Interference Cancelation in Coherent CDMA Systems Using Parallel Iterative Algorithms

Least mean square-partial parallel interference cancelation (LMS-PPIC) is a partial interference cancelation using adaptive multistage structure in which the normalized least mean square (NLMS) adaptive algorithm is engaged to obtain the cancelation weights. The performance of the NLMS algorithm is mostly dependent to its step-size. A fixed and non-optimized step-size causes the propagation of error from one stage to the next one. When all user channels are balanced, the unit magnitude is the principal property of the cancelation weight elements. Based on this fact and using a set of NLMS algorithms with different step-sizes, the parallel LMS-PPIC (PLMS-PPIC) method is proposed. In each iteration of the algorithm, the parameter estimate of the NLMS algorithm is chosen to match the elements’ magnitudes of the cancelation weight estimate with unity. Simulation results are given to compare the performance of our method with the LMS-PPIC algorithm in three cases: balanced channel, unbalanced channel and time varying channel.

💡 Research Summary

The paper addresses the problem of multi‑user interference in coherent CDMA systems by improving the well‑known multi‑stage partial interference cancelation (PIC) technique that uses the normalized least‑mean‑square (NLMS) adaptive algorithm, commonly referred to as LMS‑PPIC. In the conventional LMS‑PPIC each stage updates a set of cancellation weights with a single, fixed step‑size μ. The choice of μ is critical: a small μ yields slow convergence, while a large μ can cause instability and amplify residual interference. Moreover, when all user channels are balanced the theoretical magnitude of each weight element should be unity, but a poorly chosen μ often leads to weight vectors whose elements deviate significantly from this property, resulting in error propagation from one stage to the next.

To overcome this limitation the authors propose a Parallel LMS‑PPIC (PLMS‑PPIC) scheme. The core idea is to run several NLMS filters in parallel, each with a different step‑size (μ1,…,μM). After each iteration all M candidate weight vectors are examined, and the one whose element magnitudes are closest to one is selected as the official estimate for that iteration. The selection criterion is defined as
(J_i = \sum_{n=0}^{N-1} (| \hat w_{k,i}