How to calculate rolling volatility of returns data using Exponential Weighted Moving Average?

I have a dataset that consists of the closing price and date for 10 years. I want to calculate the rolling 3-day volatility of returns data using Exponential Weighted Moving Average. Please mention how to approach this problem and if possible give the solution in excel. The data look like this:

Date        ClosingPrice  Logreturns     Squaredlogretuns  Weights   
1-01-2000    50              -                 -               -
2-01-2000    51            0.008600172     7.3963E-05        0.28    
3-01-2000    52            0.008433168     7.11183E-05       0.2016  
4-01-2000    52.5          0.00415596      1.7272E-05        0.145152    
5-01-2000    53            0.004116566     1.69461E-05       0.10450944  
6-01-2000    59            0.046576142     0.002169337       0.075246797 


Weights * Squaredlog returns      Volatility

2.07096E-05                              -
1.43375E-05                              -
2.50707E-06                               -
1.77103E-06                        0.006270979
0.000163236                        0.013485222

In the above table, I took lambda (weights) =0.72 and used a formula to decrease the weights exponentially. Finally, I multiplied the weights with squared log returns.

I added up all the rows of (weightssquaredlogreturn) column to calculate the overall variance of the EWMA for this data but how to do it for a rolling 3 day period. Above I added up the last 3 rows of (weightsquaredlogreturns) column for each day to get the variance and took a square root of it to get the Standard Deviation or Volatility of rolling 3-day period.