Research on inter-process communication based on multi-core MCUs

In the field of embedded systems, the challenges associated with the application of multi-core MCUs are primarily manifested in the process of inter-process communication. Although communication methods between cores based on operating systems have matured to a considerable extent, achieving a satisfactory balance between communication efficiency and capacity remains a difficult endeavor. Some inter-process communication algorithms based on operating systems still largely reside in the theoretical domain, with practical implementation being some distance away. In embedded systems without an operating system, shared memory serves as a module capable of direct interaction with the MCU, offering significant advantages in terms of communication efficiency and capacity. This article is grounded in shared memory and presents a design for distributed shared memory inter-process communication, using the iterative computation model of an induction motor. Experimental comparisons are conducted between this approach and the RPMessage inter-process communication method, with respect to communication speed compared to single-core computation. The experiments demonstrate that the distributed shared memory approach can significantly reduce computation time during data processing, as opposed to single-core computation. This research holds valuable implications for the application of inter-process communication in multi-core MCUs.