Key Planning Areas parameters – TD-LTE PRACH Planning PCI Planning UL DM RS Planning
Company Confidential
PRACH Planning Principle PRACH conig!ration t"o cells m!st #e $ierent "it%in t%e PRACH re-!se $istance to increase t%e RACH $eco$ing s!ccess rate
PRACH transmission an !e separated !y"
# Time &prac%ConIn$e'( $ PRACH%P&SCH interferene" interferene" 'f PRACH resoures are separated in time within eN( $ PRACH%PRACH interfer interferene" ene" 'f same PRACH resoures are used for the ells of an eNode()
$ PRACH%PRACH interferene is preferred to PRACH%P&SCH interferene so prac%ConIn$e' o t%e cells on one site s%o!l$ #e t%e same
# )re*!ency &prac%)re*+( $ Alloation of PRACH area area should !e ne*t ne*t to P&CCH area either at upper upper or lower !order of fre+ueny !and, howe-er should not o-erlap with P&CCH area
$ A-oid separation separation of P&SCH in two two areas !y PRACH .sheduler .sheduler an only handle one P&SCH area
$ )or simplicity !se same conig!ration or all cells
# Se*!ence &PRACH CS an$ RootSe*In$e'( $ &se different se+uenes for all nei0h!our ells Company Confidential
Pream#le )ormats, )DDTDD common # PP .3S4)211 .3S4)211 speifies 5 random aess formats ommon ommon for !oth 6DD/3DD
# Differene in formats is !ased in the different durations for the yli prefi*, se+uene and 0uard time whih ha-e an effet on the t he ma*imum ell radius
# +nly )ormats . an$ / are s!pporte$ in initial releases &!p to RL0.( Reommendation"
Company Confidential
Selet 6ormat7 for ell ran0es 815)9 km
Selet 6ormat1 for ell ran0es 8::)5 km
Pream#le )ormat 1, TDD-only # PP .3S4)211 speifies a speial random aess format 5 for 3DD # 's alloated in &pP3S when it is two sym!ols lon0
# ;a*imum ell radius with pream!le format 5 is a!out 1)5km # N<3=" when usin0 pream!le format 5, only root se+uenes 7%1: are allowed) # N<3=" when usin0 pream!le format 5, allowed -alue of prahCS is 5)))4) Settin0 prahCS >4 0i-es the ma*imum ell ran0e 81)5km
# N<3=" prah6re+
# Recommen$ation3 !se PRACH $ensity / or start
# Sine PRACH performane measurement ounters are a-aila!le .R@19 $ ;771% Cell @oad it will !e possi!le to e-aluate the amount of PRACH / RACH proedures in time and adapt /optimiBe the settin0s
# )!t!re eat!res3 PRACH ;ana0ement .auto% onfi0uration, future S
Company Confidential
PRACH )re*!ency +set prac%)re*+
# 'ndiates the first PR( a-aila!le for PRACH in the &@ fre+ueny !and # PRACH area .4 PR(s should !e ne*t to P&CCH area either at upper or lower !order of fre+ueny !and to ma'imi4e t%e PUSCH area !ut not o5erlap "it% PUCCH area
# Parameter is onfi0ured !ased on the P&CCH re0ion .see P&CCH dimensionin0 i)e) its -alue depends on how many P&CCH resoures are a-aila!le)
# 'f PRACH area is plaed at the lower !order of &@ fre+ueny !and then" PRACH%6re+ueny roundup P&CCH resoures/2E
# 'f PRACH area is plaed at the upper !order of the &@ fre+ueny !and then" PRACH%6re+ueny NR( %4% roundup P&CCH resoures/2E
N<3=" prah6re+
PRACH Cyclic S%it or pream#les .-0 Prac%CS # PrahCS $eines t%e conig!ration used for the pream!le 0eneration) i)e) how many yli shifts are needed to 0enerate the pream!le
# Prac%CS $epen$s on t%e cell si4e $ Different ell ran0es orrespond to different PrahCS
# Simplifiation" 3o assume all ells ha-e same siBe .limited !y the prachConfIndex
Reommendation" Selet PrahCS !ased on the ell ran0e =)0) if estimated ell ran0e is 19km then PrahCS" 12 'f all ells in the network are assumed to ha-e same ell ran0e them PrahCS is the same for the whole network
