In this article, the recently developed Kaczmarz successive interference cancellation (KSIC) detector is generalized to the nonlinear case; specifically, a projected SIC detection scheme is developed using the projected Kaczmarz iterative method. This novel detector enjoys improved performance with minimal additional computational complexity. The new detector is first presented and then its convergence analysis is detailed. After that, the computational complexity analysis of this detector is conducted and compared to that of the KSIC detector. Finally, extensive numerical simulations are conducted in different scenarios. Simulation results show that the proposed projected Kaczmarz SIC detector can achieve lower BER and can support more users in the system compared to the KSIC detector.
INTRODUCTIONWireless fifth generation communication networks exhibit small and dense cell structures. 1,2 This small size network implementation yields various merits but at the same time suffers from various problems. The merits include high data rates and low latency whereas the drawbacks include increasing amounts of intercell, intracell, interantenna, and interuser interference. All these interferences need to be dealt with efficiently; otherwise, they will deteriorate the overall system performance.Interference cancellation detectors are essential tools for interference removal in different communication systems. Interference cancellation detectors are divided into two main categories: successive or serial interference cancellation (SIC) and parallel interference cancellation (PIC). 3,4 Recently, successive interference cancellation emerged as an important detector, especially after the rise of non-orthogonal multiple access (NOMA) as one of the key access schemes for future communication systems 5 where the SIC is adopted as the standard detector for this multiaccess scheme. 6,7 The successive interference cancellation approach is employed not only in NOMA but also in other areas where interference is a performance limiting factor. For example, it is used to suppress multi-access interference (MAI) due to the loss of orthogonality among spreading sequences in code division multiple access (CDMA). 8,9 The SIC is also used to suppress inter-antenna interference (IAI) and inter-carrier interference (ICI) in multiple-input multiple-output (MIMO) and in orthogonal frequency-division multiple-access (OFDMA), respectively. [10][11][12][13] The conventional linear SIC detection approach is regarded as a Gauss-Seidel iterative method where the columns of the system's matrix are used in a cyclic fashion to perform interference cancellation. 14 Recently, a new linear SIC detector employing the Kaczmarz iterative scheme (also known as the algebraic reconstruction technique [ART] in some literature) is devised. 15 This detector exploits the rows of the system's matrix instead of its columns to perform interference