Demystifying 10log10(2^30) in signal energy calculations.

This change removes a magic number used in NR UE rx power calculations. Received power is usually calculated using RMS^2 - each sample is squared first before taking the mean. With the samples being SQ15 (meaning 1 sign bit and 15 fractional bits) to normalize the samples would require shifting each right by 15 bits, however this can be done once signal energy was calculated and this is what the 10log10(2^30) is doing. I've tried to add some meaningful comment to the #define, hopefully this saves some headache for other people down the road.