For the successful implementation of mobile communication networks in the spectral range of 30-300 GHz, it is necessary to solve several problems related to terminal subscriber devices. And one of the main problems is to be blocked by small line-of-sight transmission objects. Ways to solve this problem are densification of cellular infrastructure in the terahertz range with the help of mobile mobile communication nodes, as well as the use of simultaneous connection of subscriber terminals to two or more cells at the same time [1; 2].
In this work, an improved NOMA algorithm based on relaying and modulation called M-СО-NOMA is proposed to solve this problem. Unlike the known ones, it differs in that, firstly, before broadcasting into the radio air in the first time slot, terminal devices are clustered in the service area of the base station of the mobile network using the algorithm for estimating the fractal dimension and, secondly, during the formation of the general transport signal for terminal devices of subscribers, the base station modulates the signals of remote users on the quadrature, and the signals of nearby users on the real component of the QPSK constellation.
The proposed improved M-CO-NOMA algorithm eliminates the main drawbacks of NOMA by introducing orthogonality at the stage of signal preparation instead of including it in the spectrum. It is shown that the proposed improved algorithm is effective in terms of such indicators as the reduction of SER, computational complexity, and interference compared to the usual NOMA.
In addition, the integrated solution based on M-NOMA and СО-NOMA made it possible to obtain new innovative opportunities for ensuring data transmission when blocking direct transmission in the terahertz frequency range due to auxiliary signal relaying between subscribers. Calculation of the information efficiency of the developed integrated solution showed that for systems with NOMA and M-СО-NOMA, this indicator is equal to 4.50 e-2 and 6.00 e-2, respectively. Thus, the calculated information efficiency value of the M-CO-NOMA system is 1.5 times greater than that of NOMA.