ABSTRACT
This chapter addresses single-realization inference in a context entailing strong assumptions on the space and the point processes. It is concerned with estimation of Palm integrals. The chapter defines a spectral representation theorem is derived, spectral measures and density functions. It discusses the spectral representation corresponding representations for reduced second moment and covariance measures, and examines estimation of frequency-domain descriptors. It also examines two approaches to linear state estimation, a time/space-domain approach applicable on general spaces and a frequency domain/spectral analysis approach for stationary point processes. In Krickeberg the fundamental role of Palm measures in inference for stationary point processes was recognized and first explored systematically.
