Indonesi an  Journa of El ect r ic al Eng inee r ing  an d  Comp ut er  Scie nce   Vo l.   23 ,  No.   3 Septem ber   2021 , pp.  1493 ~ 1500   IS S N: 25 02 - 4752, DO I: 10 .11 591/ijeecs . v 23 .i 3 . pp 1493 - 1500          1493       Journ al h om e page http: // ij eecs.i aesc or e.c om   Eva lu ation   of   dif ferent   q uantiz ation   res olu tion   lev els   on   th e   BER   pe rfo rmance   of   m assive   MI MO   s yst ems   un der   diffe rent   operatin g   s cena rios       Hayder   Kh aleel   AL - Q aysi,   Tahreer   M ahmo od ,   Kh alid   Awa ad   Hum ood   Depa rtment   of   E le c troni c   Eng ineeri ng,   Coll ege   of   Engi n ee ring ,   U nive rsit y   of   Di y a la ,   Ira q       Art ic le   In f o     ABSTR A CT   Art ic le   hist or y:   Re cei ved   M ay   1 ,   2021   Re vised   Ju l   7 ,   2021   Accepte d   J ul   14 ,   2021       The   m assive   MIM O   sy stem   is   one   of   the   m a in   technologies   in   the   fi fth   gene ra ti on   (5G)   of   te l ec om m unic ation   s y st ems ,   al so   rec ogn ized   as   a   high l y   la rge - sc ale   s y ste m .   Constant l y   in   m assive   MIMO   sy st ems ,   the   base   stat io n   (BS)   is   provide d   with   a   la rg e   num ber   of   ant ennas ,   and   thi s   la rg e   num ber   of   ant enn as   ne ed   high - quan ti z a ti on   r esolut ion   le v el s   an al o g - to - digi t al   conve rt ers   (AD Cs).   In   thi s   situ a ti on,   the r e   wil l   be   m ore   power   consum pti on   and   har dware   c osts.   Thi s   pape r   pre sents   the   sim ula ti on   per for m anc e   of   a   suggested   m et hod   to   inve stigate   and   anal y z e   the   eff e ct s   of   diffe ren t   quant i za t ion   re soluti on   le v el s   of   AD C s   on   the   bit   e rror   rat e   (BER )   per form anc e   of   m assive   MIM O   sy st em   under   diffe ren t   oper a ti n g   sce nar ios   using   MA TL AB   software .   Th e   result s   show   that   th e   SNR   ex ceeds   12   dB   ac coun ts   for   onl y   0 . 001%   of   BE R   signal s   when   the   num ber   of   a nte nnas   60   with   low   quant i za t ion   a   2   bit s’   le ve ls   AD Cs,   ap proximate l y .   Bu t   when   the   ant enn a   num ber   rises   to   300,   th e   SNR   exc ee ds   12   dB   ac count s   for   al m ost   0. 01%   of   BER   tra nsm it te d   signa ls.   Com par ably   with   the   BER   per form ance   of   high   qu ant i z at ion ,   4   bit s - qu ant i za t ion   r esolut ion   l evels   AD Cs   with   th e   sam e   diffe ren t   a nte nnas   hav e   a   s li ght   degr adatio n.   The r efo re ,   th e   num ber   of   ant enn as   is   a   v er y   important   inf lu enc e   fa ct or .     Ke yw or ds:   BER   perf or m a nce   Ma ssive   MIM O   Qu a ntiza ti on   reso l ution   le ve ls   AD Cs   RF   chai n   This   is   an   open   acc ess   arti cl e   un der   the   CC   BY - SA   l ic ense .     Corres pond in g   Aut h or :   Hayde r   Kh al ee l   AL - Qaysi   Dep a rtm ent   of   Ele ct ro nic   En gi neer in g   Coll ege   of   En gi neer in g,   U nive rsity   of   Diya la   Ba quba h,   Diya la   Prov i nce,   Ir a q   Em a il :   hay.k ha .82@ uodiya la . edu.iq       1.   INTROD U CTION   Ma ssive   m ult i - in pu t   m ulti - ou tp ut   (MIMO )   is   a   hig hly   la rg e - scal e   physi cal   layer   syst e m .   The   Ma ssive   MIM O   is   co ns ide re d   one   of   the   r el ia ble   te chn ol og ie s   in   fifth - gen e rati on   ( 5G)   te le com m un ic at ion   syst e m s   and   be yond   to   im pr ov e   sp ect ral   ef fici ency,   e nerg y   eff ic ie ncy,   t hro ughput,   sec ur it y   an d   rob ust ness,   hard war e   c omplexit y,   cost,   a nd   si gnal   proc essing   [1 ] - [ 9 ].   The   m ai n   idea   in   the   m assive   MIM O   syst em   is   to   us e   a   la r ge   nu m ber   of   tra nsm itti ng   a nten na s   ( )   in   t he   ba se   sta ti on   (B S)   t han   the   num ber   of   recei ving   anten nas   ( ),   w hich   re pr ese nt   the   us ers ,   to   pro vid e   sim ult aneous   ser vice   within   the   c overa ge   a rea   with   relat ively   si m p le   sign al s   proc essing   [6 ] ,   [ 7 ] ,   [ 10 ].   H owe ve r,   the   us e   of   a   la rg e   num ber   of   ante nnas   in   the   BS   of   m assive   MIM O   syst e m   increases   the   hardw a re   c os t   an d   com plexity ,   a nd   powe r   co nsum ption   of   the   ci rcu it   com po ne nts   in   the   rad i o   fr e quency   (RF )   c ha in   that   c onsis ts   of   sig nal   m i xer s ,   powe r   a m pl ifie rs,   ded i cat ed   filt ers,   a nd   hi gh - quantiz at ion   reso l ution   le ve l   analo g - to - digi ta l   con ve rters   (ADCs)   [ 11 ] ,   [ 12 ].   Co ns e qu e ntly ,   the   increa sed   hard war e   c os t s   an d   po wer   consum ption s   represe nt   a   substanti al   obsta cl e   in   the   pr a ct ic al   i m ple m entat io n   of   the   m assive   MIM O   sys te m s   [ 13 ].   Evaluation Warning : The document was created with Spire.PDF for Python.
