Using multiple modulated carriers that are superpositioned at different frequencies is called OFDM, so why did they coin a new term, "super channel"? I have several clarifying questions:

1. Why must the carriers be frequency-multiplexed if each carrier gets its own physical channel? This seems overly complicated. Perhaps the author made a mistake and each physical channel has multiple subcarriers at different frequencies?

2. What is the bandwidth (complex sampling rate, presuming quadrature architecture) per subcarrier?

3. Are they planning on using digital pre-distortion to compensate for degradation in the fiber optic channel? If so, how much oversampling do they think they need?

4. What CPU or special DSP ASICs are needed to support the encoding/decoding, etc?

5. Why no 1024-QAM? This is achievable with today's technology. Maybe the bandwidth and coding rate inject too much EVM?