• dongyuan1017
    了解作者
  • matlab
    开发工具
  • 7.5MB
    文件大小
  • zip
    文件格式
  • 0
    收藏次数
  • 10 积分
    下载积分
  • 30
    下载次数
  • 2020-03-10 17:02
    上传日期
数字预失真(dpd)原理matlab仿真,对学习研究比较有帮助
DPD_sim.zip
  • DPD_sim
  • fit_memory_poly_model.m
    1.2KB
  • README.pdf
    533.1KB
  • meas_pa.mat
    6.1MB
  • dpd_image.png
    18KB
  • DPD_analysis.m
    4.2KB
  • DPD.pdf
    810KB
  • dpd_adapt_verify.slx
    50.5KB
  • dpd_static_verify.slx
    40.4KB
  • ILA_CONFIGSS.slx
    17.5KB
  • license.txt
    1.3KB
  • pa_measure.slx
    44.4KB
  • rpem_dpd_coef.mat
    609B
内容介绍
<html xmlns="http://www.w3.org/1999/xhtml"> <head> <meta charset="utf-8"> <meta name="generator" content="pdf2htmlEX"> <meta http-equiv="X-UA-Compatible" content="IE=edge,chrome=1"> <link rel="stylesheet" href="https://static.pudn.com/base/css/base.min.css"> <link rel="stylesheet" href="https://static.pudn.com/base/css/fancy.min.css"> <link rel="stylesheet" href="https://static.pudn.com/prod/directory_preview_static/625d70b292dc900e6277b328/raw.css"> <script src="https://static.pudn.com/base/js/compatibility.min.js"></script> <script src="https://static.pudn.com/base/js/pdf2htmlEX.min.js"></script> <script> try{ pdf2htmlEX.defaultViewer = new pdf2htmlEX.Viewer({}); }catch(e){} </script> <title></title> </head> <body> <div id="sidebar" style="display: none"> <div id="outline"> </div> </div> <div id="pf1" class="pf w0 h0" data-page-no="1"><div class="pc pc1 w0 h0"><img class="bi x0 y0 w1 h1" alt="" src="https://static.pudn.com/prod/directory_preview_static/625d70b292dc900e6277b328/bg1.jpg"><div class="t m0 x1 h2 y1 ff1 fs0 fc0 sc0 ls0 ws0">Adaptive DPD Desi<span class="_ _0"></span>gn </div><div class="t m0 x1 h2 y2 ff1 fs0 fc0 sc0 ls0 ws0">By Kerry Schutz<span class="_ _0"></span> and Dick Benson<span class="_ _0"></span>, MathWorks </div><div class="t m0 x1 h3 y3 ff2 fs0 fc0 sc0 ls0 ws0">Digital pre-distortion (DPD<span class="_ _0"></span>) <span class="ls1">is</span> <span class="ls2">a </span>b<span class="_ _0"></span>aseband signal proces<span class="_ _0"></span>sing technique th<span class="_ _0"></span>at corrects for i<span class="_ _0"></span>mpairments inher<span class="_ _0"></span>ent <span class="ls3">to</span> <span class="ls4">RF </span></div><div class="t m0 x1 h3 y4 ff2 fs0 fc0 sc0 ls0 ws0">power amplifiers (PA<span class="_ _0"></span>s). These impairment<span class="_ _0"></span>s cause ou<span class="_ _0"></span>t-<span class="ls5">of</span>-band emissions or <span class="_ _0"></span>spectral regro<span class="_ _0"></span>wth and<span class="_ _0"></span> <span class="ls1">in</span>-band </div><div class="t m0 x1 h3 y5 ff2 fs0 fc0 sc0 ls0 ws0">distortion which correlat<span class="_ _0"></span>es with a<span class="_ _0"></span>n increased bit-error<span class="_ _0"></span>-rate (BER)<span class="ls6">. </span>Wideban<span class="_ _0"></span>d signals with <span class="_ _0"></span>a high peak<span class="_ _0"></span>-<span class="ls3">to</span>-average </div><div class="t m0 x1 h3 y6 ff2 fs0 fc0 sc0 ls0 ws0">ratio, as is characteristic of L<span class="_ _0"></span>TE/4G transmitt<span class="_ _0"></span>ers, ar<span class="_ _0"></span>e particulary susceptibl<span class="_ _0"></span>e to these unw<span class="_ _0"></span>anted effects. </div><div class="t m0 x1 h3 y7 ff2 fs0 fc0 sc0 ls0 ws0">In this paper, w<span class="ls7">e <span class="_ _0"></span></span>illustrate a w<span class="_ _0"></span>orkflow