No abstract

This paper introduces a new method of spectrum analysis of nonuniformly sampled signals based on a Dirichlet transform. This transform, which generalizes the Z-transform, is defined as:The Dirichlet transform is better suited to the spectrum analysis of nonuniformly sampled signals than the Z-transform. The Dirichlet spectrum preserves information about sampling instants because a nonuniformly sampled signal x(t,) is not treated as a sequence of samples {x(t,)} , but as a function of the sampling instants f n . The paper presents basic properties of the Dirichlet transform in case of periodical and non-periodical nonuniform sampling and also shows the inverse Dirichlet transform. The authors suggest use of this transform in analysis of the nonuniformly sampled signals processed by digital filters with constant and varying delays, systems which modulate sampling instants, sampling jitter etc. The paper also shows examples of Dirichlet transforms of several signals.

of IEEE AES SYSTEMS (5, 9, page 30) Dr. AIfonso Farina of Rome, Italy, provided a very interesting preview of this pap= At the International Radar ECCM Workshop organized by Mr. Johmton in conjunction with RADAR-90, Dr. W Kiembowski of R1.Z described ECCM features of a number of Polish radars. Accordingly, it wns suggested that Part I of this paper be revised to provide information on these newly disclosed rndars including photographs and more information on Polish marine d a m . Although this deloyed publication, it w s felt that the newly released information w s of sufJcient importance to wrranl the delay.It is hoped thut further information will be available for our readers in future issues.