                          IS S N :   2502 - 4752   Ind on esi a J  E le c Eng &  Co m Sci,   Vo l.   23 , N o.   3 Se ptem ber  2 02 1 14 93   -   15 00   1494   The   powe r   c onsu m ed   by   high - qua ntiza ti on   r esolutio n   le vel   AD Cs   re pr ese nts   the   esse ntial   par t   of   the   total   power   c onsu m ed   by   the   oth er   ci rcu it   c om po ne nts   of   the   RF   chai n   [ 14 ] ,   [ 1 5 ].   In   paral le l,   the   util izati on   of   high - quanti zat ion   res ol ution   le ve l   A DC s   co ntribute s   s ign ific a ntly   to   increasin g   t he   hard war e   c ost s   and   com plexiti es   and   reduci ng   qu a ntiza ti on   error s   in   the   m assive   MIM O   syst e m s   [1 6 ] - [ 1 8 ].   The   pow e r   consum ption   in   the   A DCs   de creases   e xpone ntial ly   with   quantiz at ion   re sol ution   le vels   ( quanti zat io n   bit s)   an d   li near ly   with   s a m pling   f requ encies   (sam pling   rates)   [1 9 ].   Partic ularly ,   AD Cs   represe nt   the   m ai n   par ts   of   receiver s   in   the   m assive   MIM O   syst em s   to   c onve rt   the   analo g   si gn al s   into   dig it al   sig nals   th rou gh   s a m pling   and   qua ntiza ti on   processes   a nd   pr e par i ng   the m   for   s ub se que nt   di gital   sign a l   processi ng   op erati on s   [ 20 ].   In   t he   li te rature,   m any   inv est igati on s   w ork   cond ucted   to   s tud y   the   e ff ect   of   dif fer e nt   quantiz at io n   reso l ution   le ve ls   of   AD Cs   on   t he   perf orm ance   of   m assive   M IMO   syst em s   un de r   diff e re nt   operati ng   scenari os ,   s uc h   as:   usi ng   a   diff e re nt   num ber   of   a nten nas,   var i ou s   m od ul at ion   schem es,   an d   sig nal   sy m bo ls   detect or s .   In   [1 4 ],   the   a utho r s   in vestigat ed   the   ef fect   of   3 - to - 12   bits - qua ntiza ti on   res olu ti on   le vels   A DCs   on   the   upli nk   e nergy   ef fici e ncy   ( EE)   in   the   m assive   MIM O   syst e m ,   with   tim e   div isi on   dupl exin g   ( TD D)   m od e   and   a   diff e rent   nu m ber   of   a nten nas   at   the   BS   us ing   li ne ar   zer o - forcin g   (ZF)   a nd   ze r o - forcin g   su c c essive   interfe ren ce   ca ncell at ion   ( ZF SI C)   si gn al   sy m bo ls   detect ors,   with   both   pe rf e ct   a nd   im per fect   c hannel   sta te   inf or m at ion   (C SI ) .   T heir   res ul ts   showe d   that   the   c on s um ed   powe r   is   si gn i f ic ant   in   t he   m assive   MIM O   s yste m   wh e n   us in g   l ow   ( 3bit s) - quan ti zat ion   res olu t ion   le vels   ADC s   due   to   the   i ncr ease   in   qua ntiza ti on   e rrors ,   w hich   needs   to   c om pen sat e   with   a   la rg e   num ber   of   ante nn a s   at   the   BS.   In   ad diti on ,   t he   us e   of   the   pe rf ect   CS I - bas e d   li near   ZFS IC   sign al   sym bo ls   detect or   is   abl e   to   i m pr ov e   t he   EE   of   the   s yst e m   m or e   th an   the   pe rf ect   CSI - base d   li near   ZF   sign al   sym bo ls   detect or.   T he   aut hors   i n   [1 6 ]   in vestigat e d   the   r ole   of   usi ng   low - qua ntiza ti on   reso l ution   le ve ls   AD Cs   wit h   the   li near   m ini m u m   m ea n   s quare d   e rror   (LMM SE )   and   no nlinear   la tt ic e   reducti on - base d   su cce ssive   in te rf ere nce   ca nc el la ti on   (LR - S I C)   sign al   sym bo ls   detect ors   in   i m pr ov i ng   t he   E E   in   the   m assive   MIM O   syst em s   unde r   dif fer e nt   operati ng   sc enar i os .   T heir   resu lt s   s how ed   that   the   bit   er r or   rate   (BER)   pe r form ance   usi ng   the   LR - SI C   si gn al   sy m bo ls   detect or s - ba sed   recei ver s   is   bette r   than   t hat   of   LMM SE   sign al   sym bo ls   detect or s - base d   receive r s   an d   that   the   EE   of   the   syst em   dep e nds   on   the   nu m ber   of   ant enn a s,   cel l   siz e,   and   sign al   pr ocessin g   eff ic ie ncy   em plo ye d   at   the   BS.   In   a dd it io n,   the   a uthors   s uggeste d   that   f or   high   EE,   sign al   sym bo ls   detect or s   sh oul d   be   care fu ll y   sel ect ed   and   acc ordin g   to   a n   ap pro pr ia t e   op erati ng   sce nar i o.   In   orde r   to   im pr ove   EE ,   they   al so   sug gested   us in g   m edium   (6 - to - 7   bits) - quantiz at io n   res olu ti on   le vels   AD Cs   to   achie ve   a   tra de - off   betwee n   powe r   c onsu m ed   a nd   no nline ar   distor ti on.   More ov e r,   S ha o   et   a l ,   [21]   propose d   to   incr ease   the   up li nk   EE   of   the   m as sive   MIM O   syst e m s   un der   diff e re nt   op e ra ti ng   scena rios,   based   on   the   us e   of   low   (1   bit) - qu a ntiza ti on   reso l ution   l evels   A DCs,   it erati ve   detect ion   a nd   decodin g   ( IDD)   te ch nique,   and   li near - low   res olu ti on   awar e   MM SE   (LRA - MM SE )   signa l   sy m bo ls   detec tor.   Thei r   res ul ts   yi el ded   a   hi gh   EE   gain   a nd   im pr ov e d   BER   perform a nce   for   the   m assiv e   MIM O   syst em .   In   [ 22 ],   the   a uthors   i nv e sti ga te d   the   EE   in   the   m assive   MIM O   syst em   us in g   ar bitrar y   (1,   2,   and   i nf init y   bi ts) - qu a ntiza ti on   res olu ti on   le vels   A DCs,   s uperim po sed   pilots   (SP)   te ch nolo gy,   an d   L MM SE   and   m axi m u m   rati o   com bin ing   (MRC )   sig nal   sy m bo ls   de te ct or s   un der   diff e re nt   op e ra ti ng   scena rios.   Their   resu lt s   sho wed   that   the   pilots   need ed   to   sp e ci fy   a   gr eat er   powe r   i n   the   m assive   MIM O   syst e m   with   hig he r   qu a ntiza ti on   r esolutio n   le ve ls   A DCs   or   with   a   la rger   num ber   of   a nten nas   at   t he   BS.   In   a dd it ion ,   SP   te chnolo gy   is   su pe rio r   to   ti m e - m ulti plexed   pilots   (TP)   t echnolo gy   for   low   (1   bit) - quantiz at io n   r es olu ti on   le vels   A DCs,   a nd   SP   te ch no l ogy   is   m or e   s uitable   w he n   us in g   higher - qua ntiza ti on   res ol ution   le vels   A DCs   an d   the   use d   num ber   of   a nten nas   in   the   BS   is   sm al l.   