Abstract :
Conventional techniques for target discrimination, by high resolution, show severe disadvantages when faced with ballistic missiles. High range resolution (HRR), for example, usually applies to a single, isolated target that moves with almost constant velocity. Ballistic missile targets are often multiple and tend not to move uniformly, so, for these targets, the usual HRR techniques must be adapted. A ballistic missile target can extend for 1 km or more while requiring a range resolution of less than 1 m to resolve it. Often the number of frequencies, available for HRR is restricted by the radar platform. For example, the case here discussed is restricted to 256 frequencies. This paper considers new methodologies for stepped frequency waveforms that have been designed to cope with ballistic missile characteristics. A 1 km profile could be generated, to a desired resolution, by a long sequence of stepped frequencies. Alternatively, it could be generated by concatenating (stitching together) a number of smaller profiles derived from shorter sequences sampled in different range bins. We have investigated two super-resolution techniques for long sequences and two for stitched sequences. These show that a long profile can be achieved. The technique of stitching short profiles has its problems. If the profiles are simply “bolted” together (“hard” stitching) then the inaccuracies of alignment can be severe. For this reason we have investigated so called “soft” stitching. This aligns the profiles and creates an image in which they overlap. The result is a 1 km long profile without severe inaccuracy due to motion
