Algorithm of 32-bit Data Transmission Among Microcontrollers Through an 8-bit Port

Abstract

This paper proposes an algorithm for 32-bit data transmission among microcontrollers through one 8-bit port. This method was motivated by a need to overcome limitations of microcontroller I/O as well as to fulfill the requirement of data transmission which is more than 10 bits. In this paper, the use of an 8-bit port has been optimized for 32-bit data transmission using unsigned long integer, long integer, and float types. Thirty-two bit data is extracted into
binary number, then sent through a series of 8-bit ports by transmitter microcontroller. At receiver microcontroller, the binary data received through 8-bit port is reconverted into 32 bits with the same data type. The algorithm has been implemented and tested using C language in ATMega32A microcontroller. Experiments have been done using two microcontrollers as well as four microcontrollers in the parallel, tree, and series connections. Based on the experiments, it is known that the data transmitted can be accurately received without data loss. Maximum transmission times among two microcontrollers for unsigned long integer, long integer, and float are 630 μs, 1,880 μs, and 7,830 μs, respectively. Maximum transmission times using four microcontrollers in parallel connection are the same as those using two microcontrollers, while in series connection are 1,930 μs for unsigned long integer, 5,640 μs for long integer, and 23,540 μs for float. The maximum transmission times of tree connection is close to those of the parallel connection. These results prove that the algorithm works well.