Space-time adaptive decision feedback neural receivers with data selection for high-data-rate users in DS-CDMA...

Area of Science:

• Electrical Engineering
• Signal Processing
• Telecommunications

Background:

• Direct-sequence code-division multiple-access (DS-CDMA) systems face challenges with multiaccess interference (MAI) and signal distortion.
• Antenna arrays are employed to enhance spatial selectivity in wireless communication.

Purpose of the Study:

• To propose a novel space-time adaptive decision feedback (DF) receiver for DS-CDMA systems.
• To jointly perform equalization and interference suppression using recurrent neural networks (RNNs).

Main Methods:

• The proposed receiver utilizes dynamically driven RNNs in the feedforward section for equalization and MAI suppression.
• A finite impulse response (FIR) linear filter is used in the feedback section for interference cancellation.
• A data selective gradient algorithm within the set-membership (SM) framework is developed for RNN coefficient estimation.

Main Results:

• The proposed neural receiver structure effectively estimates its parameters.
• Simulation results demonstrate significant performance improvements compared to existing schemes.
• The space-time adaptive receiver achieves superior equalization and interference suppression capabilities.

Conclusions:

• The proposed RNN-based DF receiver offers a powerful solution for enhancing DS-CDMA system performance.
• The novel coefficient estimation algorithm contributes to the receiver's adaptive capabilities.
• This approach represents a significant advancement in interference management for spread spectrum systems.