Fu rt her m or e,   t he   auth ors   in   [2 3 ]   inv est igat ed   the   eff ect s   of   us in g   low   (1 - to - 3   bits) - qu a ntiza ti on   reso l ution   le ve ls   AD Cs   an d   the   num ber   of   anten nas   avai la ble   at   the   base   sta ti on   us i ng   LMM SE   s ign al   sy m bo ls   detect or   in   i m pr ovi ng   the   upli nk   sp ect r um   eff iciency   (S E)   of   the   m assive   MIM O   syst e m   unde r   diff e re nt   opera ti ng   sce nar i os .   Their   res ults   pro ved   to   im pr ove   SE   by   m or e   t han   90%   us ing   low   (3   bits) - qu a ntiza ti on   re so luti on   le vels   AD Cs.   In   ad di ti on ,   the   SE   im pr ov e d   wh e n   us ing   1   bit - qu antiz at ion   res ol ution   le vels   AD Cs   with   an   inc rea se   in   the   num ber   of   anten nas   at   the   base   sta ti on .   In   [2 4 ],   the   auth or s   pro pose d   a   novel   hybr i d   i nd e pe nd e nt   c om po nen t   a naly sis   (HICA )   sig nal   sym bo ls   de te ct or   in   orde r   to   im pr ove   the   SE   and   reli abili ty   of   t he   MIM O   syst e m   un de r   diff e re nt   opera ti ng   sce nar i os .   Their   res ults   s howe d   that   t he   HI C A   sign al   sym bo ls   detect or   sho w ed   i m pr ovem e nt   in   the   BER   and   MSE   perf or m ances,   but   it   was   ineff ic ie nt   and   fail ed   to   reduc e   the   pea k   to   a ver a ge   power   r at e   (P A PR)   rat e   com par ed   to   oth e r   sta te - of - t he - a rt   sig nal   sym bo ls   detect or s   s uch   as   ICA   a nd   wa velet   de no isi ng   ICA   ( WDICA ).   In   [2 5 ],   the   au thors   inv e sti ga te d   the   j oi nt   eff ect s   of   the   low   (1   bit) - qu a ntiza ti on   res ol ution   le vels   AD Cs   an d   the   hard war e   im pairm ents   in   the   m assive   MIM O   syst em ,   us in g   a   novel   pro posed   al te rn at in g   directi on   m et ho d   of   m ulti pli ers   ( A DMM) - base d   sig nal   s ym bo ls   detect or,   qua dr at ur e   phase - s hift   ke yi ng   (Q P S K)   m od ul at ion ,   a nd   different   nu m ber s   of   a nten nas   at   the   BS.   T he   re su lt s   of   thei r   inv e sti gation s   s how e d   that   the   AD M M - base d   sig na l   sy m bo ls   detect or   achie ve d   bette r   pe rfor m ance   tha n   the   oth e r   sta te - of - t he - a rt   li near   si gn al   sym bo ls   detect ors   in   te rm s   of   i m pr o ving   BER   pe rfor m ance   a nd   the   gai n   with   an   i ncr ease   in   the   nu m ber   of   ante nn a s   a nd   us e rs   in   the   m assive   MIM O   syst em .     Evaluation Warning : The document was created with Spire.PDF for Python.
Ind on esi a J  E le c Eng &  Co m Sci     IS S N:  25 02 - 4752       Evalu atio n and  m e as ure me nt  of the effe ct of d if fe rent  quan ti za ti on  re so l ution …  ( H ay der K ha le el  AL - Qaysi )   1495   In   this   pap e r,   t he   m ai n   ob j ect ive   is   to   evalu at e   and   m easure   the   influ e nce s   of   diff e re nt   quantiz at io n   reso l ution   le ve ls   of   analo g - to - dig it al   conv erters   on   the   BER   eff ect ive ness   in   the   m assive   MIM O   syst e m   unde r   dif fer e nt   op e rati ng   sce na rios   us i ng   the   LMM SE   sign a l   sy m bo ls   detect or .   T he   rem a ind e r   of   this   pa per   is   arr a ng e d:   T he   desc riptio ns   of   t he   qu a ntize d   m assive   MIM O   syst em   e qu at io ns   with   the   L MM SE   s ign a l   sy m bo ls   detect or   are   prese nt ed   in   Sect ion   2.   Ma tl ab   sim ula ti on   res ults   that   exp la in   the   infl uen c e   of   qu a ntiza ti on   re so luti on   le vels   analog - to - di gital   con ve rters   on   the   BER   ef fici ency   in   the   m assive   m ult i ple - input,   m ulti ple - outp ut   syst em   are   pre sente d   in   Sect io n   3.   Fi nally ,   Sect io n   4   c on cl ud e s   thi s   pa per.       2.   QUANTIZ ED   M AS SI VE   M IMO   WITH   LMM SE   SI GNAL   S YM B OL S   DETE CT O R   First,   in   this   s ect ion ,   we   pr e sent   the   quant iz ed   m assive   MIM O   syst em   m od el ,   w her e   dif fer e nt - qu a ntiza ti on   re so luti on   le vels   AD Cs   a re   us ed   in   the   receive r   at   the   BS.   T he n   we   prese nt   the   LMM SE   det ect or   to   detect   t he   tr ansm itted   sig na l   sy m bo ls   vec tor   by   the   us e rs   at   the   BS   recei ver.     2.1.      Q uantiz ed   ma ssive   MI MO   mo del   w ith   ADCs   We   co ns ide r   that   the   up li nk   of   the   m assive   m ulti us er   (MU)   MIM O   s yst e m ,   as   sh own   in   Figure   1,   has   a   total   M   anten nas   at   the   BS   and   sim ultaneo us ly   ser ves   K   sin gle - a nten na   us e rs,   w here   M K 1 .   Th us ,   the   cha nnel   m a trix   ( H )   co nnect ing   t he   c hannel   coeffic ie nts   bet ween   K   si ng le - anten na   us ers   a nd   M   a nten nas   at   the   BS   is   giv e n   b y   (1)   [7 ] ,   [ 2 6 ] ,   [ 2 7 ]     H = [ h 11 h 1j h i1 h ij ]   (1)     wh e re   H M × K   a nd   the   el em ent   h ij   re pr ese nt   t he   c ha nn el   at te nuat ion   c oeffici ent   from   the   ith   r ecei ving   anten na   to   t he   j th   tra ns m it te r   anten na.   T he   c hannel   m at rix   (H)   in   the   m ass ive   MIM O   syst e m   pr ovide s   us   wit h   the   necessa ry   knowle dge   ab out   the   cha nn el   sta te   info rm at i on   (CSI)   an d   thu s   helps   s olve   m any   pr oble m s   that   we   face   in   wir el ess   com m un ic at ion   syst em s   su c h   as   path   losses,   fa ding,   sh a dowing,   sc at te ring ,   di ffra ct ion s,   and   pr op e rtie s   of   the   c ha nn el   [27 ] ,   [ 28] .             Figure   1.   U plin k   syst em   m od el   of   qua ntize d   m assive   MU - MIM O       The   i nfor m at ion   sig nals   sym bo ls   ( a K )   tra ns m it t ed   by   al l   the   K   sing le - ante nna   us e rs   a re   e ncoded   fir st   by   the   c ha nn e l   encode rs   a nd   the n   m odulate d   acco r ding   to   the   gi ven   m od ulati on   sc hem es   to   the   sign al   sy m bo ls   ( x K ).   Th ese   m od ulate d   sign al   sym bo ls   ( x K )   are   tra ns m itted   t hroug h   the   c hannels.   T he   M   ante nn a s   at   the   BS   re cei ve   these   m od ul at ed   sig nal   sy m bo ls   ( x K )   c orr up te d   by   the   channel   e ff ect s   an d   no ise .   The   corrupted   an d   unquantiz ed   si gn al   sym bo ls   ve ct or   receive d   in   the   base   sta ti on   is   gi ven   by   ( 2) ,       y = Hx + n   (2)     Evaluation Warning : The document was created with Spire.PDF for Python.
