Background and Objective: The Non-Orthogonal Multiple Access (NOMA) technology is a multi-access scheme that overcomes most of the disadvantages of its predecessor, the OMA technology. Specifically, NOMA technology supports massive connectivity of multiple users by employing the same non-orthogonal spectrum resource. In an uplink NOMA system with grant-free transmission mode, the Base Station (BS) is unaware of which users are active in the networks at a given time. Consequently, there is a need for mechanism to ensure successful recovery of users’ transmitted signals. This paper presents some new Multiuser Detector (MUD) schemes for uplink grant-free NOMA wireless communication networks with system’s model involving Multiple Measurement Vectors (MMV) rather than the Single Measurement Vector (SMV) that many previous works have considered. These MUDs include those that are based on differential Orthogonal Matching Pursuit (OMP), adaptive Simultaneous OMP (SOMP), compressive- multiple signal classification (MUSIC), and sequential compressive-MUSIC algorithms. Methods: The MUDs are employed in the detection of users’ signals in the uplink NOMA systems. Results and Discussion: Comparative performances of these MUDs with another one that is also based on the MMV system model, the SOMP-based MUD are presented for the scenarios when the system is under-loaded, fully loaded and over-loaded. Conclusion: The results suggest that the sequential compressive-MUSIC-based MUD, though shows weak performance at lower range of SNR, outperforms all the other MUDs including the SOMP-based MUD at higher SNR. Its performance is quite outstanding during the over-loaded scenarios, especially at higher SNR. However, its computational complexity is higher that the closely performing compressive-MUSIC-based MUD and SOMP-based MUD.
Read full abstract