Power economy (PE) and impairment awareness (IA) have become crucial issues in optical networks. There is a tradeoff between PE and IA under the constraint of available network resources. Transparency is needed to achieve PE in optical networks, as this requires the avoidance of the regeneration of an optical signal. This leads to an accumulation of noise due to the physical layer impairments and ultimately results in the loss of quality of the signal. We have optimized the selective usage of traffic grooming, mixed regeneratoion, and all optical wavelength conversion (AOWC) using integer linear programming to increase the PE and IA simultaneously. The optimized solution together with heuristics results are given for 24 node USNet and 19 node RailTel network of India. The utilization of heuristics for the longest path of RailTel network is shown in OptSim hardware simulation with comparative performance using various techniques. An experimental demonstration of an AOWC using semiconductor optical amplifierbased four wave mixing has been given in Section VII. A feature of this setup is that the wavelength converter subsystem remains in a lower power mode, until it is actuated by the control program based on the quality factor (Q-factor) of the signal.Index Terms-All optical wavelength conversion, four wave mixing, impairment awareness, mixed regeneration, power economy, traffic grooming.