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


Ratings: (0)|Views: 121 |Likes:
Published by Vinay Kalimi

More info:

Published by: Vinay Kalimi on Aug 25, 2011
Copyright:Attribution Non-commercial


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





 —Mobile broadband usage is taking off, demandingimproved services and increased capacity of mobile networks. Tomeet these requirements, 3GPP has defined LTE (the 3GPP LongTerm Evolution). This paper presents some key characteristics of the LTE radio interface, including physical layer and radio re-source management functions, and evaluates their impact onsystem performance. As compared to a reference system withmore basic characteristics, represented by Mobile WiMAX, re-sults point to a combined gain in spectrum efficiency of 60% indownlink and 100% in uplink. Cell-edge bitrate gains are about100% in both downlink and uplink. A closer analysis of the indi-vidual system characteristics indicates that these performancedifferences are due to rather uniform contributions from a set of distinctive features.
 Index Terms
 —LTE, Performance, WiMAX
INTRODUCTIONsage of mobile broadband services, supported by the in-troduction of High Speed Packet Access (HSPA), is tak-ing off. To meet the future increased demand for such servic-es, corresponding improvements in the supply of services arerequired, including higher bit rates, lower delays, and higher capacity. This is the target of 3GPP’s two radio access net-works HSPA and LTE [1], of which the latter is the focus of this paper. LTE brings unprecedented performance. Examplesinclude peak data rates exceeding 300Mbps, delays below10ms, and manifold spectrum efficiency gains over early 3Gsystem releases. Further, LTE can be deployed in new andexisting frequency bands, has a flat architecture with fewnodes, and facilitates simple operation and maintenance.While targeting a smooth evolution from legacy 3GPP and3GPP2 systems, LTE also constitutes a major step towardsIMT-Advanced systems. In fact, LTE includes many of thefeatures originally considered for future fourth generation sys-tem.General LTE concept descriptions are available in [1]. Inthis paper, the focus is on key characteristics of the LTE radiointerface. A set of such key characteristics are both qualita-tively discussed and quantitatively evaluated in terms of downlink and uplink user data rates and spectrum efficiencygenerated by means of system level simulations. For refer-ence, the LTE characteristics are compared to more conven-tional solutions. These are represented by corresponding func-tionalities in Mobile WiMAX with Partial Usage of Sub-Channels (PUSC) [2].The paper is outlined as follows: After an introduction tothe basic structure of the LTE radio interface in Section II,Section III provides a qualitative discussion of distinctive fea-tures of the evaluated system concepts, and their impact on performance. Models and assumptions are summarized in Sec-tion IV, followed by numerical results in Section V. Finally, asummary is provided in Section 6.II.
 Comprehensive descriptions of the LTE radio interface areavailable in [1]. In short, LTE is based on Orthogonal Fre-quency Domain Multiplexing (OFDM). The numerology in-cludes a subcarrier spacing of 15kHz, support for bandwidthsup to 20MHz, and resource allocation granularity of 180kHz x1ms (a so-called resource block pair). In the uplink, a pre-coder is used to limit peak-to-average power ratios, and there- by reduce terminal complexity. Based on channel quality,modulation (up to 64QAM) and channel coding rates are dy-namically selected. Both FDD, TDD, and half duplex FDD aresupported. A variety of antenna concepts targeting differentscenarios is included: transmit diversity for improved robust-ness of control channels, beamforming for improved channelquality in general, and multi-stream (MIMO) transmission for improved data rates in scenarios with good channel quality.On the MAC layer, dynamic scheduling is done on a resource block pair basis, based on QoS parameters and channel quali-ty. Retransmissions are handled with two loops, a fast inner loop taking care of most errors complemented with a veryrobust outer loop for residual errors.III.