Company Confidential
N CS
PRACH Cyclic S%it or pream#le 1 Prac%CS # 6or pream!le 5, prahCs onfi0uration maps to speial ta!le of yli shifts # 3he hi0hli0hted onfi0uration indies 5G4 are supported in R@19
Company Confidential
Prac%CS an$ rootSe*In$e' # PrahCS defines the n!m#er o cyclic s%its .in terms of num!er of samples used to generate m!ltiple pream#le se*!ences rom a single root se*!ence
# =*ample !ased on PrahCS>12 % num!er of yli shifts" 11F $ Root se+uene len0th is F so a yli shift of 11F samples allows R+U6DD+76 &809://9(; < yli shifts !efore makin0 a omplete rotation .si0natures per root se+uene
# 45 pream!les are transmitted in the PRACH frame) 'f one root is not enou0h to 0enerate all 45 pream!les then more root se+uenes are neessary $ 3o ensure ha-in0 45 pream!le se+uenes within the ell it is neessary to ha-e R+U6DUP &=1:<(; 17 root se+uenes per ell
# 6or pream!le format 5, root se+uene len0th is 1:)
Company Confidential
PRACH Cyclic S%it rootSe*In$e' # RootSe+'nde* points to the first root se+uene to !e used when 0eneratin0 the set of 45 pream!le se+uenes)
# =ah lo0ial rootSe+'nde* is assoiated
=*trat from PP 3S 4)211 3a!le 9):)2)%5 . Pream!le 6ormats 7%) ;appin0 !etween lo0ial and physial root se+uenes) Logical root se*!ence n!m#er
7$2
12F, :17, 157, 4FF, 127, :1F, 217, 42F, 14, 4:1, 5, :99, 179, :5, F, :54, :7, :4F, 47, ::F 2, :, 1,
25$2F
94, :, 112, :2:, 15, 4F1
7$9
7, :9F, 52, :F:, 57, :FF
4$51
9, 75, :, :44, 154, 4F
52$91
1, 7, 2, 11, 7, 7F, 2:, 12, 2F, 17
92$4
25, 19, 5, :F1, 4, ::1, :5, :49, 1:, 441, 14, :7
G)
G))
45$:9
4, :9, :, :41, 5, :F4, F, 77, 27, 1F, 21, 1
17$19
7F, 97, 249, 9:5, 2, 474
14$1F
4:, 5:2, 2F4, 95
27$:
4, 97, 79, 95, :, 544, 27, 99F, 2:F, 947, 51F, 527, 257, 9FF, 29, 91, 22F, 417
with a sin0le physial root se+uene num!er)
# 'n ase more than one root se+uene is neessary the consec!ti5e num!er is seleted until the full set is 0enerated
# N<3=" parameter rootSe+'nde* speifies the lo0ial root se+uene inde*
Reommendation" &se different rootSe+'nde* aross nei0h!ourin0 ells means to ensure nei0h!our ells will use different pream!le se+uenes Company Confidential
P%ysical root se*!ence in$e' &in increasing or$er o t%e correspon$ing logical se*!ence n!m#er(
PRACH Planning 7rap Up Steps3 % Define the prac%ConIn$e' # Depends on pream!le format .ell ran0e # 't should !e the same for eah ell of a site % Define the prac%)re*+ # Depends on the P&CCH re0ion # 't an !e assumed to !e the same for all ells of a network .simplifiation % Define the Prac%CS # Depends on the ell ran0e # 'f for simpliity same ell ran0e is assumed for all network then prahCS is the same for all ells % Define the rootSe*In$e' # 't points to the first root se+uene # 't needs to !e different for nei0h!our ells # rootSe+'nde* separation !etween ells depends on how many are neessary per ell .depends on PrahCS
Company Confidential
PRACH Planning e'ample Ass!mptions3 % prac%ConIn$e';0 for all ells # pream!le format >7 # 9)9km # =ah ell onsumes 5 root se+uenes
Company Confidential
E'ercise
# Plan the PRACH Parameters for the sites attahed in the e*el
# Assumptions" $ P&CCH resoures >4 $ Cell ran0e > 12km .all ells ha-e same ran0e
$ (I"17;HB
