Try to read the help of the xcorr function: it can be used to plot temporal cross-correlation as well as autocorrelation of a discrete-time signal. Note that such function provides to you only temporal informations (autocorrelation versus time lag). To get  statistical informations, you must have different time-realizations and assume, for example, their indipendence: following this approach, the sample mean of the autocorrelation function follows a well known (gaussian) distribution from who is easy to get coincidence intervals. 

Thank you Marco Bilucaglia, I will try it.

Hi,

Here is a snippet of the code,

t=0:3:24

y="Your correlation value"

plot(t,y)

@R.A. Joseph,  as you know that the Correlation coefficient has value in between 

[-1 to 1], so how can I select the Correlation coefficient value to get a plot like the one shown in the question. thanks