Results from a Four-Element Adaptive Antenna Experiment

Experimental results from a four-element, linear, half-wavelength spacing, adaptive-array antenna under the control of the least mean square (LMS) algorithm are presented. The array is found to be capable of nulling a 70-MHz signal to -35 dB below a desired signal over a 5-MHz bandwidth. The antenna processing gain is constant over a desired signal-to-jammer signal power ratio range from -20 dB to 5 dB. A sharp reduction in processing gain is observed for angular separations between jammer and desired signal of less than 10°. Antenna patterns taken with weights set in 300 iterations of the LMS algorithm show that the one strong, one weak jammer combination has a longer weight convergence time and reduced processing gain compared with a two strong jammers combination. Contours of constant desired signal-to-jammer signal power ratio, after adaptive antenna processing, reveal a complex shape for communication between air and ground due to the finite angular resolution of the adaptive antenna.     

Some properties of retrodirective beam of multiple sidelobecanceler

The retrodirective or perturbation beam concept is often used to aid in explaining the nulling of interference sources. The classical explanation for cancellation of a single-zero-bandwidth source is that of a perturbation beam steered to and peaking at that interferer angle. The subtraction of this perturbation beam from the main beam causes the nulling. It is shown that, for discrete point main and auxiliary antennas, this explanation can be generalized to non-zero bandwidth sources. It is demonstrated that if the main and auxiliary antennas are spatial points, then the retrodirective beam is nonuniformly weighted, steered to the interference source, and peaks at the interferer angle. For the more realistic assumption of a nonpoint main antenna, the retrodirective beam does not peak at the interferer angle for any bandwidth     

Analysis of a Signal Processor for an Antenna Array

A signal processor that provides ratio-squared predetection combining, has been investigated for application in an adaptive antenna array. The analysis and experimental data presented here pertain to the signal processing gain when the antenna array is illuminated by a coherent signal source and a partially coherent noise source. For a noise source which is coherent, the processing gain depends on relative strength of the signal and noise, relative directions of arrival, and the usual "array factor." The array exhibits capturing effects much as in an FM receiver. The effective antenna pattern is a superposition of two beams of different magnitudes, one directed to the signal source and the other to the noise source. When the noise is partially coherent, the behavior of the signal processor is quite complex. Analytical prediction and experimental simulation measurement on a four-channel system indicate that the partially coherent noise may be regarded as the source of an incoherent noise component plus a coherent noise component with the magnitude of the latter determined by the coherence coefficient for the noise source.     

Constrained Least-Squares Synthesis of Coefficients for Arrays of Sensors and FIR Digital Filters

A general method is presented for synthesizing weighting coefficients for arrays of sensors or for finite-impulse-response (FIR) digital filters. The beam pattern of the array or the frequency response of the digital filter is a weighted, least mean-square (LMS) approximation to a desired function, subject to constrained pattern values at specified points. The method is applied to the problem of producing beam patterns of a line array for the rejection of interfering point sources. For this case, we show that adaptive array processing can be based on modification of the roots of the array polynomial.     

Adaptive processing at the subarray level

This paper describes a class of partially adaptive arrays with adaptive processing applied to the outputs of steered subarrays. The problem is to detect a signal or estimate its direction of arrival in the presence of jammers. The advantage of applying adaptive processing to subarrays is that it requires much less CPU time than the corresponding fully adaptive processing. The subarray processing equations for the two kinds of problem are described. In this paper, we compare partially adaptive processing performance with fully adaptive processing performance in the case of the following antenna and signal sources:

• •- array with regularly spaced sensors ;
• •- between 20 and 100 elements ;
• •- between 3 and 20 subarrays ;
• •- a single jammer ;
• •- desired signal from the antenna zenith.

We suggest a method for determining the optimal subarray configurations in this case. An example is given to show that the performance of an antenna with five subarrays is comparable to that of a fully adaptive thirty-element array for eliminating a single jammer with a target at the zenith.     

An Lms Adaptive Array Using a Pilot Signal

A least mean square (LMS) adaptive array requires a reference signal. When the desired signal contains a pilot signal, it may be used as the reference signal. In this paper the steady-state performance of an LMS adaptive array in which the pilot signal is used as the reference signal is examined. It is shown that the LMS adaptive array occasionally suppresses the desired signal. The loop gain, which is an important parameter, is also considered.     

