In this paper, we propose a non-coherent CFAR decision rule named OSTM-CFAR (Order Statistic Trimmed Mean Constant False Alarm rate) detector in a Pareto type II clutter with unknown scale and shape parameters. This is obtained after adequate selection of two scale invariant functions in the general hypothesis test. The resulting test statistic provides rapid calculation because is independent of any estimation method or numerical integration. Through simulated and real data, CFAR properties of suggested CFAR detector are maintained against both scale and shape parameters of clutter model. Besides, detection probability of both OSTM-CFAR and Bayesian-CFAR procedures are sketched together where the proposed one demonstrates its superiority in presence of interfering targets.