Non-arbitrary frame of reference for acceleration?

The reason we can choose a frame of reference arbitrarily is that a physical system is not dependent on an absolute position, orientation or time.
According to Noether's Theorem, the invariance of a system under a change of position is equivalent to the momentum conservation law.
In the same way, invariance under a change in orientation is the angular momentum conservation law, and time invariance is the law of
conservation of energy.
In Newtonian Mechanics, momentum conservation is expressed as the action-reaction law; If for every force (change of momentum) there's an equal and opposite
reaction force, momentum is conserved.
For fictitious forces however, like the force that acts when the frame of reference itself is accelerating, there's no corresponding
reaction force; Hence there seems to be no conservation law in that case.
Does this not point in the direction of a non-arbitrary frame of reference for acceleration?
I've looked around on google, but I've always heard that this is an unsolved mystery in physics.

Any ideas?

via International Skeptics Forum http://ift.tt/1NcIcuZ