You should estimate how long your frig and other loads run each day and hence how many amp hours hours they use. Then compare that with REAL output amp hours from two solar cells. I'd GUESS you can only replace 10% to 20% of your daily requirement at the 100 amp hour daily load rate. (Don't use that guess, work the actuals.)

The WIFI-only model will not do this since it estimates > its position from cellular towers and wifi hot-spots. > > I wouldn't want to run the length of the ICW with only an iPad, but you > certainly could do it.

My algorithm is supposed to estimate and correct the different effects that an OFDM symbol might experience.

That'll make it possible that the channel estimate done on the early part of the OFDM symbol is valid throughout the symbol.

. > > I use the averaging of the two half-ranges of the 3*m block tecnique > rather than the minimum square estimator as recommended.

That removes position uncertainty from the navigation equations, and allows all the "information" from the satellite measurements to be used to improve the time estimate. Errors in this position create errors in timing, with a magnitude scaled by the speed of light (one ns per foot, three ns per meter).

The attached picture is data from an optical plethysmograph sensor for heart rate, the estimated beat/beat interval, and the motion of the chest wall measured by a radar.

-VLBI: Among other observables, VLBI can be used to get very good estimations of the clock difference between stations.

It does this based on it’s internal estimates of lock quality. Unless you are already at maximum time constant and have a good internal estimate, changing the time constant has no immediate impact. On most GPSDO’s and with most OCXO’s under most conditions, the step out process takes days or weeks.

A bunch of multiples is assumed from the orbit, as there is at least the delay of the shortest distance, and then you can make a rough estimate by the sub-code phase of the birds, as they hook up like a set of clock-work gears.