Prediction of Adaptive Array Performance

A direct relationship between the conventional properties of an array and the array performance in an adaptive mode is given. Expressions are provided to obtain the output signal-to-interference-plus-noise ratio (SINR) of an adaptive array in terms of its conventinal pattern and the locations of the desired signal and jammers. These expressions permit one to evaluate the performance of an adaptive array without an exhaustive search for all possible scenarios and parametric values to ascertain that the required performance levels be met. In fact, one can predict the jammer locations for which the array will provide its best and worst performance by observing the conventional pattern. Several examples are provided to demonstrate the relationship between the conventional pattern and the adaptive array performance. The examples include both linear and planar arrays.     

The Performance of an LMS Adaptive Array with Frequency Hopped Signals

The performance of an LMS adaptive array with a frequency hopped, spread spectrum desired signal and a CW interference signal is examined. It is shown that frequency hopping has several effects on an adaptive array. It causes the array to modulate both the amplitude and the phase of the received signal. Also, it causes the array output SINR (signal-to-interference-plus-noise ratio) to vary with time and thus increases the bit error probability for the received signal. Typical curves of the desired signal modulation and the time-varying SINR at the array output are presented. It is shown how the array performance depends on hopping frequency, frequency jump size, interference frequency, signal arrival angles, and signal powers.     

Convergence Rate of the Extended SMI Algorithm for Narrowband Adaptive Arrays

An extended sample matrix inversion (SMI) algorithm can be used for minimizing the mse (mean-squared error) between the output of an N-element adaptive array and a desired reference signal. This algorithm is shown to yield mse within 3 dB of minimum (on the average) after (2N - 1) observations of the antenna element outputs.     

Adaptive Array Behavior with Periodic Envelope Modulated Interference

A method is presented for determining the effects of envelope modulated interference on a least mean square (LMS) adaptive array. The interference is assumed to have periodic envelope modulation with a bandwidth that is small compared with the carrier frequency. For such interference, the method allows one to calculate the periodic steady-state behavior of the array weights and the resulting array performance. As an example, we compute the effects of an ordinary amplitude modulated (AM) interference signal on the array. It is shown that such interference causes the array to modulate the desired signal envelope but not its phase. With a differential phase-shift-keyed (DPSK) desired signal, AM interference is found to have about the same effect on bit error probability as CW interference.     

The Effect of a Pulsed Interference Signal on an Adaptive Array

The performance of a least mean square (LMS) adaptive array in the presence of a pulsed interference signal is examined. It is shown that a pulsed interference signal has two effects. First, it causes the array to modulate the desired signal envelope (but not its phase). Second, it causes the array output signal-to-interferenceplus-noise ratio (SINR) to vary with time. The desired signal modulation is evaluated as a function of signal arrival angles, powers and interference pulse-repetition frequency (PRF) and pulsewidth. It is shown that the signal modulation is small except when the interference arrives close to the desired signal. To evaluate the effect of the time-varying SINR, it is assumed that the array is used in a differential phase-shift keyed (DPSK) communication system. It is shown that the SINR variation causes a noticeable but not disastrous increase in the bit error probability.     

Directivity-Bandwidth Limitations on Retrodirective Data Links

In a retrodirective array, frequencies differing from the pilot signal retrodirect at different angles from that of the pilot. For the retrodirected beamwidth to cover the reception point over the desired bandwidth, the product of directivity and fractional bandwidth squared cannot exceed a maximum value. This value is dependent only upon the angular retrodirection range, with smaller angular ranges allowing a higher product. For example, a ±30 degree angular range and 5 percent bandwidth allows 36-dB directivity.     

Instrumental variable adaptive array processing

An instrumental variable (IV) approach is presented for estimating the weights of an adaptive antenna array. Theoretical analysis of the IV method shows that the antenna gain weights are independent of finitely correlated noise, so that unbiased estimation of signal arrival angles is possible. Only matrix inversions are required to compute the weight estimates. In this sense, the IV method provides performance comparable with eigenvector techniques but with lower computational burden. Both minimal and overdetermined IV estimators are derived. The overdetermined estimators give the same theoretical array weights as minimal estimators, but yield more accurate weight estimates in real data situations. Simulation results are presented to compare these IV methods with one another and with conventional matrix inversion weight estimators. In these examples it is seen that IV methods are able to resolve closely spaced interference sources when conventional matrix inversion techniques cannot. It is also shown that overdetermined methods are capable of providing weight estimates with lower variances than those of minimal methods     

Performance of a Modified Applebaum Adaptive Array

The performance of a modified Applebaum adaptive array is studied. The new array is obtained by removing the desired signal component from the output signal fed back to the correlator of an Applebaum type adaptive array. Various signal scenarios, including single desired signals or multiple simultaneous desired signals are examined. The new array is less sensitive to beam pointing errors and does not cause power inversion of desired signals. In the case of multiple simultaneous desired signals, the new array does not degrade the SNRs of strong desired signals.     

