Fourier Analysis: Creating A "Virtual Laboratory" Using Computer Simulation

Introduction As the 21st century approaches, businesses continue to demonstrate their near-insatiable demand for newer computer technology. Annual investment in hardware and software products is estimated to be $350 billion [1996 figures] (Maddox, 1997). This continued growth has created tremendous job opportunities for MIS graduates (Hube, 1996) and put enormous pressure on colleges and universities to not only keep their IS programs up to date, but to expand them as new applications for technology are developed The costs associated with "keeping up" can be formidable Gone are the days of decades past when a fledgling IS program could get by with some dumb terminals, access to a PDP-11, and perhaps a small cluster of Apple II computers Learning to program in the languages of the day (e.g., FORTRAN, COBOL, Assembly) and some familiarity with the common operating systems created a marketable MIS graduate. Today, a technologically conscious school (with a bottomless endowment) could build specialized laboratories to support activities in Multimedia, Networking, CASE, Collaborative Technology, CAD/CAM and so on. Sadly, most institutions cannot afford such an array of facilities and are typically limited to clusters of general-purpose personal computers Instructors that wish to develop learning activities beyond what these general-purpose clusters can support often must look for creative solutions to their problems This paper reports on efforts to create a "virtual" computer laboratory where specialized activities can be simulated using general-purpose facilities. Specifically, this report will discuss how spreadsheet packages may be used to simulate digital signals and how common Local-Area Network (LAN) media might affect them. However, there are numerous other possibilities that this approach may be adapted to and the goal of this report is to provide an example of how learning activities may be developed or enhanced through computer modeling and simulation that might otherwise not be possible Problem The example used in this report was originally developed for use in a senior-level LAN class in a business-based MIS program The particular class is applied in nature and the students are required to complete weekly lab activities in addition to their normal class work. Budget limitations dictated that only inexpensive Ethernet cards be used. Compatible with the IEEE 802.3 standard these network cards operate at transfer rates of 10 Mega-bits/second. When originally developed this 10 Mbit rate was consistent with the speed of the popular computers of the day. However, as the speed of desktop computers has jumped from tens to hundreds of MHz in recent years the older Ethernet standard has become a bottleneck for many network users. 100 Mbit technology, so-called Fast Ethernet, is increasingly available, but its higher price is still a matter of concern. In addition to the higher hardware costs, the migration to Fast Ethernet also requires improved wiring, which adds its own set of costs to the equation One of the most common questions raised in the class is why the speed of an older Ethernet card cannot simply be increased (i.e., add a faster clock chip). The usual follow-up question is why the network media must also be upgraded when Fast Ethernet cards are introduced. As there is no simple answer to the questions a common response might be some short discussion about bandwidth along with the hope that the students would be satisfied and that the lecture could proceed. However, most students recognize that the coaxial cable used in LAN applications is similar to that used to carry 100 cable television stations into their homes and apartments and the simple answer was usually unsatisfactory [FIGURE 1 OMITTED] [FIGURE 2 OMITTED] [FIGURE 3 OMITTED] At the root of such questions lies the broader question of how are digital signals affected when their speed is pressed into the VHF and UHF ranges. …