A Framework for Human Activity Recognition Based on WiFi CSI Signal Enhancement

Algorithm 1

Antenna selection.

	Input: S—the sequential data of CSI which contain 3 antennas with 30 subcarriers
	W—the size of sliding window
	step—the step size of window movement
	Output: Sj—the sequential data of the antenna which are most sensitive to activity
	Step 1: for each antenna Sj in S
	Step 2: calculate the mean sequential data (Fj) of 30 subcarriers in Sj
	Step 3: Nj = the length of Fj
	Step 4: Ej = Ø
	Step 5: for (int k = 0; k + W ≤ Nj; k = k + step)
	Step 6: calculate the variance () of sequential data in sliding window
	Step 7: append to Ej
	Step 8: end
	Step 9: Rj = max (Ej) − min (Ej)
	Step 10: end
	Step 11: return antenna Sj whose corresponding Rj is maximum