP activities for the DMs.   Antonia Papandreou-Suppappola has expertise in time-varying signaling and she will develop the CRM, JSM, and DM on signaling for communications. She will also coordinate the minority outreach program with WISE (see letter).  Tolga Duman has expertise in channel coding and will develop the CRM, JSM, and DM for the channel coder.  Cihan Tepedelenlioglu has expertise in multi-carrier systems and ultra wideband communications.   He will develop the CRM, JSM, and DM for multicarrier modulation.

Table 1: Principal Investigators and Collaborators

Management plan project; CRCD Time-line and milestones

      The CRCD project director is Andreas Spanias. As the PI he will have the overall responsibility for overlooking all tasks, milestones, and all aspects of this projects.  The project is scheduled for three years. It involves module development, software infrastructure development in Java and in MATLAB, extensive assessment, dissemination, and coordination with the oversight committee and various other entities that will collaborate with us on CRCD tasks.  A tasks/milestone/time-line table is given below.