Some of the more fundamental features discussed in the previous section are not unique to LTE. E.g. OFDM, multi-antenna transmission, and adaptive modulation and coding arestandard techniques used by many systems. On a more de-tailed level however, LTE distinguishes itself by using moresophisticated solutions than other systems. A list of such cha-racteristics is presented in Table I. For reference, the corres- ponding solutions used in more basic systems are also listed.This is represented by Mobile WiMAX Wave 2. It should benoted that there are several other features differing betweenthese systems which are not listed, e.g. control signaling ro- bustness, higher layer overhead, and mobility aspects.
The LTE Radio Interface – Key Characteristicsand Performance
Anders Furuskär, Tomas Jönsson, and Magnus LundevallEricsson Research, Sweden
978-1-4244-2644-7/08/$25.00 © 2008 IEEE
 Models and assumptions are aligned with the NGMN re-commendations in [3]. Table II contains a brief summary. Theevaluation methodology is based on time-dynamic, multi-cellsystem simulations.V.
 This section presents downlink and uplink user throughputand spectrum efficiency for a selection of system configura-tions and scenarios. More specifically, the following subsec-tions cover (A) baseline configurations with 2x2 and 1x2 an-tenna configurations, (B) more advanced multi-antenna confi-gurations, and (C) results for file transfer (non-full buffer)traffic models.
 LTE and Mobile WiMAX – Baseline Scenario
Downlink user throughput and spectrum efficiency figuresfor LTE FDD, LTE TDD, and Mobile WiMAX are summa-rized in Fig. 1. Note that in this special case, as there are 10full-buffer users per sector in average, and the spectrum allo-cation is 10MHz, the spectrum efficiency, measured in bps/Hz/sector, and the average user throughput, measured inMbps, are the same. For the TDD systems, the spectrum effi-ciency is calculated by down-scaling the denominator (system bandwidth) with the relative time utilization in the direction inquestion (measured in data symbols). Distributions of user throughput normalized with spectrum allocation and TDDutilization are also presented.It is seen that LTE is some 60% better than Mobile WiMAXin the average metrics, and about a factor two better in cell-edge performance. The reasons for these differences are acombination of the distinctive features presented in Table I.
(slogan used in Fig. 1-2)
 LTE Mobile WiMAX wave 2 Performance impactMultiple access
 OFDM in DL,DFT-spread OFDM in ULOFDM in DL and UL DFT-spread OFDM reduces the peak-to-average power ratio and reduces terminal complexity,requires one-tap equalizer in base station receiver Uplink power control
 Fractional pathloss compensa-tionFull pathloss compensation Fractional pathloss compensation enables flexibletrade off between average and cell-edge data ratesScheduling
 Channel dependent in timeand frequency domainChannel dependent in timedomainAccess to the frequency domain yields larger scheduling gainsMIMO scheme
 Horizontal encoding (multiplecodewords), closed loop with precodingVertical encoding (singlecodeword)Horizontal encoding enables per-stream link adaptation and successive interference cancella-tion (SIC) receiversModulation and cod-ing scheme granulari-ty
 Fine granularity(1-2dB apart)Coarse granularity(2-3dB apart)Finer granularity enables better link adaptation precision
Hybrid ARQ II
 Incremental redundancy Chase combining Incremental redundancy is more efficient(lower SNR required for given error rate)Frame duration
(CQI delay)
 1ms subframes 5ms frames Shorter subframes yield lower user planedelay and reduced channel quality feedback delaysOverhead / controlchannel efficiency
(OH / CCH eff)
