how can we explain the negative frequency physically?
Physicists leverage to their advantage many seemingly paradoxical concepts, as negative temperature or energy. Do not know if they use negative frequency for special treatment of problems. For engineers, the generalization of frequency values onto a negative semi-axis and even into the complex plane is an important concept in both the practical and the theoretical aspects of their work. Some examples:
The carrier frequency is either 2.4GHz or something else, but not both. Also, it makes no sense to speak of the negative frequency in the context of EM radiation frequency.
Your representation of a carrier wave by "two tones" is not valid (item 1) and so your consideration of power splitting between waves. Still, notice: when multiple coherent monochromatic electromagnetic waves arrive at the receiver antenna, you cannot separate the total power into the contributions of individual waves: it is the participating electromagnetic fields that sum up (superposition principle), while the energy and power are quadratic functions of the field strength.
Because your "two tones" are not defined, we cannot discuss their "rotation". To mention rotation: in a circularly polarized electromagnetic wave the field vectors rotate. This rotation can be either clockwise or counter clockwise w.r.t. the wave vector direction (propagation direction).
Phasors. Let you have an AC current source, a real "hardware implemented" device. Let its output be I(t) = I0·sin(ω·t). The source is loaded by a high-quality, practically lossless, inductor L. Apply an AC Ohm's law to calculate the voltage across the inductor. Start calculation, and you immediately realize, that the AC source produces not one, but two frequencies, +|ω| and -|ω|:
$$
V_L = I_0{1\over{2j}}(exp(-jωt)·Z_L(-ω) - exp(jωt)·Z_L(ω)) = \\
I_0{1\over{2j}}(exp(-jωt)·(jωL) - exp(jωt)·(-jωL)) = \\
-I_0·cos(ωt)·ω·L
$$
You cannot do without the negative frequency concept when applying AC Ohm's law -- it is possible only at the expense of undue complication in calculations and analysis.
Double Balanced Mixers. Although the presence of sum and difference frequencies in the mixer output spectrum can be explained with angle sumof electromagnetic signals and difference identities, the negative frequency concept is needed to understand how a baseband frequency spectrum is transferred to aelectromagnetic radiation spectrum transmitted via a high-frequency communication channel (radio)are different things. The analysis and calculations of double sideband and single sideband modulation schemes operate with negative frequencies inTo understand why the basebandsignal spectrum.
Eb/No. Citing https://en.wikipedia.org/wiki/Eb/N0:
Sometimes, the noise power is denoted by N0/2 when negative frequencies and complex-valued equivalent baseband signals are considered rather than passband signals.
It is not a nonessential remark, made in passing. The contains both positive and negative frequency concept streamlines the noise analysis of radio communication channels.
Alsovalues, natural (non-computer) language habits may defer acceptance ofconcentrate on the "negative frequency" concept by EE students. Compare definitions of ordinary frequency (f) vs. angular frequency (ω)discipline. For ordinary frequency -- the number of cycles per second -- even the non-integer values are the result of generalization (because the fractional value requires a quantization within a single cycle) or computation (if we define the frequency as the inverse of a periodLearn phasors, f=1/T). There is no such problem with angular frequency measured in radians per second: it is real-valued by definitionBode plots, and can be negative as well. (Recommendation for textbook authors: avoid negative values for ordinary frequencies measured in Hz.)
The next jump into generalization is to get used to the complex (angular) frequencyconstellation diagrams, where also some "insights" can be sought and might be found outmodulation/demodulation etc.