                          IS S N :   2502 - 4752   Ind on esi a J  E le c Eng &  Co m Sci,   Vo l.   23 , N o.   3 Se ptem ber  2 02 1 14 93   -   15 00   1496   wh e re   x = [ x 1 , x 2 , x 3 , . , x K ] T   is   the   vect or   of   the   tra ns m i tt ed   sign al   sym bo ls   f ro m   al l   use rs,   y = [ y 1 , y 2 , y 3 , . , y M ] T   i s   the   vect or   of   the   recei ved   sign al   sym bo ls ,   an d   n   is   the   a dd it ive   w hite   Gau s sia n   nose   vecto r.   T he   m od el   giv e n   in   ( 2)   re prese nts   the   m assive   MIM O   syst e m   mo del   th rou gh   wh ic h   it   is   po s sible   to   fin d   the   cha nn el   m at rix   (H)   ba sed   on   the   pro vid e d   in f or m ation   a bo ut   t he   x   an d   y   si gn al   s ym bo ls   vect or s .   G e n e r a l l y ,   t h e   s t r u c t u r e   of   A D C s   c o n s i s t s   of   f i l t e r s ,   s a m p l i n g   a n d   h o l d   c i r c u i t s ,   a n d   q u a n t i z e r s .   P a r t i c u l a r l y ,   t he   q u a n t i z e r s   r e p r e s e n t   t h e   c o r e s   of   A D C s   t h a t   p e r f o r m   t h e   q u a n t i z a t i o n   p r o c e s s e s   for   t h e   r e c e i v e d   s i g n a l   s y m b ol s   v e c t o r   ( y )   g i v e n   in   (2)   i n t o   the   q u a n t i z e d   s i g n a l   s y m b o l s   v e r s i o n   v e c t o r   (r)   [1 4 ] ,   [ 1 8 ].   As   s h o w n   in   F i g u r e   1,   by   u t i l i z a t i o n   of   two   b   b i t - q u a n t i z a t i o n   r e s o l u t i on   l e v e l s   A D C s   for   t h e   r e a l   a nd   i m a g i n a r y   p a r t s   of   t h e   c om pl e x   r e c e i v e d   s i g n a l   s y m b ol s   v e c t o r   ( y ) ,   t h e   o u t c om e   q u a n t i z e d   s i g n a l   s y m b ol s   v e c t o r   (r)   is   g i v e n   by   (3),      r = ( y ) = [ e ( y ) + j   Ι m ( y ) ]   (3)     wh e re     (.)   de s cribes   the   qu a ntiza ti on   pr oc ess   for   the   re al   ( e ( y ) )   and   im aginar y   ( Ι m ( y ) )   pa rts   of   the   com plex   receiv ed   si gn al   sym bo ls   vector   (y) ,   and   it   is   a   non - li near   process.   The   cha racteri sti cs   of   3   bits   reg ular   m id   as cend i ng   qu a ntiza ti on   proces s es   are   sh ow n   in   Fig ur e   2.   Assum ing   that   (.)   re presenti ng   b   bits   re gula r   m id   ascend i ng   qu a ntiz at ion   proce sse s   ac hieve d   by   the   qu a ntize r   with   N = 2 b   res olu ti on   le vels   of   t he   A D Cs.   C on se que nt ly ,   the   quanti zer   pe rfor m s   the   co nversi on   process   to   the   real   pa rt   of   the   input   sig nal   sy m bo ls   vect or   (y)   into   a   re al   evaluated   outp ut   sig nal   sy m bo ls   vecto r   (r),   w he re   r = [ r 1 , r 2 , r 3 , , r ]   a nd   s = 1 , 2 , 3 , , N .   T he   en dpoi nts   of   the   in put   inte r val   ( y s )   are   giv e n   by   ( 4),      y s = {                                                           for   s = 1             ( N 2 1 + s )                                     f or   s = 2 , 3 , 4 , , N +                                                               for   s = N + 1   (4)     wh e re     re prese nts   the   per i od   siz e   of   qu a ntiz at ion   process .   The   pro duced   sign al   sym bo ls   val ues   ( r s )   of   the   qu a ntize r   a re   gi ven   by   (5),       r s = ( N 2 1 2 + s )                                   for   s = 2 , 3 , 4 , , N   (5)           Figure   2.   Stai r case   re pr ese nta ti on   of   a   th ree   bits   re gu la r   m i d   asce ndin g   quantiz at ion   proc esses       2.2.      L MMSE   sign al   s ymbol s   detec t or   As   a   res ult   of   the   interuse r   interfe ren ce   a m on g   the   us e r s,   the   sig nal   sy m bo ls   detect or   is   use d   to   evaluat e   the   ve ct or   of   the   t ransm it te d   sign al   sy m bo ls   ( x )   from   the   vector   of   t he   recei ved   sign al   sym bo ls   ( y ) .   In   this   wor k,   we   hav e   a dopt ed   the   LMM S E   sig nal   sym bo ls   detect or   fro m   the   oth e r   s ta te - of - the - art   detect or s   because   of   its   a dv a ntage s,   i nclud i ng   [7 ] ,   [ 1 6 ],      It   has   lo wer   co m pu ta ti on al   co m plexit y.     It   has   noise   im pro vem ent.     It   outpe rfor m s   the   ZF   sig nal   s ym bo l   detect or .   Evaluation Warning : The document was created with Spire.PDF for Python.
