Text
Consistent Optical and Electrical Noise Figure
The optimum noise figure of an electrical amplifier is Fe = 1 and the optimum traditional noise figure of an optical amplifier is Fpnf = 2. This irresolvable conflict is due to the fact that Fe measurement requires electrical powers, proportional to squared amplitudes (voltages), while Fpnf measurement requires squares and variances of photocurrents, proportional to 4th powers of amplitudes (fields). In line with this an electrical receiver can receive I&Q parts of an electric carrier while a direct-detection photoreceiver can detect only power and not phase of an optical carrier. Optical amplifiers cause Gaussian field noise. Photodetection causes shot noise. Spontaneous-spontaneous beat noise in direct detection is taken into account by negative binomial or chi-squared photoelectron distributions, without errors of a Gaussian approximation. Coherent receivers linearly sense the Gaussian field noise. The sensitivity of an ideal coherent I&Q receiver is not degraded if it gets an ideal optical preamplifier, while the corresponding Fpnf = 2 suggests degradation. Coherent I&Q or heterodyne receivers have electrical output powers proportional to squared amplitudes (fields). This way one has the same metric in electrical and optical domain. One gets an optical noise figure Fo , IQ . For large amplifier gain it is Fpnf /2. In an ideal amplifier, Fo , IQ = 1. For true optical homodyne receivers and for direct detection receivers with Gaussian approximation it can be converted into Fpnf and vice versa. Phase-sensitive amplifiers are also covered. With Fe and the I&Q optical noise figure Fo , IQ a consistent unified noise figure is derived, valid and usable in electrical and optical domain.
Barcode | Tipe Koleksi | Nomor Panggil | Lokasi | Status | |
---|---|---|---|---|---|
art145718 | null | Artikel | Gdg9-Lt3 | Tersedia namun tidak untuk dipinjamkan - No Loan |
Tidak tersedia versi lain