New Power Estimation Methods for Highly Overloaded Synchronous CDMA Systems

TL;DR

This paper introduces new power estimation methods for highly overloaded CDMA systems, addressing the limitations of conventional techniques and demonstrating their effectiveness through simulations and signature set design.

Contribution

It proposes an optimal and a suboptimal power estimation method for overloaded CDMA systems, including iterative binary input estimation, and introduces a class of near-optimal signature sets.

Findings

Proposed methods work for systems with up to 2 users per chip.

Performance of the methods is validated through simulations.

A new class of signature sets is identified as optimal for power estimation.

Abstract

In CDMA systems, the received user powers vary due to moving distance of users. Thus, the CDMA receivers consist of two stages. The first stage is the power estimator and the second one is a Multi-User Detector (MUD). Conventional methods for estimating the user powers are suitable for underor fully-loaded cases (when the number of users is less than or equal to the spreading gain). These methods fail to work for overloaded CDMA systems because of high interference among the users. Since the bandwidth is becoming more and more valuable, it is worth considering overloaded CDMA systems. In this paper, an optimum user power estimation for over-loaded CDMA systems with Gaussian inputs is proposed. We also introduce a suboptimum method with lower complexity whose performance is very close to the optimum one. We shall show that the proposed methods work for highly over-loaded systems (up to m(m + 1) =2 users for a system with only m chips). The performance of the proposed methods is demonstrated by simulations. In addition, a class of signature sets is proposed that seems to be optimum from a power estimation point of view. Additionally, an iterative estimation for binary input CDMA systems is proposed which works more accurately than the optimal Gaussian input method.