Ind on esi a J  E le c Eng &  Co m Sci     IS S N:  25 02 - 4752       Evalu atio n and  m e as ure me nt  of the effe ct of d if fe rent  quan ti za ti on  re so l ution …  ( H ay der K ha le el  AL - Qaysi )   1497   Im ple m entation   the   li nea r   sig nal   detect or   by   us in g   K   ma trix   ( A )   to   the   receive d   sig na l   sy m bo ls   vecto r   (y ),   t he   transm itted   sig nal   sym bo ls   ve ct or   is   giv e n   by   (6),       x ̂ = A H y = A H ( Hx + n )   (6)     wh e re   x ̂ = [ x ̂ 1 , x ̂ 2 , x ̂ 3 , . , x ̂ K ] T   is   K × 1   si gnal   cha racters   ve ct or   co ns ist s   of   data   in form at ion   st ream   of   K   sing le - a nten a   us er .   Em plo yi ng   the   LMM S E   sig nal   sym bo ls   detect or   by   def i ning   t he   detect io n   m atr ix   data   stream   (A)   as   gi ven   by   (7),       A = H   [ H + n 2 x 2 I K ] 1   (7)     wh e re   x 2   a nd   n 2   are   the   tra ns m itted   ori gi nal   sig nal   a nd   t he   receive d   noise   sig na l   var ia nces   corres pondin gly .   For   a   qua ntize d   m assive   MU - MIM O   syst e m ,   we   hav e   the   Kth   el em ent   of   x ̂   as   gi ven   by   ( 8)     x ̂ = A H r = A H ( y ) = A H   ( Hx + n ) = a K H   h K x K + a K H   h i x i K i k + a K H   n   (8)     wh e re   the   first   te rm   in   (8)   re presents   the   de s ired   sig nal,   the   second   te rm   r epr ese nts   the   inter us er   inter fe ren ce ,   and   the   t hir d   te rm   rep rese nts   t he   no ise .       3.   SIMULATI O N   RESU LT S   In   t his   sect io n   of   the   pa per,   MATLAB   sim ulati on s   are   a ppli ed   to   ex plai n   the   ef fects   of   di ff e ren t   qu a ntiza ti on   re so luti on   le vels   of   a nalo g - to - dig it al   co nvert er   ( A DCs)   on   the   BER   pe rfor m ance   in   m assive   m ul ti ple   input   m ul ti ple   ou t p ut   (MIM O)   syst em s   un der   di ff e ren t   operati ng   scenari os .     3.1.       BER   per fo rm an ce   usin g   di ff eren t   sc h emes   of   m odu lation   Figures   3   a nd   4   show   the   BE R   per f or m ances   us in g   two   di ff e ren t   schem es   of   quad ratu r e   a m plit ud e   m od ulati on ,   4   QAM   a nd   16   QA M,   an d   dif f eren t   qu antiz at ion   res olu ti on   l evels   of   AD Cs   an d   LMM SE   sign a l   sy m bo ls   detect or   ha ve   been   us e d   at   the   B S.   Sim ulati on s   res ults   are   im plem ented   with   80   ante nn as   at   the   transm itter   an d   ser ving   5   ante nn a s   as   a   us e r’s.   T he   ef fects   of   dif fer e nt   quantiz at ion   resol ution   le vels   of   A DCs   on   the   BER   pe rfor m ance   are   sh ow n   as   a   sig nal   to   noise   rat io   (SNR)   get   bi gg e r,   the   re su l t   dem on strat e   t hat   the   BER   de gr a dation   ex pone ntial ly   in   a   sim i la r   way   for   bo t h   schem es   of   m od ulati on s   w hen   usi ng   the   sam e   qu a ntiza ti on   re so luti on   le vels   of   A DCs.   Ne ve rtheless,   the   r esults   il lustrate   that   by   i ncr ea sing   the   quanti zat ion   reso l ution   le ve ls   of   A DCs,   the   BER   pe rfor m ance   beco m e s   m or e   ideal .             F i g u r e   3.   B E R   p e r f o r m a n c e   u s i n g   d i f f e r e n t   q u a n t i z a t i o n   r e s o l u t i o n   l e v e l s   of   A D C s ,     =   8 0 ,     =   5,   a n d   4   Q A M       In   orde r   to   stu dy   highe r - orde r   m od ulati on s ,   we   obse rv e   t hat   with   us in g   2   an d   3   bits - qu a ntiza ti on   reso l ution   le ve ls   of   AD Cs ,   the   BER   sh ow i ng   degrade s,   wh il e   us in g   the   conditi on   4   bits - qu a nt iz at ion   reso l ution   le ve ls,   the   BER   beco m e s   acce ptab le .   Furthe rm or e,   ha ving   5   bits - qua ntiza ti on   reso l ution   le ve ls   and   Evaluation Warning : The document was created with Spire.PDF for Python.
