Massively Parallel, Nonlinear Processing for 6G: Potential Gains and Further Research Challenges

In order to support the plethora of the new, pending, transformative use-cases, 6G systems will need to be capable of providing not only high-throughput but also ultra-massive device connectivity. In this direction, 6G systems will need to fully exploit the available spatial, frequency and time resources of the wireless channel. Moving from linear to nonlinear processing has the potential to substantially improve the capabilities of such 6G systems. However, the processing complexity and latency requirements of traditional nonlinear processing become impractical when a large number of concurrently transmitted information streams is targeted. This work discusses how by efficiently massively parallelizing nonlinear processing, we can enable practical 6G systems with substantially improved capabilities compared to current systems. Examples of such gains are demonstrated and further research challenges are discussed towards unlocking the full potential of future massively parallel, nonlinear processing. In addition, how massively parallel, nonlinear processing can redefine the way we design future physical layer architectures is also examined.