STOCHASTIC RESPONSE ANALYSIS OF A SOUND LEVEL METER TAKING INTO ACCOUNT A CHANGE IN DIMENSION IN SHANNONS SIGNAL SPACE

This paper discusses how the probability distribution of the power flu ctuation of an arbitrary non-Gaussian signal is affected by a change i n dimension (2 WT, with W: frequency bandwidth; T: duration of a signa l) in Shannon's signal space. Specifically, the probability distributi on of the mean squared output of a sound level meter in the slow mode, with dimension 2 WT (T=1 (s)) is predicted from that in the fast mode , with dimension 2 WT0 (T-0=0.125 (s)). The effectiveness of the theor y has been verified by application to actual road traffic noise data. The experimental results appear to have provided substantial proofs fo r its practical use.