Adaptive Array Behavior with Sinusoidal Envelope Modulated Interference

The behavior of a LMS (least mean square) adaptive array with modulated interference is described. An interference signal with sinusoidal, double-sideband, suppressed-carrier modulation is assumed. It is shown that such interference causes the array to modulate the desired signal envelope but not its phase. The amount of the desired signal modulation is determined as a function of signal arrival angles and powers and the modulation frequency of the interference. Such interference also causes the array output signal-to-interference-plus-noise ratio (SINR) to vary with time. However, it is shown that when the desired signal is a digital communication signal, the averaged bit error probability is essentially the same as for continuous wave (CW) interference.     

A Fast Beamforming Algorithm for Large Arrays

This beamforming algorithm is written specifically for array radars in which the number of array elements K is very large compared with the number of jammers L the radar is designed to suppress. It uses a set of M noise vectors to construct a basis for the jammer component of the antenna output vectors. The component of the quiescent weight vector orthogonal to each basis vector is calculated, renormalized to unit length, and identified as the adapted weight vector. This algorithm is effective in the suppression of many types of jammers. The number of noise samples M required in the construction of the adapted weight vector is approximately equal to L. In the special case of L narrowband noise jammers, for example, a choice of M = L usually reduces the receiver output jammer power to a few dBs above the white noise background. It is permissible to have M相似文献   

Angle measurement in the presence of mainbeam interference

The problem of reducing interference impinging on an antenna array when the sources lie in the main beam is addressed. Adaptive antenna arrays are incorporated to form adapted sum and difference beams in which the interference signals are suppressed. Monopulse error curves are then obtained, providing the necessary distortion correction curves across the entire mainbeam tracking angle region. New Cramer-Rao (C-R) bounds on the angle estimation error are derived with generalized assumptions on the signal amplitude and phase. The bounds previously derived by others are valid under different conditions. With these generalized assumptions on the signal characteristics, a Monte Carlo simulation is performed, based on the estimation procedure presented, to determine the angle estimation error. These errors are compared with the C-R bounds. Good performance is shown for sufficient S/N₀ and angular separation between the target and the interference sources     

确定相干机器多噪声源位置的方法研究

提出利用均匀线阵接收的数据构造一种矩阵进行奇异值（SVD）分解,以实现对相干噪声源的方位估计。利用这种方法,通过改变阵列中心位置便可确定机器多噪声源位置。计算机模拟和声学实验证实了这种方法的可行性。     

基于空域稀疏性的宽带DOA估计

利用宽带阵列接收信号的空域稀疏性,将宽带信号的波达方向(DOA)估计转化为一个稀疏信号重构的问题,提出了一种新的宽带信号DOA估计算法。该算法将宽带信号分解为多个子带信号,联合利用多个子带信号的空域稀疏性进行重构。它是对用于稀疏重构的标准的稀疏贝叶斯学习算法的推广,可适用于多冗余字典的信号模型。另外,通过对多快拍的阵列接收信号进行奇异值分解(SVD),提取信号子空间作为算法的输入数据,可以在有效减少运算复杂度的同时,提高对噪声的稳健性。与传统的宽带阵列DOA估计方法相比,该算法能够用于低信噪比、快拍有限和信源相关性较高的场合,同时算法的性能对信源个数的估计值不太敏感。仿真实验表明,该算法相对现有的基于子空间类的方法,具有更好的DOA估计性能。     

机载共形阵天线阵列信号处理的关键技术分析

机载共形阵是机载有源相控阵天线发展的下一个目标。共形阵天线的阵元具有不同的方向性,空间分辨率算法必须考虑来波的极化问题。本文研究了空间信号波达方向(Direction of Arrival,DOA)估计算法的现状,分析了一种适用于共形阵天线空域自适应波束形成的高速滤波算法。