The first operational system has helped to maintain a virtually continuous colour service without dropout caused by excessive CCI. Experimentation and measurements indicate that the aerial system is capable of rejecting up to twelve or so independent sources of interference. Note that independent channel frequencies of the same transmitter location correspond to independent sources. Figure 15a shows the effect of a picture of co-channel interference as obtained when the SABRE aerial pattern is held at its preset value and Fig. 15b shows the result after optimisation of the aerial. The improvement in CCI rejection is from around 20 dB for the preset pattern to around 45-50 dB for the optimised pattern.

Measurements have also been made during 1978-79 of the levels of interference on the output of the adaptive array and also of a fixed aerial array used as a reference. Results for a period of relatively high interference are shown in Fig. 16.

It can be seen that, under these conditions, the adaptive array achieves a performance 25 dB better than the fixed array and, if the rejection of the latter to interference is assumed to be 20dB, then it can be seen that the required performance of 45 dB or better is achieved.

By using the results to date, we can state with confidence that propagation mechanisms such as sea scatter and tropospheric scatter do not degrade the performance of the adaptive aerial, and in fact the adaptive array has considerable advantages over a fixed array. This is as would be expected; a fixed aerial system is unable to track the changes in apparent CCI direction caused by scattering processes. Since the rate of change of CCI bearing with time has been found to be relatively small, although the total change may be fairly large, the adaptive aerial, which adjusts itself by making small fixed steps at a constant rate, can track the sources of CCI.

Fig. 15 (above): Unretouched off-screen photographs of pictures [of Peter Sissons] obtained at Alderney using the SABRE array under conditions of high level co-channel interference: (a) before optimisation: (b) after optimisation. To obtain (a) the SABRE aerial pattern was held at its preset value. The CCI rejection improves from about 20 dB in the preset pattern to around 45-50 dB in the optimised pattern.