Understanding Phase Modulation For 100G Transmission

by

There are a number of challenges to achieving 100Gb/s optical transmission rates using the technologies that have been deployed at 10Gb/s and even 40Gb/s.

Using traditional transmission technologies such as NRZ modulation on the line will result in very poor spectral efficiency and a 100G signal using NRZ would not fit into the standard 50GHz grid. In addition it would be almost impossible to achieve NRZ transmission due to a non-tolerance to chromatic dispersion.

Tolerance to Chromatic Dispersion at 2.5Gb/s, 10Gb/s, 40Gb/s

 

In order to increase spectral efficiency and reduce the effects within the fibre it has been necessary to use either phase modulation or phase and amplitude modulation in combination with polarisation multiplexing to reduce the symbol rate on the line to the extent that even large amounts of chromatic dispersion can be tolerated.

Using standard non return to zero (NRZ) modulation the amplitude on the line is set high when a 1 bit is transmitted and the line is set to low when a 0 is transmitted. Phase modulation instead transmits symbols with a changing phase instead of a changing amplitude. Take the example of binary phase shift keying (BPSK). In this modulation a 0 bit can be sent with the “standard” phase (with no phase offset), when a 1 bit is to be transmitted the phase is offset by 180°.  These phase changes are detected by the optical receiver to recover and decode the transmitted signal.

Amplitude based NRZ modulation and phase based BPSK modulation comparison

Both the NRZ and BPSK modulations have a symbol rate identical to the line bitrate. What this means is that for every single bit that is sent down the line the transmitter will be sending a single symbol (a high or low amplitude for a time period in the case of NRZ and a particular phase for a time period in the case of BPSK). So in the case of a 100G OTU4 interface running at a bitrate of 112Gb/s the symbol rate sent by the transmitter would also be 112Gbaud/s.

QPSK uses 4 phase states to represent 2 bits

In order to reduce the symbol rate most vendors use Quadrature Phase Shift Keying (QPSK) as the phase modulation. QPSK is similar in function to BPSK in that it uses specific phase states to represent bits on the line. However with BPSK there are only two phase states – 0° and 180°  –  whereas QPSK uses 4 distinct phase states – 45°, 135°, 225° and 315°. As there are 4 separate phase states every time period on the line can now be used to represent 2 bit – so QPSK as a 2 bit encoding method can already reduce the symbol rate on the line by 50% whilst maintaining the bitrate.

Other modulations exist such as DQPSK which will be examined on this site in the future, and other such as 16QAM allow increased numbers of bits per symbol to be transmitted.

 Phase modulation is not used in isolation for 100G signalling and instead is used in conjunction with polarisation multiplexing which will be discussed in a separate article in the near future.

Leave a Reply

PageLines