PRACH Confg Exercise
Company Confidential
PCI Planning Intro$!ction # 3here are 975 uni+ue Physial Cell 'Ds .PC' P%ysical Layer Cell I$entity ; &0 > 6ID/( 6ID? 6ID/3 Physial @ayer Cell 'dentity 0roup) Ran0e 7 to 14: $ Defines SSS se+uene 6ID?3 'dentity within the 0roup) Ran0e 7 to 2 $ Defines PSS se+uene
# PC' impats the alloation of resoure elements to the referene si0nal and the set of physial hannels Resoure element alloation to the Referene Si0nal
# Alloation pattern repeats e-ery 4th Physial @ayer Cell 'dentity Company Confidential
PCI Planning #
Analo0ous to sram!lin0 ode plannin0 in &;3S
$
;a*imum isolation !etween ells with the same PC' J
#
3o ensure that &= ne-er simultaneously reei-e the same identity from more than a sin0le ell
Physial Cell 'dentity is defined !y the parameter phyCell'D"
#
Parameter
+#@ect
Range
phyCell'd
@NC=@
7 to 97
Dea!lt Not Applia!le
3here should !e some le-el of o%ordination aross international !orders when alloatin0 PC's) $
3his will help to a-oid operators alloatin0 the same identity to ells on the same R6 arrier and in nei0h!ourin0 0eo0raphi areas
Company Confidential
PCI Planning Recommen$ations or ?T' 'd > 7 'd > 2
'n priority or$er, num!er 1 most important in 6DD .all four should !e fulfilled, ideally) Num!er 2 important in 3DDK
nei0h!our ells
'd > 'd > 1
'd > 'd > 9
1) A-oid assi0nin0 the same PC' to
'd > 4
'd > : 'd > F
'd > 11 'd > 5
'd > 17
E'ample / PCI I$entity Plan
? A5oi$ assigning t%e same mo$0 &PCI( to Bneig%#o!r cells ) A-oid assi0nin0 the same mod4.PC' to Lnei0h!ourM ells 5) A-oid assi0nin0 the same mod7 .PC' to Lnei0h!ourM ells E'ample ? PCI I$entity Plan Company Confidential
PCI Planning e'ample uiB" how to alloate PC' for a sin0le%ell 3D%@3= omni site to !e plaed at the red irleO
Company Confidential
UL Reerence Signal Planning % &lse+Hop % &lrpHop % 0rpAssi0P&SCH % ulRsCs % Se+uene roup Num!er .u
Company Confidential
UL Reerence Signal +5er5ie" Types of &@ Referene Si0nals
# Demodulation Referene Si0nals .D; RS $ P&SCH/P&CCH $ata estimation
# Soundin0 Referene Si0nals .SRS $ ;ainly &@ c%annel estimation &@ .R@57/R@19
D; RS is c%aracterise$ !y"
# Se+uene .adoff Chu odes # Se+uene len0th" e+ual to the Q of su!arriers used for P&SCH transmission
# Se+uene 0roup" J 7 options J Cell speifi parameter
# Cyli Shift" &= and ell speifi parameter Company Confidential
&@ D; RS alloation per slot for Normal Cyli Prefi*
UL DM Reerence Signal 6ee$ or Planning Iss!e3 # D; RS oupy always the same slot in time domain
# 'n fre+ueny domain D; RS of a 0i-en &= oupies the same PR(s as its P&SCH/P&CCH data transmission
# Possi!le inter ell interferene for RS due to simultaneous &@ alloations on nei0h!our ells
$ No intra ell interferene !eause users are separated in fre+ueny
$ Possi!le inter ell interferene
Scope o planning3 # DM RS in co-site$ cells nee$s to #e $ierent Company Confidential
&@ D; RS alloation per slot for Normal Cyli Prefi*
RS Se*!ences an$ RS Se*!ence ro!ps Se*!ence ro!p I$, B! # RS se+uenes for P&SCH ha-e different len0ths dependin0 the &@ !andwidth alloated for a &=
# 7 possi!le se+uenes for eah PR( alloation len0th of 1%177 PR(s # Se+uenes are 0rouped into 7 0roups so they an !e assi0ned to ells # Se+uene 0roup num!er L!M" 0rpAssi0P&SCH" 0roup u
Company Confidential
=
( PCI + grpAssigPUSCH ) mod 30
