Optimal Frequency Hopping Patterns for FH-SCMA

Non-orthogonal multiple access (NOMA) techniques are an effective tool for increasing the user capacity in 5G networks. NOMA employs different interference cancellation strategies to improve the performance. This is necessary since the users occupy a same resource in any given time and the underlying user codes cannot achieve full orthogonality due to their short length compared to the number of users. In Sparse Code Multiple Access, Frequency Hopping (FH) can be used to randomize interference but the patterns need to be carefully designed in order to achieve effective randomization. In the paper, we first give the algebraic construction of optimal FH patterns when it comes to randomization, and then using simulations we establish the performance improvement when our patterns are compared with random pattern assignment.