                          IS S N :   2502 - 4752   Ind on esi a J  E le c Eng &  Co m Sci,   Vo l.   23 , N o.   3 Se ptem ber  2 02 1 14 93   -   15 00   1498   ov e r head,   the   BER   eff ect ive ness   is   ve ry   cl ose   to   the   in finity   bits - qua ntiza ti on   r esol ution   le vels.   T he   ou t com es   sh ow   wh e n   us i ng   highe r   orde r   of   m od ulati on   (4   Q AM   a nd   16   Q AM),   t he   BER   increas es   corres pondin gly   for   sp eci fic   quanti zat ion   res olu ti on   le vels   of   A DCs.   Co ns eq ue ntly ,   hig he r   quantiz at io n   res olu ti on   le vels   of   A DCs   are   ob li gat or y   for   a dv a nce d   orde r   m od ulati ons   to   at ta in   the   equ i valent   perf or m ance.           F i g u r e   4.   B E R   p e r f o r m a n c e   u s i n g   d i f f e r e n t   q u a n t i z a t i o n   r e s o l u t i o n   l e v e l s   of   A D C s ,     =   8 0 ,     =   5,   a n d   16   QAM       3.2.       BER   per fo rm an ce   usin g   di ff eren t   nu mber   of   anten na s     Furthe r   sim ul at ion   is   acc ompli sh e d   to   dete rm ine   BER   per f or m ance   as   a   functi on   of   qu a ntiza ti on   reso l ution   le ve ls   of   AD Cs   with   a   diff e re nt   nu m ber   of   BS   ante nn as ,   a nd   LMM SE   si gnal   sy m bo ls   detect or   hav e   bee n   us e d.   We   exam ine   a   con sta nt   nu m ber   of   us e rs   as   a   receiver   =   5,   an d   the   num ber   of   ante nn as   at   the   BS   will   be   ta ken   from   60   to   300   ante nna s .   Fig ures   5   an d   6   ex plain   t ha t   increasin g   t he   num ber   of   ant enn a s   at   the   tra ns m itt er   will   decr eas e   the   BER   re gressi on   cau sed   by   the   low   qu a ntiza ti on   res olu ti on   le vels   of   AD C s .     Figure   5   il lustrate   a   2   bits - quantiz at io n   res olu ti on   le vels   AD Cs   with   th r ee   differe nt   nu m ber   of   the   a nt enna   (60,   120 ,   an d   300)   has   an   obvi ous   cha ng e   with   the   incr ease   the   nu m ber   of   a nten nas .   Com par ably ,   with     F i g u r e   6,   t h e   B E R   p e r f o r m a n c e   of   4   b i t s - q u a n t i z a t i o n   r e s o l u t i o n   l e v e l s   A D C s   w i t h   t h e   s a m e   di f f e r e n t   a nt e n n a s   h a v e   a   s l i g h t   d e g r a d a t i o n .   W h e r e f o r e ,   t h e   h i gh e r   n u m b e r   of   a n t e n n a s   at   t h e   BS   is   prop o s e d   to   r e c o v e r   t h e   B E R   p e r f o r m a n c e   r e g r e s s i o n   in   t h e   s i t u a t i o n   of   h a v i n g   c o n s t a n t   l o w   q u a n t i z a t i o n   r e s o l u t i o n   l e v e l s   A D C s .             Figure   5.   BER   perform ance   usi ng   2   bit - qu a nt iz at ion   res olu t ion   le vels   ADC s   an d   4   QAM   Evaluation Warning : The document was created with Spire.PDF for Python.
Ind on esi a J  E le c Eng &  Co m Sci     IS S N:  25 02 - 4752       Evalu atio n and  m e as ure me nt  of the effe ct of d if fe rent  quan ti za ti on  re so l ution …  ( H ay der K ha le el  AL - Qaysi )   1499       Figure   6.   BER   perform ance   usi ng   4   bit - q ua nt iz at ion   res olu t ion   le vels   ADC s   an d   4   QAM       4.   CONCL US I O N   In   this   w ork,   we   eval uated   and   m easur e d   the   eff ect s   of   diff e re nt   quant iz at ion   res olu ti on   le vels   of   AD Cs   on   t he   BER   per f orm a nce   of   m assive   MIM O   syst em s   un de r   di ff e ren t   ope rati ng   scenari os .   T he   two   ty pical   m od ula ti on s,   4   Q AM   and   16   Q AM   ha ve   bee n   us e d   with   LMM SE   sign al   sym bo ls   detect or   to   ev al uate   the   syst e m .   The   resu lt s   dem on strat ed   t hat   the   perform ance   of   BER   in   m as sive   MIM O   te chnolo gy   is   affe ct ed   directl y   by   the   qu a ntiza ti on   r esolutio n   le vel s   of   AD Cs ,   t he   nu m ber   of   anten nas ,   a nd   m od ulati on   sc hem es.   Lo w   (1   bit   a nd   2   bits) - qua ntiza ti on   re so l ution   le vels   A DCs   can   reali ze   appr opriat e   BER   perfor m ances   com par ed   to   the   high   qua ntiza ti on   res olu ti on   le vels   A D Cs   wh e n   the   nu m ber   of   ant enn a s   at   the   BS   high.   More o ve r,   we   ob s er ve   that   us ing   4   bits - quan ti zat ion   reso l ut ion   le vels   ADC s   in   16   Q AM ,   the   S NR   of   a rou nd   15.5   dB   is   obli gato ry   to   at ta in ed   the   B ER   of   10 3 ,   w her ea s   usi ng   4   bits - quanti zat ion   r esol ution   le vels   AD C s   in   4   QAM   re quire   only   6   dB   SN R.   We   c on c lud e   th at   incre asi ng   the   m od ulati on   or der   will   sign ific ant ly   aff ect   the   BER   perf or m ance   of   hi gh - qua ntiza ti on   r esol ution   le vels   AD Cs   wh il e   sli ghtl y   aff ect e d   the   BER   perform ance   w hen   low - quanti zat ion   reso l ution   le vels   ha ve   been   use d.       REFERE NCE S   [1]   Y.   M.   Ta b ra   an d   B.   M.   Sabb ar,   "H y brid   MVDR - LMS   Bea m form ing   for   Mass ive   MIM O,"   Indone sian   Journal   of   El e ct rica l   Enginee ring   and   Computer   Sci e nce   ( IJE ECS) ,   vol.   16,   no .   