No abstract

The problem of obtaining 10ng sequences with finite alphabet and peaky aperiodic auto-correlation is important in the context of radar, sonar and system identification and is called the coded waveform design problem, or simply the signal design problem in this limited context. It is good to remember that there are other signal design problems in coding theory and digital communication. It is viewed as a problem of optimization. An algorithm based on two operational ideas is developed. From the earlier experience of using the eugenic algorithm for the problem of waveform design, it was realised that rather than random but multiple mutations, all the first-order mutations should be examined to pick up the best one. This is called Hamming scan, which has the advantage of being locally complete, rather than random. The conventional genetic algorithm for non-local optimization leaves out the anabolic role of chemistry of allowing quick growth of complexity. Here, the Hamming scan is made to operate on the Kronecker or Chinese product of two sequences with best-known discrimination values, so that one can go to large lengths and yet get good results in affordable time. The details of the ternary pulse compression sequences obtained are given. They suggest the superiority of the ternary sequences.

No abstract

An instrument which uses a new sampling technique has been constructed and used to study the electromagnetic induction effects in a laboratory scale model representing conductors in the presence of host rock. It measures, in a dipole‐dipole setup, the in‐phase and out‐of‐phase components of the anomalous field directly, and has provision for recording the same components on a chart recorder. Idealized ore bodies can be represented by solid conductors and the host rock, by an aqueous salt solution. The operating frequency can be smoothly varied and set at any value between 10 khz and 100 khz. The operating separation between the source and the receiver in the coplanar coil setup may be up to 21 cm at 10 khz and up to 29 cm at 30 khz or above. Measurements may be made to within ±0.1 percent of the free‐space field, though absolute accuracy is less. It is found that the presence of a conducting host rock may substantially enhance the observed EM anomaly if the introduction of the target in the host rock alters the geometry of induced current paths. If this geometry is unaltered, a simple screening effect involving a reduction and phase rotation of the anomalous field occurs.

No abstract