Correction to 'Optical orthogonal codes: Design, analysis, and applications' (May 89 595-604)

where the definition of Cl( * , . , ) is obvious from above. We note that C depends explicitly on XI that contradicts the analysis in the paper’. Hence, the test which minimizes .J+, is not necessarily an LRT. For the n-stage problem ( n 2 2) , it is easily seen that p.b.p.0. local strategies at time n alone are LRT’s in general. Also, if we restrict the local decisions made by each sensor to depend on its present observation and all its past decisions, then it can easily be shown that p.b.p.0. local strategies are all LRT’s with thresholds depending on the past decisions. In this setting, the results claimed in the paper’ are provable.