Wing Back Sweep
The idea of sweeping back the wing as an arrowhead came in the 1930s (Busemann), as a consequence of the fact that only the nirmal velocity component matters on an infinite swept wing (principle of independence). Thus swept wings (or infinite yawed wings) can always be designed to operate behind the Mach cone. Wave drag is completely avoided only by infinite swept wing.
Since the velocity component parallel to the leading edge does not contribute to the
change in aerodynamic behavior (Fig. 1), the critical Mach number is reduced to the
value M'=M/cos 
Figure 1: swep tback wing
The use of sweep-back reduces the critical Mach number to the value M' = M/cos(sweep). This technical solution not only delays the transonic drag rise, but it also reduces the rate at which the drag increases in the transonic regime.
While the back sweep is ideal at transonic and low supersonic speeds, it does have some drawbacks at low speeds: high induced drag and loss in lift coefficient. To fly at both regimes wings of variable sweep may be required.
Figure 2: CD vs Mach number for different sweeps
A similar effect can be achieved with a forward sweep, although there appear some stability problems with consequent difficulty in maneuvering the aircraft.
While the back sweep is ideal at transonic and low supersonic speeds, it does have some drawbacks at low speeds: high induced drag and loss in lift coefficient. To fly at both regimes wings of variable sweep may be required (these wings, though, have their own problems).
Related Material