for mod<span class="_ _0"></span>eling and simu<span class="_ _0"></span>lating PAs an<span class="_ _0"></span>d DPDs. The m<span class="_ _0"></span>odels shown in th<span class="_ _0"></span>is </div><div class="t m0 x1 h3 y8 ff2 fs0 fc0 sc0 ls0 ws0">article are based on<span class="_ _0"></span> two techn<span class="_ _0"></span>ical papers, [1] and [2]<span class="ls6">. </span>W<span class="_ _0"></span>e start <span class="_ _0"></span>from PA measurements. F<span class="_ _0"></span>rom measure<span class="_ _0"></span>ments, <span class="ls8">we </span></div><div class="t m0 x1 h3 y9 ff2 fs0 fc0 sc0 ls0 ws0">derive a static DPD de<span class="_ _0"></span>sign based on<span class="_ _0"></span> a memory polynomial. <span class="_ _0"></span>Such a<span class="_ _0"></span> polynomial <span class="_ _0"></span>corrects for both <span class="_ _0"></span>the non-linearities </div><div class="t m0 x1 h3 ya ff2 fs0 fc0 sc0 ls0 ws0">and memory effects in<span class="_ _0"></span> the PA<span class="ls6">. </span>For <span class="_ _0"></span>simulation purposes, we<span class="_ _0"></span> construct a <span class="_ _0"></span>system-level <span class="_ _0"></span>model <span class="ls3">to </span>evaluate the DPD<span class="ff3">&#8217;<span class="_ _0"></span>s </span></div><div class="t m0 x1 h3 yb ff2 fs0 fc0 sc0 ls0 ws0">effectiveness. Becau<span class="_ _0"></span>se any PAs characteri<span class="_ _0"></span>stics will vary over<span class="_ _0"></span> time and <span class="_ _0"></span>operating cond<span class="_ _0"></span>itions, we extend<span class="_ _0"></span> the static </div><div class="t m0 x1 h3 yc ff2 fs0 fc0 sc0 ls0 ws0">DPD design <span class="ls3">to an</span> ad<span class="_ _0"></span>aptive one<span class="ls6">. </span>We evaluat<span class="_ _0"></span>e two adaptive <span class="_ _0"></span>DPD designs, o<span class="_ _0"></span>ne based on the (Lea<span class="_ _0"></span>st-Mean Squ<span class="_ _0"></span>are) </div><div class="t m0 x1 h3 yd ff2 fs0 fc0 sc0 ls0 ws0">LMS algorithm and <span class="_ _0"></span>a second using th<span class="_ _0"></span>e Recurive Predictor Er<span class="_ _0"></span>ror Method (RPEM) algorith<span class="_ _0"></span>m.<span class="_ _0"></span> </div><div class="t m0 x1 h3 ye ff2 fs0 fc0 sc0 ls0 ws0">The examples used in th<span class="_ _0"></span>is article are av<span class="_ _0"></span>ailable for download<span class="_ _0"></span>. </div><div class="t m0 x1 h2 yf ff1 fs0 fc0 sc0 ls0 ws0">The Workflow </div><div class="t m0 x1 h3 y10 ff3 fs0 fc0 sc0 ls0 ws0">The goal of this work is a s<span class="_ _0"></span>imulation model which a<span class="_ _0"></span>ccurately models the PA&#8217;<span class="_ _0"></span>s impairments<span class="_ _0"></span> and an adap<span class="_ _0"></span>tive DPD </div><div class="t m0 x1 h3 y11 ff2 fs0 fc0 sc0 ls0 ws0">design which mitigates tho<span class="_ _0"></span>se impairments. <span class="_ _0"></span><span class="ls9">We</span> divide th<span class="_ _0"></span>e modeling effort<span class="_ _0"></span> into four p<span class="_ _0"></span>arts. </div><div class="t m0 x2 h4 y12 ff2 fs0 fc0 sc0 lsa ws0">1.<span class="ff4 ls0"> <span class="_ _1"> </span><span class="ff2">Model and simulate t<span class="_ _0"></span>he PA. </span></span></div><div class="t m0 x2 h4 y13 ff2 fs0 fc0 sc0 lsa ws0">2.<span class="ff4 ls0"> <span class="_ _1"> </span><span class="ff2">Derive DPD coefficients.