Optimized Sequences for Nonlinearity Estimation and Self-Interference Cancellation

In wireless communication systems, nonlinearity is the root cause of many undesired effects. In order to compensate these effects, the nonlinearity must be estimated accurately. This article deals with nonlinearity estimation methods and especially focuses on sequences that are optimized for that purpose. Based on the Cramér-Rao bound of the nonlinearity estimator, we derive an objective function for sequence optimization. An algorithm is proposed to solve this optimization problem while satisfying constraints on the sequence format to produce sequences that are compatible to current wireless LAN standards. The obtained sequences are thoroughly analyzed and compared against sequences defined in IEEE 802.11ax. Their superiority confirms that dedicated sequences for nonlinearity estimation are valuable in future communication standards. Moreover, we provide an analysis of estimation error components to show that an optimal number of estimated polynomial coefficients can be found that minimizes estimation error. To further show the benefits of optimized sequences in the context of an application, a method for digital nonlinear self-interference cancellation suited for frequency-division duplex systems is proposed and evaluated. With that, we demonstrate that cancellation performance is significantly enhanced using optimized estimation sequences.