1、外文原文：http:/ 外文文献原文 24 中文 4840字 外文文献原文 ： Rapid MIMO-OFDM Software DefinedRadio System Prototyping Amit Gupta, Antonio Forenza, and Robert W. Heath Jr. Wireless Networking and Communications Group Department of Electrical and Computer Engineering, The University of Texas at Austin 1
2、University Station C0803, Austin, TX 78712-0240 USA Phone: +1-512-232-2014, Fax: +1-512-471-6512 agupta, forenza, rheathece.utexas.edu AbstractMultiple input-multiple output (MIMO) is an attractivetechnology for future wireless systems. MIMO communication,enabled by the use of multiple transmit and
3、multiplereceive antennas, is known for its high spectral efficiency as well as its robustness against fading and interference. CombiningMIMO with orthogonal frequency division multiplexing (OFDM),it is possible to significantly reduce receiver complexity as OFDMgreatly simplifies equalization at the
4、 receiver. MIMO-OFDM iscurrently being considered for a number of developing wireless standards； consequently, the study of MIMO-OFDM in realisticenvironments is of great importance. This paper describes anapproach for prototyping a MIMO-OFDM system using a flexiblesoftware defined radio (SDR) syste
5、m architecture in conjunctionwith commercially available hardware. An emphasis on softwarepermits a focus on algorithm and system design issues rather thanimplementation and hardware configuration. The penalty of thisflexibility, however, is that the ease of use comes at the expenseof overall throug
6、hput. To illustrate the benefits of the proposed architecture, applications to MIMO-OFDM system prototypingand preliminary MIMO channel measurements are presented.A detailed description of the hardware is provided along withdownloadable software to reproduce the system. I. INTRODUCTION Multipl
7、e-input multiple-output (MIMO) wireless systems use multiple transmit and multiple receive antennas to increase capacity and provide robustness to fading 1. To obtain these benefits in broadband channels with extensive frequency selectivity,MIMO communication links require complex space time equaliz
8、ers. The complexity of MIMO systems can be reduced, however, through orthogonal frequency division multiplexing(OFDM). OFDM is an attractive digital modulation technique that permits greatly simplified equalization at the receiver. With OFDM, the modulated signal is effectively transmitted in parall
9、el over N orthogonal frequency tones.This converts a wideband frequency selective channel into N narrowband flat fading channels. Currently OFDM is used in many wireless digital communication systems, such as the IEEE 802.11a/g 2, 3 standards for wireless local area networks(WLANs). MIMO-OFDM techno
10、logy is in the process of being standardized by the IEEE Technical Group 802.11n4 and promises 外文文献原文 25 to be a strong candidate for fourth generation(4G) wireless communication systems 5. As the theory behind MIMO-OFDM communication continues to grow, it becomes increasingly important to dev
11、elop prototypes which can evaluate these theories in real world channel conditions. During the past few years, a number of MIMO-OFDM prototypes have been developed 612.These implementations make use of FPGAs or DSPs, which require a large amount of low level programming and a fixedpointimplementatio
12、n. This is the preferred solution when developing high-speed implementations； however, it hinders the flexibility of the platform as these systems are not easily reconfigurable. As a result when experimenting with many different space-time coding schemes or receiver designs, a more flexible solution
13、 may be preferred. In this paper we propose a MIMO-OFDM system architecture based on the software defined radio (SDR) paradigm. The advantage of this approach lies in the fact that the user is not required to have in depth hardware knowledge and may implement a number of different schemes by simply
14、reconfiguring the software. The platform uses National Instruments radiofrequency (RF) hardware in conjunction with the LabVIEW graphical programming language. With this architecture, it is possible to define and simulate a system in a high level programming language and then seamlessly apply that c
15、ode towards the hardware implementationthis greatly reduces the time involved in system prototyping. Compared with 612, our prototyping platform can easily be reduplicated as it consists of commercial-off-the-shelf hardware and publicly available software. A user who purchases the RF hardwarefrom Na
