This purpose of this project was the design and construction of a testbed network for experimentation with spread spectrum communications in the 900 and 2400 MHz band. The current network consists of five fixed nodes and two portable nodes. Of the fixed nodes, three are located in the Purdue MSEE building, one is located in the Harold L. Michael Traffic Operations Laboratory, and one is located in the experimental traffic signal cabinet at the intersection of Stadium and Northwestern Avenues in West Lafayette. The testbed has been used to evaluate spread spectrum radio technologies from vendors Microwave Data Systems, GINA, and EnCom, as regards radio performance as a function of link distance and with varying levels of interference. The project has produced software that can be used to test any radio presenting a standard RS-232 interface to customer equipment. The testbed can also be interfaced with wireless channel emulator equipment located in the Wireless Communications Research Laboratory in order to test radio performance in multipath and fading environments. The field tests conducted to date have yielded the following conclusions regarding the design of spread spectrum interconnect for the traffic signal control application. First, as regards the need for line-of-sight between radio antennas, we were able to establish a link even when line of sight was not available although the resulting communication was somewhat error prone. If an additional robust protocol were designed these radios could work even in non-line-of-sight applications. The protocol would need to emphasize an efficient retransmission scheme as opposed to the use of more powerful forward error control codes (this observation comes from the link error statistics gathered from field testing). Second, it was noted that the RSSI (which stands for received signal strength indicator) value was not a completely informative measure of link quality. In other words, the link testing software found many instances of links where RSSI was relatively large yet many errors were incurred. Such situations always involved interference or suspected interference. Third, as regards the choice of antenna (Yagi versus omni) it is concluded that Yagi antennas (with their higher gain) were preferable in long distance rural environments. But in the presence of multipath propagation or interference (as in a town or city scenario) the omni was often preferable. Finally, interference tests showed that the MDS 9810 handled the interference well at the price of the throughput; the EnCom radio had a high throughput but also a high data loss rate in certain situations. It was concluded that if the application only needs between 6 and 14 kbps, the MDS radio is an ideal choice. For higher rate applications the EnCom radio would be preferred.

Report Number



wireless communications, spread spectrum, traffic signal control, SPR-2395

SPR Number


Project Number


File Number


Performing Organization

Joint Transportation Research Program

Publisher Place

West Lafayette, IN

Date of this Version