Skip to Main Content
One of the major claims in the paper of Gershman et al. (see ibid., vol.44, p.361-6, 1996) is that Hung-Turner (1983) projections do not adequately suppress wide-band interference. This statement is only true if too few snapshots are used to estimate the interference subspace. For Hung-Turner projections the use of more snapshots than the rank of the interference is called "averaging over the dimension". The authors implicitly and wrongly claim that even with "averaging over the dimension", Hung-Turner projections perform poorly in the presence of wide-band interference. In this comment, we show theoretically and by simulation that Hung-Turner projections adequately suppress wide-band jammers if the number of snapshots is equal to or larger than the rank of the interference. Here, the same sensor array processing model and assumptions are used. Gershman et al. reply that they proposed a new modification of the Hung-Turner (HT) algorithm improving the robustness of this algorithm against wide-band and moving jammers. They also add that their approach can be applied to other adaptive array algorithms as well, and the critical comment contains a misinterpretation of their results and conclusions.