16、tional Instruments and downloads the available MIMO software toolkit along with the prototyping code developed by the authors (available at 13, 14) can realize the same rapid prototyping benefits which we discuss in this paper. The flexibility of the current implementation of the prototype is limite
17、d by some hardware constraints, such as the bandwidth of the PCI bus, which prevents fully real-time transmission over the wireless link, and software constraints like our lack of complete synchronization algorithms in the software, which causes us to use a wired synchronization channel. The spirit
18、of this contribution is to summarize our method and describe our first efforts towards the developmentof a complete MIMO-OFDM platform designed for systemvalidation and Fig. 1. A 2 2 MIMO-OFDM spatial multiplexing system channel measurements. More work needs tobe investig
19、ated to overcome the limitations and expand thecapabilities of our initial design. This paper is organized as follows. Section II exploresthe signal model for a MIMO-OFDM system and our specificMIMO-OFDM implementation. Section III discusses thespecifics of the hardware and software platform. Sectio
20、n IVshows preliminary results from our system implementation aswell as channel Spatial Multiplexing OFDM modulator OFDM modulator OFDM domod OFDM demod OFDM Equalizer Rec0m bine Bit Streams 外文文献原文 26 measurements in indoor environments. II. MIMO-OFDM IMPLEMENTATI
21、ON In this section we review the MIMO-OFDM signal modeland then describe our specific MIMO-OFDM system implementation. A. MIMO-OFDM Signal Model In a MIMO-OFDM system (see 8 and the referencestherein) MIMO space-time codes are combined with OFDMmodulation at the transmitter while complicated space-t
22、imefrequencyprocessing is employed at the receiver. For simplicityof explanation, we consider spatial multiplexing asillustrated in Fig. 1 though it will be apparent that othertransmission techniques can be implemented in the proposedarchitecture. In a MIMO-OFDM system with MT transmit antennasand M
23、R receive antennas, the sampled signal at the receiver(after the FFT and removing the cyclic prefix) of a spatialmultiplexing MIMO system for OFDM symbol period n andtone k can be expressed by the following equation (assumingperfect linearity, timing, and synchronization.) 1 , , , ,n k n k n k n kY
24、H S W (1) The equalization in MIMO-OFDM systems may be enabledthrough different procedures such as zero-forcing equalizer,minimum mean-squared error equalizer, V-BLAST successivecancelling equalizer, sphere decoder, and maximum likelihooddecoder (see 1 for an overview). In our p
25、rototype wecurrently implement the zero-forcing equalizer； the flexibilityof the proposed architecture though allows us to prototypemore sophisticated equalization strategies. B. System Implementation and Specifications The first implementation features spatial multiplexing with two transmit and two
26、 receive antennas, as illustrated in Fig. 1.Other MIMO schemes are already available in the LabVIEW MIMO Toolkit 14, and we are planning to use this to implement other space-frequency codes in the future. The specifications of the system are listed in Table I. In our MIMO-OFDM implementation, OFDM w
27、ith 64 tones is employed over a 16MHz bandwidth. The cyclic prefix is 16 samples long. This corresponds to an OFDM symbol duration of 5s, with a guard interval of 1s and a data portion of 4s. We transmit our OFDM symbols in 200ms data packets.This 200ms was determined by our hardware as memory const
28、raints at the receiver prevented longer acquisition periods. The system is equipped with an adjustable carrier frequency. We chose to run our system at 2.4GHz, which is the carrier frequency used for WLANs 2, 3. Various modulation schemes arepossible (BPSK, QPSK, 16-QAM, 64-QAM) along with optional
29、convolutional coding. Channel estimation is carried out by periodically transmitting an OFDM training symbol. The frequency at which training symbols are sent can be programmatically changed in the software and depends on the expected variation of the channel. The estimation at the receiver is enabl
30、ed by the pilot symbols, sent out over orthogonal tones across the transmit antennas. We then use a linear interpolation across the tones to estimate the channels full frequency response. Once wehave a channel estimate, the data is demodulated by a MIMO zero-forcing linear receiver. Due to space limitations, in this contribution we do not provide
