Linear complexity multiuser detector using joint successive interference cancellation

We present a linear complexity multiuser detector that employs a joint successive interference cancellation (SIC) technique to achieve better performance than the standard SIC detector. The key feature of joint successive interference cancellation (JSIC) is that it considers two users jointly at each stage of successive decoding. For the two-user case, the JSIC detector gives the maximum likelihood (ML) solution. Simulations also show that for higher number of users, JSIC gives consistently better performance than the standard SIC detector, both in terms of error probability of a given user and the joint error probability for K users.