2,   pp.   715 - 723,   2019 ,   do i:   10. 11591/ijeecs. v16. i2. pp715 - 72 3 .     [ 2 ]   P.   S u n i t a ,   R o s a l i n   S a m a n t a r a y,   P r a d yu m n a   K u .   M o h a p a t r a ,   R.   N.   P a n d a   a n d   P a d m a   S a h u ,   "A   N e w   C o m p l e x i t y   R e d u c t i o n   M e t h o d s   of   V - B L A S T   M I M O   S ys t e m   in   a   C o m m u n i c a t i o n   C h a n n e l , "   I n t e r n a t i o n a l   J o u r n a l   of   I n f o r m a t i cs   and   C o m m u n i c a t i o n   T e c h n o l o g y   (IJ - I C T ) ,   v o l .   8,   n o .   1,   pp.   29 - 38,   2 0 1 9 ,   d o i :   1 0 . 1 1 5 9 1 / i j i c t . v 8 i 1 . p p 2 9 - 38 .     [3]   O.   Agboje ,   Ns i kan  Nkordeh ,   U za iru Stan ley   I dia ke ,   Olol ade  Olado y in,  Kenn ed y   Okokpuji and   Ibin abo  Bo b - Manue l ,   "M IMO   Channe ls:   O pti m iz ing   Throu ghput   and   R ed uci ng   Outag e   by   In creasing   Multi pl exi ng   Ga in , "   TEL KOMNIKA   Tele communic a t ion,   Computing ,   El e ct ronics   and   Control ,   vo l.   1 8,   no.   1,   pp .   41 9 - 426,   2020 ,   doi :   10. 12928/telkomnika. v18 i1. 8720 .     [4]   M.   K.   Hus sein   a nd   N.   N.   Kh amiss ,   "Inte gra t ing   Mill imete r   W av e   with   H y br id   Pr ec oding   Multi us er   Mass ive   MIM O   for   5G   Com m un ic a ti on, "   TEL KOMNIKA   Tel ec om municat i on ,   Co mputing,   El e ct ro nic s   and   Con trol ,   vo l.   18,   no.   1,   pp.   90 - 98 ,   2020 ,   doi:   10. 12928 /t e lkomnika. v18i1 . 13674 .     [5]   A.   Nalba nd ,   M.   Sarva g y a,   and   M.   R.   Ahm ed,   " Pow er   Saving   a nd   Optimal   H y b rid   Prec oding   in   Mill imet er   W a ve   Mass ive   MIM O   Sy st ems   for   5G . "   TEL KOMNIK A   Tele communic ati on,   Computin g,   El e ct ronics   a nd   Control ,   vol.   18,   no .   6,   pp .   28 42 - 2851,   2020 ,   doi:   10 . 12928/te lkomnika. v18i6 . 15952 .   [6]   L.   Van   d er   Perre ,   L.   Li u ,   and   E.   G.   La rss on,   "Eff ic i ent   DSP   and   Circ uit   Ar chi t ec t ure s   for   Mass ive   MIM O:   Stat e   of   the   Art   and   Fut ure   Dire ct ions , "   IEEE   Tr ansactions   on   Signal   P roce ss ing ,   vol .   66,   no.   18,   pp.   4717 - 4736,   Sept .   2018 ,   doi :   10 . 11 09/T SP . 2018. 28 58190 .   [7]   M.   A.   Albre em,   M.   Juntti   and   S.   Shahabuddi n,   "M assive   M IMO   Dete c ti on   Te chni qu es:   A   Surve y , "   IEEE   Comm unic ati ons   Surve ys   &   Tuto rials ,   vol .   21 ,   no .   4,   pp .   3109 - 31 32,   2019 ,   doi :   10 . 1109/COMS T. 2 019. 2935810 .   [8]   C.   Stergi ou,   E.   P.   Kos ta s,   B.   B.   Gupta   and   Y.   Ishibashi,   "S ec ur ity ,   Priva c y   &   Eff iciency   of   Sus ta ina bl e   Clou d   Com puti ng   for   Big   Data   &   IoT,"   Sustainabl e   Computing:   Informatic s   and   Syste m s ,   vol.   19,   pp.   17 4 - 184,   2018 ,   doi :   10. 1016/j.sus com.2018. 06. 003 .   Evaluation Warning : The document was created with Spire.PDF for Python.
                          IS S N :   2502 - 4752   Ind on esi a J  E le c Eng &  Co m Sci,   Vo l.   23 , N o.   3 Se ptem ber  2 02 1 14 93   -   15 00   1500   [9]   R.   Yang,   Shuqi   Xi ,   Qibo   C ai ,   Z hiz hou   Chen ,   Xi aoha ng   W ang ,   Gui   Liu ,   "A   Co m pac t   Plan ar   D ual - Band   Multi p le - Input   and   Multi ple - Output   An tenna   with   High   Is ola ti on   for   5G   a nd   4G   Applications , "   Mic rom ac hine s ,   vo l.   12 ,   n o.   5,   pp .   1 - 8,   Ma y   2 021,   doi :   10 . 33 90/mi12050544 .     [10]   P.   Varz aka s,   "A ver age   Ch anne l   Capa c ity   for   Ra y l ei gh   Fading   S pre ad   Spectrum   MIM O   Sy st ems , "   Inte rnational   Journal   of   Comm unic ati on   System s ,   vol.   19,   no.   10,   pp .   108 1 - 10 87,   De c.   2006 ,   d oi:   10 . 1002/dac.784 .   [11]   C.   W ang,   C.   W en,   S.   Jin,   and   S.   Tsa i,   "F ini te - A lpha be t   Prec od i ng   for   Mass ive   MU - MIMO   with   Low - Resoluti on   DA Cs,"   IEE E   Tr ansacti ons   on   Wireless   Comm unic ati ons ,   vo l.   17 ,   no .   7,   p p.   4706 - 4720 ,   Jul y   2018 ,   do i:   10. 1109/t wc . 201 8. 2830343 .   [12]   S.   Jac obss on,   Giuseppe   Durisi ,   Mika el   Co ldrey ,   Ulf   Gus ta vss on ,   Christoph   S tude r ,   "Through put   Anal y sis   of   Mass ive   MIM O   Uplink   wi th   Lo w - Resolut ion   A DCs , "   IEE E   Tr ansacti ons   on   W i rele ss   Comm u nic ati ons ,   vol .   16 ,   no.   6,   pp .   4038 - 4051,   June   2017 ,   doi :   10 . 1109/TW C. 2017. 2691318 .     [13]   J.   Choi,   Y.   Cho ,   and   B.   L.   Eva n s,   "Q uant ized   M assive   MIM O   Sy stems   with   Mu lt icell   Coordin ated   Be amform ing   and   Pow er   Con trol , "   I EEE   Tr ansacti ons   on   C omm unic ati ons ,   vol.   69,   no .   2,   pp.   946 - 961,   Feb.   2021 ,   doi:   10. 1109/T COM M.2020. 303668 9 .   [14]   T.   Li u,   Jun   Ton g,   Qinghua   Guo,   Jiang ta o   Xi ,   Yangua ng   Yu   a nd   Zhi t ao   Xi ao ,   "Ene rg y   Eff ici ency   of   Mass ive   MIM O   Sy stems   with   Low - Reso lut ion   AD Cs   an d   Succe ss ive   In te rfe r ence   Can c el l at ion , "   IEEE   Tra nsacti ons   on   Wirel ess   Comm unic ati ons ,   vol .   1 8,   no .   8,   pp .   