<span class="_ _0"></span> </span></span></div><div class="t m0 x2 h4 y14 ff2 fs0 fc0 sc0 lsa ws0">3.<span class="ff4 ls0"> <span class="_ _1"> </span><span class="ff2">Evaluate static DPD d<span class="_ _0"></span>esign. </span></span></div><div class="t m0 x2 h4 y15 ff2 fs0 fc0 sc0 lsa ws0">4.<span class="ff4 ls0"> <span class="_ _1"> </span><span class="ff2">Extend static DPD design <span class="_ _0"></span>to adaptive one.<span class="_ _0"></span> </span></span></div><div class="t m0 x3 h4 y16 ff2 fs0 fc0 sc0 ls2 ws0">a.<span class="ff4 ls0"> <span class="_ _1"> </span><span class="ff2">E<span class="_ _0"></span>valuate least mean square (<span class="_ _0"></span>LMS)-<span class="_ _0"></span>variant </span></span></div><div class="t m0 x3 h4 y17 ff2 fs0 fc0 sc0 lsb ws0">b.<span class="ff4 ls0"> <span class="_ _2"> </span><span class="ff2">Evaluate recursive p<span class="_ _0"></span>redictor error method (<span class="_ _0"></span>RPEM)-variant<span class="_ _0"></span> </span></span></div><div class="t m0 x1 h2 y18 ff1 fs0 fc0 sc0 ls0 ws0">State Your Assump<span class="_ _0"></span>tions </div><div class="t m0 x1 h3 y19 ff2 fs0 fc0 sc0 ls0 ws0">Models are never perfect repr<span class="_ _0"></span>esenta<span class="_ _0"></span>tions of reality. Th<span class="_ _0"></span>e key is to simplify reality wh<span class="_ _0"></span>ile maintainin<span class="_ _0"></span>g sufficient </div><div class="t m0 x1 h3 y1a ff2 fs0 fc0 sc0 ls0 ws0">relevance. We made the <span class="_ _0"></span>following simplification<span class="_ _0"></span>s in o<span class="_ _0"></span>rder to a<span class="_ _0"></span>ccelerate the simulations<span class="_ _0"></span>. </div><div class="t m0 x2 h4 y1b ff2 fs0 fc0 sc0 lsa ws0">1.<span class="ff4 ls0"> <span class="_ _1"> </span></span><span class="ls9">We<span class="ls0"> model the PA as a<span class="_ _0"></span> discrete time syst<span class="_ _0"></span>em. In practice, the P<span class="_ _0"></span>A is an analog circuit. <span class="_ _0"></span> </span></span></div><div class="t m0 x2 h4 y1c ff2 fs0 fc0 sc0 lsa ws0">2.<span class="ff4 ls0"> <span class="_ _1"> </span><span class="ff2">We model the PA signal as<span class="_ _0"></span> baseband<span class="_ _0"></span> complex. In real<span class="_ _0"></span>ity, the PA signal <span class="lsc">is</span> <span class="ls2">a </span>real pas<span class="_ _0"></span>sband signal (Fi<span class="_ _0"></span>gure 8)<span class="ls6">. </span> </span></span></div><div class="t m0 x2 h4 y1d ff2 fs0 fc0 sc0 lsa ws0">3.<span class="ff4 ls0"> <span class="_ _1"> </span><span class="ff2">We use double-preci<span class="_ _0"></span>sion data typ<span class="_ _0"></span>es and math. <span class="_ _0"></span>In reality, in<span class="_ _0"></span>teger data types and math<span class="_ _0"></span> would b<span class="_ _0"></span>e used<span class="ls6">. </span> </span></span></div><div class="t m0 x2 h4 y1e ff2 fs0 fc0 sc0 lsa ws0">4.<span class="ff4 ls0"> <span class="_ _1"> </span><span class="ff2">We do not includ<span class="_ _0"></span>e quan<span class="_ _0"></span>tization effects due to th<span class="_ _0"></span>e ADC and DAC (Fi<span class="_ _0"></span>gure 8). </span></span></div></div><div class="pi" data-data='{"ctm":[1.568627,0.000000,0.000000,1.568627,0.000000,0.000000]}'></div></div> </body> </html>
评论
  • nibonibo 2021-12-30 16:04:41
    比较有用,点赞
相关推荐