EFFECTS OF RADIOFREQUENCY EXPOSURE (100-KHZ TO 500-KHZ) ON VERY-HIGH-SPEED DIGITAL DATA-TRANSMISSION SYSTEM USING A COPPER LOOP

This paper describes our investigation of the effect of radio frequenc y (RF) exposure on the performance of a very-high-speed digital data t ramission system called the Asymmetrical Digital Subscriber Lines (ADS L) using the uncoded quadrature modulation (QAM) scheme at a rate of a pproximately 1.6 Mb/s and which operates within the 100- to 500-kHz fr equency band. The effect of different wideband emissions within the AD SL band, includong the background rtdaio noise in the environment and emissions from light dimmers, flurosecent lamps, motors, and other hou sehold appliances on the customer's premises were investigated. Also, the effect of RF exposure caused by radiation from high-power licensed transmitters, such as LORAN-C, coastguard transmitters, and radiobeac ons, were investigated. A complete database of information for these t ransmitters was acquired. The coupled voltage caused by radiation from the transmitters was predicted by using a detailed electromagnetic in terference (EMI) model that was developed and experimentally verified. Field surveys near LORAN-C stations in New York, New Jersey, and Cali fornia were performed. For the study on spurious-emission-like backgro und radio noise, light dimmers, invisible fences, and emission from di esel engines, it was found that these sources inside the customer's pr emises may not generate potential interference to the system. Minimum distances (critical distance) that should be maintained between the AD SL receiver and the interfering radio transmitters (LORAN-C Radiobeaco ns) were obtained for the 18 ADSL test loops. Initial study of multipl e exposure caused by several transmitters radiating at different frequ encies, but all within the ADSL band, showed that such an effect shoul d not generate any potential interference within the 100- to 500-kHz b and. However, the effect should be studied in detail at higher frequen cy bands (such as the commercial AM broadcasting band). The conclusion on our study for an uncoded QAM ADSL system without any special signa l processing at the receiver is as follows: 16-QAM signaling generally performs better than 64-QAM signaling, even in the presence of RFI. T he ADSL system performed satisfactorily, when it was deployed at locat ions greater than the critical distances. Database indicated that on m ajority of the ADSL test loops, areas within the critical distances ar e mostly rural, where substantial amount of ADSL services may not be r equired.