Interference Cancellation and DOA Estimation by Generalized Receiver Applying LMS and MUSIC Algorithms
Jingui Liu, Modar Safir Shbat, Vyacheslav P. Tuzlukov · 2013
Under implementation of the generalized receiver (GR) constructed based on the generalized approach to signal processing (GASP) in noise there is a need to cancel an interference to improve the GR performance. In this paper, we discuss an interference cancellation technique based on the non-blind beamfroming algorithm, namely, the least mean square (LMS) algorithm employed by GR. The direction of arrival (DOA) estimation is also used by LMS algorithm to provide a priori knowledge about the desired signal in GR. The simulation results demonstrate a superiority of GR performance in comparison with the Neyman-Pearson (NP) receiver. The generalized receiver (GR) is constructed based on the generalized approach to signal processing (GASP) in noise discussed in (1{5). The GR is a combination of Neyman-Pearson (NP) detector that is optimal for detection of signals with known parameters and the energy detector that is optimal for detection of signals with unknown parameters. With the purpose to improve the GR performance under noise and interference conditions, the least mean square (LMS) algorithm is proposed for interference cancellation (6). LMS beamforming algorithm is a non-blind beamforming algorithm that needs a reference signal to update the weight vectors of array antenna with the purpose to form a desired direction vector and generate nulls towards an interference direction (7). Employment of GR with LMS algorithm requires a prior knowledge about the desired signal. The direction of arrival (DOA) estimation is used to provide the GR with the required information about arrival angles of the received signals. For the past few decades, a wide variety of techniques have been proposed for the DOA estimation. The subspace algorithms such as the multiple sig- nal classiflcation (MUSIC) and estimation of signal parameter via rotational invariance technique (ESPRIT) algorithms are widely used owing to their high resolution (8). In this paper, the MUSIC algorithm for DOA is employed with LMS algorithm with the purpose to cancel interference. The simulation results demonstrate a good performance at the output of GR with LMS algorithm and MUSIC DOA estimation under the interference cancellation. The rest of this paper is organized as follows: Section 2 presents the simple GR ∞owchart and main functioning principles. Section 3 discusses the LMS beamformer employed by GR. Implementation of DOA estimation algorithm is introduced in Section 4. The simulation results are discussed in Section 5. The conclusion remarks are given in Section 6. 2. GR STRUCTURE GR ∞owchart is presented in Fig. 1. Here, AF is the additional fllter used to generate the reference noise and PF is the preliminary fllter that can be considered as a band pass fllter matched by bandwidth with the desired signal. The AF resonance frequency is detuned relative to the PF resonant frequency on such a value that both the signal and noise can be appeared at the PF output, whereas only the noise is appeared at the AF output. The detuning value must be more than 4 » 5 signal bandwidth. In this case, the correlation coe-cient between the processes forming at the PF and AF outputs is not more than 0.05. MSG is the model signal generator generating the reference signal or model signal a M. The stochastic process at the GR output takes the following form: