This paper proposes a feed-forward neural network predictor to adaptively estimate and suppress the narrowband interference (NBI) in the Direct Sequence-Code Division Multiple Access (DS-CDMA) signal. The iterative code-aided estimation is used to further improve the system performance. Simulation results reveal that the proposed algorithm outperforms conventional linear prediction filtering and recurrent neural networks (RNN) based NBI rejection methods, in different interference models.

Iterative Narrowband Interference Suppression for DS-CDMA Systems Using Feed-Forward Neural Network

Adaptive modem and interference suppression based on deep learning

In recent years, modern systems are highly reliant on Global Navigation Satellite Systems (GNSS) as reliable positioning, navigation, and timing services have become crucial in many safety-critical...

A Comprehensive Survey on GNSS Interferences and the Application of Neural Networks for Anti-jamming

This paper proposes a narrowband interference (NBI) suppression algorithm for Direct Sequence-Code Division Multiple Access systems. The NBI is considered from heterogeneous networks, and predicted based on its cyclostationary characteristic using a nonlinear feed-forward neural network predictor which eliminates the nonlinearity of the spread spectrum (SS) signal in the NBI prediction. To further improve the suppression performance, this paper exploits the structure of the spreading code, and proposes an iterative code-aided algorithm to jointly estimate the NBI and the SS signal. Simulation results reveal that the proposed algorithm largely outperforms the conventional linear prediction filtering and linear-conjugate linear polyperiodically time-varying filtering methods in both the signal to interference plus noise ratio improvement and the bit error rates, when it operates in NBI-contaminated environments.

https://link.springer.com/content/pdf/10.1007%2Fs11277-011-0495-0.pdf

Cyclostationarity-Based Narrowband Interference Suppression for Heterogeneous Networks Using Neural Network

To suppress the interference in the direct sequence spread spectrum (DSSS) system, a transform domain message data adaptive identify (TISI) algorithm is proposed in this paper, based on two improved algorithms: Power distributing predominance wavelet packets transform (PDP-WPT) and extended BP neural network (EBPNN). Firstly, PDP-WPT is presented to track the interference signal effectively, which improves the convergence rate of TISI. Secondly, the message data can be identified in transform domain by adaptability EBPNN, which has simple structure and enhanced numerical robustness. Simulation results show that TISI can improve the capability of interference suppression by 32% compared with widely used conventional algorithms.

Interference Suppression for Satellite Navigation Based on Neural Network and Wavelet Packets Transform

A new adaptive neural network predictor to eradicate the narrowband interference in the spread spectrum system is proposed in this paper. The effectively robust Kalman filter (RKF) algorithm is adopted to adjust the synaptic weights in the nonlinear recurrent architecture and thereby estimate the narrowband interference. The main characteristics of the proposed RKF-based canceller are its rapid convergence rate and precise prediction. Simulation results reveal that the RNNP based on RKF algorithm has large improvement on the interference suppression capability compared with conventional LMS, ACM and RTRL-based canceller in CWI and ARI environments, respectively.

Hong Zhao

Papers

Narrowband Interference Suppression Using RKF-Based Recurrent Neural Network in Spread Spectrum System