Adaptive Digital Pre-distorter for Higher-Order Modulation Nonlinearities and Data Jitter

Recently, there has been a lot of interest in the adoption of variable coding & modulation (VCM); and adaptive coding & modulation (ACM) for the communication systems of near-Earth satellites.  These spectral efficient techniques increase data rates by utilizing the available link margins and pairing various channel coding schemes with higher-order modulations, but have a number of implementation and performance challenges.  We propose to develop an adaptive digital pre-distorter modulator that would overcome these challenges.  Recently, there has been a lot of interest in the adoption of variable coding & modulation (VCM); and adaptive coding & modulation (ACM) for the communication systems of near-Earth satellites.  These spectral efficient techniques increase data rates by utilizing the available link margins and pairing various channel coding schemes with higher-order modulations, but have a number of implementation and performance challenges.  We propose to develop an adaptive digital pre-distorter that would overcome these challenges.

The Consultative Committee on Space Data Systems (CCSDS) has adopted two VCM/ACM systems, Digital Video Broadcasting (DVB-S2) and Serially Concatenated Convolutional Coding (SCCC), but the fundamental concept of VCM/ACM can be applied to existing sets of CCSDS protocols, such as Low Density Parity Check (LDPC) codes with higher order modulations.  These systems try to operate with all available link margins while maintaining a constrained bandwidth, by increasing data rates through the use of various channel coding and modulation combinations.

Fundamentally, higher-order modulations are heavily sensitive to channel distortions, producing large losses in link performance and power efficiency.  The distortions are typically handled by pre-distortion of the waveforms prior to the amplifier, with one of the most practical approaches being the application of a static inversion of the measured amplifier characteristics.  However, these pre-distortion approaches have limited capability due in part to improper characterization of the amplifier response or the limitation in the pre-distortion algorithm.  Typically, manufacturers’ input amplitude modulation to output amplitude modulation (AM/AM) and input amplitude modulation to output phase modulation (AM/PM) curves are measured with a single tone radio frequency carrier that is varied in power.  However, the data modulated transmitted signal is dynamic, with a spectrum of side tones across the bandwidth.  The non-linearity will produce intermodulation products that are not considered in these measurements, thus designing a pre-distorter with these measurements is inaccurate and will produce a large penalty in the error performance of the link.

Alternatively, the input drive level of the amplifier can be backed off to minimize nonlinearities.  However, as the order of the modulation increases, so does the amount of backed off power.  Having a transmitter operating with a large amount of required input back off translates into a tremendous amount of direct current (DC) power required on the spacecraft to support these modulations.

The objective of this proposal is to develop an adaptive digital pre-distorter that can overcome the challenge of requiring excessively high DC power for higher-order modulations that are required for high-data rates links, e.g. VCM/ACM links.  In addition, the fundamental approach can also be applied to mitigate phase noise by reducing symbol-to-symbol jitter.  The advantages of digital pre-distortion are that the dynamic and memory behavior of the AM/AM can be measured in real-time while the adaptive algorithm can mitigate these effects to a designed precision based on a learning algorithm.  Therefore we can effectively eliminate any distortions created by the amplifier.