Autocorrelation is a function of time difference as it matches the signal with its delayed version.

According to the theorem spectral density of a signal x(t) and auto correlation function makes a fourier tranform pair.

According to its properties autocorrelation is maximum at origin.

A power signal is periodic signal and its function is a real even and non negative function as per the definition.

According to Parseval/s theorem the energy spectral density function can be given as equal to the square of the magnitude of the signal x(t).

Autocorrelation is a function that matches a signal with its delayed version.

Energy spectral density gives the signal energy equal to the area under the waveform energy spectral density versus frequency curve.

Power spectrum gives the distribution of variance of a signal in the frequency domain, sampled into spectral components.

Spectral density function is plotted against frequency domain and spectogram is a graph which is plotted against time domain.

A power signal usually is a periodic signal. But power spectral density of non periodic signal can be calculated by truncating it and observing it in the range of (-T/2,T/2).