Welcome to Scribd, the world's digital library. Read, publish, and share books and documents. See more
Download
Standard view
Full view
of .
Look up keyword
Like this
1Activity
0 of .
Results for:
No results containing your search query
P. 1
Mimo Chapter

Mimo Chapter

Ratings: (0)|Views: 1|Likes:

More info:

Published by: Ghionoiu Marius Catalin on Apr 10, 2013
Copyright:Attribution Non-commercial

Availability:

Read on Scribd mobile: iPhone, iPad and Android.
download as PDF, TXT or read online from Scribd
See more
See less

05/15/2014

pdf

text

original

 
FPGA in Wireless CommunicationsApplications
Kiarash Amiri
1
, Melissa Duarte
1
, Joseph R. Cavallaro
1
, Chris Dick
2
, RaghuRao
2
, and Ashutosh Sabharwal
1
1
Rice University
{
kiaa,mduarte,cavallar,ashu
}
@rice.edu
2
Xilinx Inc.
{
chris.dick,raghu.rao
}
@xilinx.com
1 Introduction
In the past decade we have witnessed explosive growth in the wireless commu-nications industry with over 4 billion subscribers worldwide. While first andsecond generation systems focussed on voice communications, third genera-tion networks (3GPP and 3GPP2) embraced CDMA (code division multipleaccess) and had a strong focus on enabling wireless data services. As we reflecton the rollout of 3G services, the reality is that first generation 3G systemsdid not entirely fulfill the promise of high-speed transmission, and the ratessupported in practice were much lower than that claimed in the standards.Enhanced 3G systems were subsequently deployed to address the deficiencies.However, the data rate capabilities and network architecture of these systemsare insufficient to address the insatiable consumer and business sector demandfor the nomadic delivery of media- and data-centric services to an increasinglyrich set of mobile platforms.With these considerations in hand, the move to 4G technologies like 3GLTE (long term evolution) [1] and WiMAX [2] is proceeding at an extremelyrapid rate. The goal of next generation systems is to provide high-data rate,low-latency, high reliability (minimize outages and connection drops) employ-ing packet-optimized radio access technology supporting flexible bandwidthallocation. Additional key objectives are to drive down the cost of infras-tructure equipment and consumer terminals and to employ a more efficientmodulation scheme than the CDMA technology used in 3G systems in orderto make more optimal use of precious communication bandwidth.To meet all of these requirements a significant re-structuring of both thephysical layer (PHY) and network architecture is required. Increased spectralefficiency is delivered in all 4G cellular broadband sysems through the use of OFDM (orthogonal frequency division multiplexing) as the preferred modu-lation scheme. The technology center-piece for delivering improved data ratesand communication link robustness is more aggressive use of the spatial di-mension of a communication system through multiple-input multiple-output
 
2 Authors Suppressed Due to Excessive Length
(MIMO) [3] techniques. 4G wireless systems will employ MIMO processingto equip next generation systems with spatial multiplexing (improved datarates) and diversity (improved reliability) capabilities. MIMO systems (Fig1) can increase the data rate and provide multiplexing gain by transmittingdifferent data on different antennas [4], also known as spatial multiplexing.MIMO systems can also improve the reliability and error performance in thereceiver through diversity gain, i.e. providing the receiver with multiple copiesof the transmitted signal. One practical way to realize such diversity gains is tobeamform a message across multiple transmit antennas [2, 5]. The beamform-ing procedure utilizes limited information about the channel state information;this channel information is usually provided through a limited feedback linkfrom the receiver to the transmitter.The implementation of 4G infrastructure equipment has several challenges.The first is related to the compute requirements of MIMO-OFDM systems.The more sophisticated MIMO-OFDM advanced receiver architectures havecompute requirements that are several orders of magnitude greater than thatof 3G CDMA systems. A second requirement is related to bridging the gapbetween
legacy 
standards (e.g. 3GPP) and next generation 4G protocols andsystems. In many cases 4G infrastructure equipment will need to be multi-mode in nature and support not only the MIMO-OFDM PHY of WiMAXand 3G LTE systems, but also the W-CDMA (wideband CDMA) PHY of 3GPP-based networks. To first order, the silicon technologies deployed in nextgeneration systems will not only need to supply enormous compute capability(many hundreds of giga-operations/second), but they will also require sig-nificant flexibility in order to support multi-mode basestation infrastructureequipment.Field programmable gate arrays (FPGAs) with their inherently parallelstructure, are increasingly the technology of choice for addressing the com-pute and flexibility requirements of next generation systems. One of the fun-damental areas in many academic and industrial research organizations is inthe area of hardware realization of advanced MIMO receivers.
Transmitter Receiver
12
...
MT12
...
MR
Fig. 1.
Multiple antenna (MIMO) system.
 