assi0nment for P&SCH Ran0e 7G2FE, step 1
Cyclic S%it #
Additional se+uenes an !e deri-ed from a !asi se+uene !y applyin0 a yli shift
#
Cyli shifts of an e*tended C se+uene are not fully ortho0onal, !ut ha-e low ross%orrelation
#
3he atual &@ referene si0nal yli shift ncs used !y &= is different for e-ery 7)9ms time slot ncs
=
[n
Cell%speifi stati yli shift defined !y LNCEL/ ulRsCs and !roadast on (CCH
(1) DMRS +
( 2)
nDMRS
+
nPRS ( ns )
33'%speifi yli shift si0nalled to &= on PDCCH DC'7 in eah uplink shedulin0 0rant .defined !y sheduler
] mod12 Pseudorandom yli shift offset that han0es e5ery time slot Depends on the PC', slot num!er ns and u -ia @NC=@/grpAssigPUSCH
cinit
Company Confidential
cell N ID = ⋅ 32 + 30
u
UL DM Reerence Signal Hopping Tec%ni*!es # Se*!ence Hopping $ 'ntra%Su!frame hoppin0 !etween two se+uenes within a se+uene 0roup f or alloations lar0er than 9PR(s
$
$ 'f 0roup -aries !etween slots" roup hoppin0 $ roup Hoppin0 not ena!led in R@17/R@27/R@7" UlGrpHop = false J roup is the same for all slots
# Cyclic S%it Hopping $ Always used $ Cell speifi yli shift added on top of &= speifi yli shift Company Confidential
Planning )rom T%eory to Practice &/:?( T%eory3 # 't should !e possi!le to assi0n to the cells o one site the same se*!ence gro!p B! and LdifferentiateM the se+uenes usin0 different ell speifi yli shifts i)e) alloatin0 $ierent !lRsCs
Remem!erK" Cyli shifts of an e*tended C se+uene are not fully ortho0onal, !ut ha-e low ross%orrelation Company Confidential
Planning )rom T%eory to Practice &?:?( Practice3 # 't doesnMt seem to work # &@ 3hrou0hput 0ets onsidera!ly affeted if &@ traffi in nei0h!our ell $ 6rom 57 ;!ps to 22 ;!ps in the e*ample
Company Confidential
Planning 6e" r!le # Alloate $ierent se*!ence gro!p u or e5ery cell, incl!$ing cells o t%e same site $ Cross%orrelation properties !etween se+uenes from two different 0roups are 0ood !eause of se+uene 0roupin0 in the PP spe
# ulRsCs does not matter .it is only rele-ant for se+uenes within one se+ 0roup u
Company Confidential
Planning Res!lts # &@ 3hrou0hput still suffers from &@ interferene in nei0h!our ell !ut the effet is lower
Company Confidential
Pros an cons o t%e ne" planning r!le # FG3 Results seem to !e !etter # FG3 @ess parameters to plan, only PC' plannin0 needed $ UlRsCs only rele-ant when usin0 se+uenes of the same 0roup $ LuM will !e different if PC' module rule is followed) 'n that ase LgrpAssigPUSCHM -alue is not rele-ant u
= ( PCI + grpAssigPUSCH) mod30
# F -G3 Redued 0roup reuse distane ompared to the ase of assi0nin0 the same 0roup per eah site
Company Confidential
UL DM RS Planning 7rap !p # Principle3 DM RS nee$s to #e $ierent in cells o t%e same e6o$e # Current plannin0 approah" $ Assign $ierent se*!ence gro!p n!m#er B! to t%e cells o t%e same site) Ran0e" 7G2FE) grpAssigPUSCH an !e onstant >no need for plannin0 u
=
( PCI + grpAssigPUSCH ) mod 30
$ 'f ells of the site follow the PC'mod rule, the se+uene 0roup num!er LuM will !e different $ 'f PC'mod rule is not followed, hek PC'mod7 rule J 'f pro!lems use grpAssigPUSCH to differentiate the LuM % se+uene 0roup num!er% $ 'f same LuM has to !e used in nei0h!ourin0 ells and annot !e han0ed usin0 0rpAssi0P&SCH then assi0n different ulRsCs to the ells of a site) Ran0e 7G:E
Company Confidential
UL DM RS Planning e'ample, PCI-#ase$
Company Confidential