 Relatively low OH (whilecontrol channels arerobust)Relatively high OH Lower overhead improves performance
 Parameter ValueTraffic Model a) Full buffer (10 users per sector) or  b) File transfer (100KB fixed file size) withvariable loadUser location Uniform distributionSite-to-site distance 500mCarrier frequency 2.0GHzCarrier bandwidth 10MHzDistance-dependent pathlossL = I + 37.6·log
(R) + P, R in km, I = 128.1 for 2GHz, P = 20dB penetration lossLognormal shadowing8dB std dev, 50m correlation distance, 0.5 corre-lation between sitesChannel model3GPP SCM, Urban Macro High Spread (15 deg),extended to 10MHzTerminal speed 3km/hBS / Terminal power 46dBm / 23dBmAntenna configura-tionsBS: 2-4 transmit and receiveTerminal: 1 transmit, 2-4 receiveScheduler LTE: DL: Proportional fair in time and fre-quency, UL: Quality-based FDMWiMAX: DL: Proportional fair in time domain,UL: FDMMIMOLTE: Codebook-based pre-coded adaptive rank MIMOWiMAX: Dynamic switching between spatialmultiplexing MIMO and STCPower controlLTE: Open loop with fractional pathloss com- pensation (
=0.8), SNR target 10dB at cell edgeWiMAX: Open loop, SNR target 15dB (full pathloss compensation)Receiver type LTE: MMSE with SIC in DLWiMAX: MMSETDD asymmetry LTE: 4:3, WiMAX: 22:15
 The individual impact of each such feature has been assessed by, in the simulations, replacing the LTE functionality withthe corresponding WiMAX functionality. The result is shownin the lower bar graph in Fig. 1. The percentage figure to theleft represents the individual feature impact, and the percen-tage figure to the right the accumulated impact of the featurescombined. It is seen that the total difference is not due to asingle distinctive feature, but rather a combination of distinc-tive features, headed by frequency domain scheduling, faster channel quality feedback, and control channel efficiency. Notealso that when all distinctive features are replaced, the perfor-mance is the same, confirming that these features are indeedthe reason for the overall difference in performance. A similar analysis can be made for the cell-edge metric.Similar results for the uplink are summarized in Fig. 2. Inthis direction, it is seen that LTE is more than a factor two better than Mobile WiMAX in both average and cell-edgemetrics. Also here, the distinctive features jointly make up the
   A  v  g  c  e   l   l   t  p   [   b  p  s   /   H  z   /  c  e   l   l   ]
   C  e   l   l  -  e   d  g  e  u  s  e  r   t  p   [   b  p  s   /   H  z   ]
0.0520.0500.028 User Throughput [bps/Hz]
   C .   D .   F .   [   %   ]
LTE FDD mc/mu/ce 1.73/0.173/0.052 bps/HzLTE TDD mc/mu/ce 1.70/0.170/0.050 bps/HzWiMAX TDD mc/mu/ce 1.06/0.106/0.028 bps/Hz
00.511.5WiMAX PUSCLTE WiMAX-likeHARQMCSMIMO: vertical/no SICMIMO: precodingOH / CCH effCQI delaySchedulingLTEAvg cell throughput [bps/Hz/cell]Downlink 1.06-39%1.03-40%1.03-1%-40%1.05-2%-40%1.06-2%-39%1.09-4%-37%1.13-11%-35%1.28-13%-26%1.47-15%-15%1.730% 0%
Fig. 1. Summary of baseline downlink normalized user throughput and spec-trum efficiency results, and feature analysis.
   A  v  g  c  e   l   l   t  p   [   b  p  s   /   H  z   /  c  e   l   l   ]
   C  e   l   l  -  e   d  g  e  u  s  e  r   t  p   [   b  p  s   /   H  z   ]
0.0490.0450.018 User Throughput [bps/Hz]
   C .   D .   F .   [   %   ]
LTE FDD mc/mu/ce 1.05/0.105/0.049 bps/HzLTE TDD mc/mu/ce 0.98/0.098/0.045 bps/HzWiMAX TDD mc/mu/ce 0.43/0.043/0.018 bps/Hz PUSCLTE WiMAX-likeMA (OFDM)HARQMCSSchedulingPCCQI delayOH / CCH effLTEAvg cell throughput [bps/Hz/cell]Uplink 0.43-59%0.44-58%0.42-1%-60%0.43-1%-59%0.43-6%-59%0.46-9%-56%0.50-11%-52%0.56-14%-46%0.66-37%-37%1.050%0%
Fig. 2. Summary of baseline uplink normalized user throughput and spectrumefficiency results, and feature analysis.

Activity (6)

You've already reviewed this. Edit your review.
1 hundred reads
davi mart marte liked this
yb4in liked this
Podg liked this
eutopia2211 liked this
eutopia2211 liked this

You're Reading a Free Preview

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