Fortunately, this type of interference is rare and unlikely to be caused by normal transmissions. However, radar transmitters in low flying aircraft are capable of physically damaging the electronics of a radioastronomy receiving system. Typically a power level at the radiometer input of 0.1 W would burn out the varactor in a parametric amplifier. This corresponds to a power flux of 10 W/m2 (10 dB(W/m2)) at 1400 MHz, if the interfering source is outside the main beam of the antenna. The corresponding power flux-density, assuming a bandwidth of 27 MHz (1400 to 1427 MHz) would be 3.7 x 10-7W/(m2. Hz) or -64 dB(W/(m2. Hz)). If the antenna is accidentally pointed at such a strong interference source, the power flux-density that could burn out the input varactor must be reduced by the gain of the antenna. Category 2: Relatively strong interference which is easy to recognize. Usually this is the case if the interference power is stronger than the noise power in the radiometer input. This type of interference is fatal to observations, and there is no doubt that it is interference. There is no choice but to discard the data. Typically, interference power fluxes above - 110 dB(W/m2) belong in this category. With an assumed bandwidth of 27 MHz (the 1413.5 MHz band), the corresponding power flux-density is - 184 dB(W/(m2 . Hz)). Category 3: Very low-level interference with a very low interference to noise ratio (less than —20 dB) that cannot be recognized. The long integration times required to bring the wanted signal out of the noise will mask the characteristic features of the interfering signal. It cannot be recognized as interference and erroneous data result; this type of low-level interference is therefore particularly harmful. Furthermore, because the radio astronomer cannot determine by examination of the data that he has encountered interference, there is no possibility of identifying the source. 9.2 Interference reduction techniques A number of techniques designed to reduce the effects of interference can be tried by the radio astronomer. Some of these are obvious and straightforward, and some are clever, complicated and often time consuming. They all suffer from the problem of being of limited usefulness. In general, the employment of interference reduction techniques leads to the need for more observing time. 9.2.1 Filtering techniques. Unwanted signal energy outside the observed band is rejected in the radioastronomy receiver by using bandpass filters. Normally, when the interfering signals are of low intensity and do not cause non-linear operation anywhere in the system, limiting the observed passband by a filter in the IF channel is useful. Since the IF frequencies are relatively low, typically between 100 to 300 MHz, relatively steep skirt selectivity is possible. However, limiting the observing band decreases the sensitivity of the system which is proportional to yAZ Filtering in the radiometer input is also used, particularly when the potentially interfering signals are strong. Again decreased sensitivity is the result, both because of the narrower bandwidth and because of the insertion loss of the filter which, when inserted in the receiver input, adds to the loss and the noise temperature of the system. Since the filtering takes place at the observing frequency, adequate skirt selectivity may be a problem. Typically, about 75% of an allocated band remains available after a reasonable IF filtering (see Report 696) which corresponds to a sensitivity loss of about 13%. If input filtering is needed, sensitivity reduction of a factor of 2 or more could be expected. In order to obtain 100 dB reduction of the midband response at the band edge of the radio- astronomy band, one needs three or more 8-section filters in the IF channel. Although such a filter system is feasible, there are important phase considerations to be taken into account when observing with an antenna array or interferometer. This makes it questionable whether filtering really is a viable general solution to the band edge interference problem. Bandpass filters do not, of course, alleviate the in-band interference problem. 9.2.2 Observing techniques It is possible, and often necessary, to reduce the effect of interference by using special observing techniques. One possibility is to repeat the observation several times, with the assumption that the interference is present only occasionally. It is also useful to move the telescope on and off the source during an observing run, assuming that the interference is present all the time. Both methods are of limited usefulness, because of the assumptions one has to make of the behaviour of the celestrial source as well as of the interference. They also increase the required observing time by a sizeable factor. Furthermore, the
RkJQdWJsaXNoZXIy MTU5NjU0Mg==