Update KalmanForWave.h

bbn_wave_freq_m5atomS3/KalmanForWave.h

@@ -58,6 +58,13 @@
          T,
          0       ]
 
+  The issue with this model is that while it is hinted to oscilate around mid sea level,
+  there is no hint on for a scale of oscillations. So we depend on selecting variances and
+  covariances properly. Another way would be to use this method for finding wave frequency,
+  use trochoidal wave model in which apmplitude is derived from acceleration and frequency
+  (or from ratio of heave and acceleration) and use heave estimate from trochoidal wave
+  model to fuse with accelerometer data. Basically feed something into F matrix on each step
+  (or every few steps) from estimate provided by trochoidal wave model.
          
 */