398 7 - 4002,   Aug.   20 19 ,   doi :   10 . 1109 /T W C. 2019. 292 0129 .   [15]   J.   Chen,   S.   Che n,   Y.   Qi ,   and   S.   Fu,   "Int el l ige n t   Mass ive   MIM O   Antenna   Se lecti on   Us ing   Mo nte   Ca rlo   Tree   Sear ch, "   I EE E   Tr ansacti ons   on   Signal   Pro ce ss ing ,   vol .   6 7,   no.   20,   pp .   5380 - 5390,   Oct.   2019 ,   do i :   10. 1109/T SP . 20 19. 2940128 .     [16]   Z.   Xi ao,   Jinca n   Zha o ,   T i anle   L iu ,   L ei   Geng ,   F.   Z hang   and   Jun   To ng ,   "On   the   En er g y   Eff i ci en c y   of   Mass ive   MIMO   S y stems   with   Low - Resolut ion   AD Cs   and   La tt i ce   Reduction   Ai ded   Dete c tors, "   Symmet ry ,   vol .   12,   no. 3,   pp .   1 - 20,   2020,   doi :   10 . 33 90/s y m 1203040 6 .   [17]   H.   W ang,   W.   S hih,   C.   W en,   a nd   S.   Jin,   "Reli abl e   OFDM   Re ce iv er   with   Ultr a - Low   Resolut i on   AD C, "   IEE E   Tr ansacti ons   on   Comm unic ati ons ,   vol .   67 ,   no .   5,   p p.   3566 - 3579 ,   M a y   2019 ,   doi :   10 . 1109/T COM M.2019. 2894629 .   [18]   J.   Li u,   Z.   Luo,   a nd   X.   Xiong,   "Low - Resolut ion   A DCs   for   W ire le ss   Comm unic at ion:   A   Com pre h ensive   Surve y , "   in   IEE E   Acce ss ,   vo l.   7,   pp .   91291 - 9 1324,   2019 ,   doi :   10. 1109/ACCESS . 2019. 2927891 .   [19]   J.   Zha ng ,   Li ng l ong   Dai ,   Ziy an   He ,   Shi   Jin ,   and   Xu   Li ,   "P erf or m anc e   Anal y sis   of   Mixed - ADC   Mas sive   MIM O   S y stems   Over   R ic i an   Fading   Channe ls,"   IE EE   J ournal   on   Sel ec t ed   Areas   in   Comm unic ati ons ,   v ol.   35,   no.   6,   pp .   1327 - 1338,   201 7 ,   doi :   10 . 1109/J SA C. 2017. 2687278 .   [20]   T.   W aday am a   a nd   S.   Ta k abe,   " Quanti z er   Opti m iz at ion   B ase d   on   Neura l   Quan ti z erf or   Sum - Pr odu ct   De code r , "   in   2018   IEE E   Glob al   Comm unic ations   Confe renc e   ( G LO BE COM ) ,   Abu   Dhabi,   Unit ed   Arab   Emirat e s,   2018 ,   pp.   1 - 6 ,   doi:   10 . 1109/GL OCO M.2018. 8647503 .     [21]   Z.   Shao,   R.   C.   de   La m are,   and   L.   T.   N.   La nd au,   "Ite r ative   D et e ct ion   and   De codi ng   for   Larg e - Scale   Multi p le - Antenna   S y s te m s   with   1 - Bit   AD Cs,"   IEE E   Wireless   Comm unic ati ons   Lett ers ,   vo l.   7,   no.   3,   pp .   476 - 479,   June   2018 ,   doi :   10 . 1109/LW C. 2017. 2787159 .   [22]   C.   Chen,   W.   Zh ang,   and   X.   B ao ,   "A chi eva b le   R at e   Anal y sis   on   the   Uplink   of   Mass ive   MIM O   w it h   Superimpos ed   Pilot s   and   Arbit rar y - Bit   AD Cs,"   in   2019   IE EE   19th   In te rnatio nal   Conf ere nce   on   Comm unic a ti on   Te chnol og y   ( ICCT ) ,   Xi' an ,   C hina ,   2019 ,   pp.   6 74 - 678 ,   doi :   10 . 1109/ICCT 4680 5. 2019. 8947297 .   [23]   Y.   Dong   and   L.   Qiu,   "S pectra l   Eff iciency   of   M assive   MIM O   Sy stems   with   Lo w - Resolu ti on   A DCs   and   MM S E   Rec e ive r , "   IE EE   Comm unications   Letters ,   vol.   21,   n o.   8,   pp.   1 771 - 1774,   Au g.   2017 ,   doi :   10. 1109/L COM M.2017. 269327 6 .     [24]   R.   Bh anda r i   a nd   S.   Jadha v ,   " Spect ral   Eff i ci en t   B li nd   Ch anne l   Est imati o n   T ec hniqu e   f or   MIM O - OF D M   Com m unic at ions, "   Inte rnat iona l   Journal   of   Ad va nce s   in   Applied   Sci en ce s   ( IJA AS ) ,   vol.   7,   no.   3,   p p.   286 - 297,   201 8 ,   doi:   10 . 11591/ij aa s.v7.i3. pp286 - 297 .     [25]   Ö.   T.   Dem ir   an d   E.   Björnson,   "ADMM - Based   One - Bit   Qua n ti z ed   Signal   Det ection   for   Mass ive   MIM O   Sy st ems   with   Hardware   Im pai rm ent s,"   in   ICASSP   2020 - 2020   IEE E   Int ernati onal   Conf ere nce   on   Ac ou stic s,   Speech   and   Signal   Proce ss in g   ( ICASSP ) ,   Bar ce lon a,   Spain,   2 020 ,   pp .   9120 - 9 124 ,   doi :   10 . 110 9/i c assp40776.2020. 9053984 .   [26]   S.   Jac obss on,   Giuseppe   Durisi,   Mikae l   Coldr e y ,   Ulf   Gus ta vss on   a nd   Christoph   Studer ,   "O ne - Bit   Mass ive   MIM O :   Channe l   Esti m ation   and   High - O rde r   Modulation s,"   in   2015   IE E E   Inte rnat ional   Confe renc e   on   Comm unic ati on   Workshop   ( ICC W) ,   London,   UK,   2015 ,   pp .   1304 - 13 09 ,   doi :   10 . 1 109/ICCW.2015. 7247358 .   [27]   V.   Bhat ia,   M.   R.   Tri pat h y ,   and   P.   Ranj an ,   "D ee p   Le arn ing   for   Mass ive   MIM O:   C hal l enge s   and   Future   Pros pec ts, "   in   2020   IEE E   9th   Inte rnationa l   Confe renc e   on   Comm unic ati on   Syste ms   and   Net work   Technol ogie s   ( CSNT ) ,   Gw al ior,   In dia,   pp.   26 - 31 ,   2020 ,   doi:   10. 1109 /CSN T48778. 2020. 9115783 .   [28]   L.   V.   Ngu y en ,   D u y   Trong  Ngo ,   Nghi  H.  Tra n ,   A .   L ee  Sw indl ehu rst  and  Du y   H .   N.  Ngu y en ,   "S uper vised   and   Se m i - Supervised   Lear ning   for   MIM O   Bli nd   Det ec t io n   with   Low - Res olut ion   AD Cs,"   IEE E   Tr ansact ions   o n   Wirel es s   Comm unic ati ons ,   vol .   19 ,   no .   4,   p p.   2427 - 2442 ,   A pril   2020 ,   do i:   1 0. 1109/t wc . 2020 . 2964661 .   Evaluation Warning : The document was created with Spire.PDF for Python.