1401阅读
5回复

求：哪位英语高手帮我翻译一下，好难啊！！！

离线save1987

UID ：60345

注册：2010-05-25
登录：2010-06-15
发帖：41
等级：仿真新人

0楼发表于: 2010-06-05 20:41:01

Abstract =1o_:VOG
The experimental system PALES is described: a multi-functional array antenna in the X band with broadband beamforming by a Rotman lens. Studies that have been conducted about possible broadband antenna elements have implied the use of flared-notch antennas (Vivaldi antennas), which are specified in the paper. A short description of the Rotman lens design used is given. The receiver branch of the PALES system, the applied signal model, and different broadband array-processing methods, which can be applied to the received signals of a multibeam system, are discussed. To separate broadband signals with a small angular separation, the ICA (Independent Component Analysis) method shows superior performance compared to ML (maximum likelihood) estimation, based on superresolution by the STCM (Steered Covariance Matrix) method.
Keywords: Vivaldi antennas; flared notch antennas; Rotman lens; lens antennas; antenna arrays; array signal processing; maximum likelihood estimation
1. Introduction {nH. _
Spatial restrictions, combined with an increasing number of antenna systems on the same platform, the weight of the antenna systems and their costs, as well as the mutual interference of the different antenna systems, suggest the use of a shared multi-functional antenna ("shared aperture") for different tasks. An antenna system is preferred that can be used simultaneously for radar, electronic support measures (ESM), and for other tasks, such as communication or active protection (e.g., jamming, HPM, etc.). The desired shared antenna has to fulfill the requirements of the different tasks and to meet the scopes of future missions.
In future radars, passive modes are preferred, to avoid detection. If an active radar mode cannot be avoided, the detection should be made difficult by frequency hopping over a broad frequency region, or by using special pulse modulations, e.g., pseudo-noise signals. Digital components and synthesizers allow for the desired high agility of the transmitted signals in time and frequency. On the other hand, radar imaging methods, such as SAR (synthetic aperture radar) or ISAR (inverse synthetic aperture radar) are required for reconnaissance and non-cooperative target identification (NOI), which imply a broad instantaneous band-width. To guarantee effective energy management during the radar mode, and to simultaneously conduct different radar tasks, such as search, confirmation, and tracking of multiple targets, an electronic-scanning-array antenna (ESA) is required. The real options of an electronic scanning array, if in an active or passive mode, can not be exploited until array signal processing is applied. Interference suppression by adaptive beamforming (ABF), angular super-resolution of target or jammer signals, adaptive beamforming on transmission to avoid the detection by hostile electronic support measure receivers, and the detection of slowly moving targets in clutter by spatial temporal adaptive processing (STAP), can only be realized by array signal processing.
When conducting electronic support measure tasks, signal acquisition requires a broad instantaneous bandwidth of the system. To analyze the received signals of an imaging radar, an instantaneous bandwidth of at least 500-2000 MHz is required. An over-all bandwidth of multiple octaves is desired in the electronic support measures mode. Exact signal analysis of the received signals requires a high dynamic range. High sampling rates and bit depths are desired for electronic support measures (a "digital receiver"). The high agility in time and frequency of future radars expands the observation space and reduces the number of samples. Thus, the integration loss has to be compensated for -at least partially - by antenna gain. Because the angles of arrival (AOA) of possible target signals are not known in advance, an electronic support measures receiver should provide a full field of view, or at least a major angular sector should be simultaneously observed. In electronic support measures applications, the estimation by super resolution of the angles of arrival of received signals with small angular separation, the estimation of the parameters of multiple simultaneously received signals, as well as signal sorting ("blind signal separation") for classification and identification, and the activation of necessary countermeasures, can only be realized by using a multi-channel system.
To achieve the above requirements, a multi-channel system, applying array processing, is a basic demand. The system has to have a broad instantaneous bandwidth (imaging radar, electronic support measures). The high dynamic range of the receiver's channels, the numerical load having a high sampling rate (>1 GHz), forced by the broad instantaneous bandwidth, and the resulting costs, imply a reduction of the number of receiver channels. The number of processed receiver channels should be chosen to be as small as possible. The whole interesting angular section has to be observed simultaneously (a full field of view in electronic suppor measures).
All of the above-noted postulates can be realized by a multibeam system. A multibeam system guarantees the necessary antenna gain to compensate for the integration loss of the missing time samples. A multibeamn system operates like a spatial bandpass filter. Only a small number of beams is necessary to obtain nearly optimum results in jammer suppression, spatial superresolution, and signal sorting.
In general, Butler or Blas matrices can only be realized as narrowband devices. While a change of the look direction can be realized by true time delays in the case of antenna arrays, in the broadband case, phase shifters (e.g., a shift by 7r12) cannot be built from constant time delays. For a given runtime of an electromagnetic wave, the phase changes with frequency.

Electromagnetic lenses, such as Rotman lenses or Luneberg lenses, are possible candidates for the desired multibeamn system. In both types of electromagnetic lenses, beamforming relies on true time delays, i.e., both lens types have the desired broadband properties. Considering Luneberg lenses, by construction the angular distance of the look directions of neighboring beams is given by the aperture of the beam-port antennas mounted on the surface of the sphere. In contrast to Luneberg lenses, Rotman lenses can be constructed with transmission and reception beams that have a small angular separation. A distribution of the power on many antenna elements with their own transmitter/receiver (TR) modules can not be realized with Luneberg lenses, in contrast to Rotman lenses. Whether Luneberg lenses and the beam-port antennas used can be used for high-power transmission (2-4 kW) has to be checked.

At the Army Research Laboratory (ARL), a multifunctional antenna system was built [1] that is similar to the experimental system PALES (Phased Array with an Electromagnetic Lens and Array Signal Processing"), described in this paper. Array processing was not investigated at ARE，while in the PALES project, array signal processing is one of the main research topics.

In this paper, the experimental system PALES is described, a multi-functional. array antenna operating in the X band, with broadband beamnforming by a Rotman lens.。The studies conducted about possible broadband antenna elements have implied the use of flared-notch antennas. The Vivaldi antennas used are described in the second section of the paper. A short description of the Rotman lens design used is given in the third section. The receiver branch of the PALES system and the applied signal model is sketched in the fourth section. In the fifth section, broadband array processing that can be applied to the received signals of a multibeam system is discussed. Because of the good results obtained by the Independent Component Analysis (ICA) method, Independent Component Analysis signal processing ..

未注册仅能浏览部分内容，查看全部内容及附件请先登录或注册