FPGA in Wireless Communications Applications 3
In this chapter, we discuss the architectural challenges associated withspatial multiplexing and diversity gain schemes and introduce FPGA archi-tectures and report experimental results for the FPGA realization of thesesystems. We introduce a flexible architecture and implementation of a spatialmultiplexing MIMO detector, Flex-sphere, and its FPGA implementation. Wealso present a hardware architecture for beamforming in a WiMAX system asa way to enhance the diversity and performance in the next-generation wire-less systems. Finally, we introduce the Rice University Wireless Open AccessResearch Platform (WARP), an FPGA-centric scalable and programmable re-search platform, and how it is used for studying the effects of MIMO systemsin real-world scenarios.
2 Spatial Multiplexing MIMO Systems
One of the challenges of MIMO systems is detecting the transmitted vector,which often times, translates into finding a minimum distances, and requiresconsiderable resources. For instance, most of the MIMO detection requirelarge sizes of searching and sorting in hardware. In the next two sections, wepresent the spatial multiplexing MIMO systems as well as a reduced complex-ity MIMO detector, Flex-Sphere.We assume a MIMO OFDM system with
transmit antennas and
R
receive antennas. We denote the
subcarriers by the subscript
k
=1
,...,K 
throughout this chapter. The input-output model is captured by
˜y
k
=
˜H
k
˜s
k
+
˜n
k
(1)where
˜H
k
is the complex-valued
R
×
channel matrix,
˜s
k
= [˜
s
1
,k
,
˜
s
2
,k
,...,
˜
s
,k
]
is the
-dimensional transmitted vector, where each ˜
s
j,k
,j
= 1
,...,M 
, ischosen from a complex-valued constellation
Ω 
j
of the order
p
j,k
=
|
Ω 
j,k
|
,
˜n
k
isthe circularly symmetric complex additive white Gaussian noise vector of size
R
and
˜y
k
=
y
1
,k
,
˜
y
2
,k
,...,
˜
y
R
,k
]
is the
R
-element received vector. Notethat we do not restrict all the parallel
streams to use the same modulationorder; rather, each stream, which corresponds to one of the antennas of oneof the users, may be using either the 4
,
16 or 64-QAM modulation. Also, notethat even though we focus on one transmitter with multiple antennas, theresults in this section can be extended to multiple users such that the sum of the number of antennas of all of them equal
.We assume that the complete channel information, i.e. the channel matrixcoefficients for each subcarrier, is known in the receiver. Moreover, since thedetection procedure for each subcarrier is performed independent of othersubcarriers, we will drop the
k
subscript unless it is needed. Therefore, theoriginal MIMO model can be simplified to
˜y
=
˜s
+
˜n
.The preceding MIMO equation can be further decomposed into real-valuednumbers as follows [5]:

You're Reading a Free Preview

Download
scribd
/*********** DO NOT ALTER ANYTHING BELOW THIS LINE ! ************/ var s_code=s.t();if(s_code)document.write(s_code)//-->