Per Pelin

Detection of sources using bootstrap techniques

Source detection in array processing can be viewed as a test for equality of eigenvalues. Such a test is proposed, based on a multiple test procedure that considers all pairwise comparisons between eigenvalues. Using the bootstrap to estimate the null distributions of the test statistics results in a procedure with minimal assumptions on the nature of the signal. Simulations show that the proposed test is superior to information theoretic criteria such as the MDL, which are based on Gaussian signals and large sample sizes. Performance in most cases exceeds the more powerful sphericity test.

Decoupled blind symbol estimation using an antenna array

The problem of separating superimposed digitally modulated signals using an array of antennas is considered. The proposed method exploits the finite alphabet structure to demodulate one signal at the time, resulting in a computationally efficient solution. The resulting signal estimates are shown to be exact in the noise-free case. In noisy scenarios, the performance is comparable with that of the iterative least squares approach, which demodulates all signals simultaneously at a higher computational cost.

A fast minimization technique for subspace fitting with arbitrary array manifolds

A new technique for minimization of a multidimensional subspace fitting criterion is proposed. An approximation is made to the exact criterion function, giving an explicit least squares solution for location parameters. Ideas from alternating projections are employed for initialization. The new technique gives global convergence with a high probability, has a low computational cost, is applicable to general array manifolds, and provides simple solutions for many problems with intricate parameterization.

Performance of decoupled direction finding based on blind signal separation

A number of methods for blindly separating superimposed digitally modulated signals arriving at an antenna array have been proposed. These techniques are efficient at the up-link (mobile to base) in a mobile communication system. However, for solving the base-to-mobile beamforming problem it may be necessary to also estimate the directions-of-arrival (DOAs) of the various signal paths. We present an optimal decoupled DOA estimation procedure based on information from the blind separation algorithm. Its performance is evaluated in the presence of spatially correlated noise and array modeling errors. The proposed technique has computational advantages as compared to traditional DOA estimation, because the different signal waveforms are treated in a separated fashion. Yet, the decoupled approach is shown to be substantially less sensitive to modeling errors and interference.

Joint symbol detection and channel parameter estimation in asynchronous DS-CDMA systems

A method for jointly estimating the time delay and complex gain parameters, as well as detecting the transmitted symbols in an asynchronous multipath DS-CDMA system, is presented. A short training sequence is used to obtain a coarse estimate of the channel parameters, which is consequently used to detect the symbols. By exploiting structure in the digitally modulated signals, the method iterates to (i) improve the estimate of the channel parameters and (ii) reduce the probability of incorrect detection. The methods efficacy is demonstrated by numerical simulations.

Detection of sources in array processing using the bootstrap

A hypothesis testing methodology for determining the number of narrowband sources impinging on an array is presented. Using multiple hypothesis tests the multiplicity of the smallest ordered eigenvalues of the sample correlation matrix and hence the number of sources, is determined. The finite sample null distributions of the test statistics are estimated using bootstrap resampling. By removing the assumption of Gaussianity and large sample size that the traditional MDL approach is based on, we are able to gain improvements in the small sample case or when there are deviations from Gaussianity.

Decoupled direction finding: detection

Antenna arrays are likely to be an important feature of future mobile communication systems. With an antenna array, mobile users can be separated by a spatial filtering procedure allowing several users on the same carrier frequency. The uplink part (mobile to base) not only can, but is better solved without using any spatial knowledge in terms of direction of arrival (DOA). However, DOA estimation remains an important issue in the overall system, both for downlink beamforming, as well as channel allocation. Previous results have shown that DOA estimation is best performed in a post-detection manner, i.e., using the estimated symbol sequence for DOA estimation. In this way, the estimation problem can be decoupled to treat individual users separately. To estimate the number of propagation paths from a specific user, a detection scheme is derived based on the DOA estimation criterion function.

Time delay estimation for multipath CDMA-systems based on a fast minimization technique for subspace fitting

A low complexity algorithm for parameter estimation in a block fading multipath DS-CDMA system is presented. The main contribution is a novel technique for minimizing a subspace fitting criterion which is obtained as a large sample approximation of the maximum likelihood (ML) estimator. The minimization procedure is based on an approximation of the exact criterion function, allowing a direct analytic solution. For the acquisition phase, an initialization procedure similar to alternating projections is employed which exhibits remarkable global convergence properties. The efficacy of the proposed method is demonstrated by means of numerical simulations.

A decoupled WLS approach to DS-CDMA multiuser detection

A new algorithm for multiuser detection in a DS-CDMA system is presented. Based on an algorithm originally developed in the array-processing context, it minimizes a weighted least-squares criterion, and detects one signal at a time. A performance analysis in terms of the bit-error rate is carried out, and demonstrates that the proposed method is asymptotically identical to the LMMSE-estimator, but without assuming or requiring any knowledge of the signal amplitudes. Only the desired users signature sequence and timing is required. Both a batch and a recursive formulation of the algorithm is described and simulation results are included in addition to the analysis, to highlight the comparison to other methods.

Using the bootstrap for robust detection in array signal processing

A critical problem in many signal processing applications is the determination of the correct model order for example, the number of multipath components in a received communication signal. One approach to detect the model order is to use the distribution of the criterion function of the estimator applied to find interesting parameters. Unfortunately, the nominal distribution of such a criterion function relies heavily on a correct model of the observed signal. In practice, with modeling errors present, the distribution is unknown. For robust detection based on the criterion function of a certain class of estimators, a two-step procedure is proposed. First, an alternative representation of the residuals is found using a predictor. Second, using bootstrap resampling, a parameter is estimated from the new residuals. This parameter transforms the criterion function to pivotal form. Numerical experiments show robustness to a range of possible modeling errors. An example from real measured array data is included.