Communications and Network, 2013, 5, 1-5
http://dx.doi.org/10.4236/cn.2013.53B2001 Published Online September 2013 (http://www.scirp.org/journal/cn)
Downlink Scheduling and Rate Capping for
LTE-Advanced Carrier Aggregation
Mieszko Chmiel, Jin Shi, David X. Zhou
LTE System Design & Architectu r e, Nokia Siemens Networks, Beijing, China
Received April, 2013
ABSTRACT
Long Term Evolution (LTE) Carrier Aggregation (CA) was introduced by the Release-10 3GPP specifications. CA al-
lows aggregation of up to 5 cells for a terminal; both downlink (DL) CA and uplink (UL) CA are supported by the
3GPP specifications. However, the first commercial deployments focus on the aggregation of two cells in the downlink.
The benefits of LTE CA are increased terminal peak data rates, aggregation of fragmented spectrum and fast load ba-
lanci ng. In t his pa per, we analyze different strategies of DL scheduling for LTE CA including centralized, independent
and distributed schedulers, we provide the corresponding simulation results considering UE data rate limitations and
different traffic models. Also, we compare the performance of a single LTE carrier with LTE CA using the same total
bandwidth.
Keywords: Long Term Evolu tion; Carrier Aggregation; Scheduling; Downlink; Bandwidth
1. Introduction
Carrier Aggregation is one of the Long Term Evolution
Advanced features introduced by 3GPP in order to meet
IMT-Advanced requirements of peak data rates of up to 1
Gbit/s in the DL a nd 500 Mbit/s in the UL [1-3]. In addi-
tion to the User Equipment (UE) p eak data rate increase,
another benefit of CA is the possibility for operators to
aggregate fragmented spectrum. Also fast load balancing
can be achieved with LTE-Advanced CA because of a
UE with a ggre gated cells; the traffic can be scheduled on
any of the aggregated cells on a Transmission Time In-
terval (TTI) basis.
The overview of LTE-Advanced CA is given in [4]
and [5] while the CA impact on Radio Resource Man-
agement algorithms is presented in [6]. In [7], perfor-
mance results with high number of DL aggregated cells
are provided; furthermore, UL C A simulations results are
reported in [8]. However, in this paper we focus on the
aggregation of two DL cells since this is the first com-
mercial deployment scenario for CA.
The performance of CA is highly dependent on the
scheduling method used by the eNode B (eNB). The fol-
lowing three general scheduling principles can be used
for CA.
• One centralized scheduler for all aggregated cells.
• Independent schedulers per aggregated cell [9, 10].
• Distributed and coordinated schedulers per cell [9,
10].
In this paper, we compare the above principles taking
into account real-life effects such as traffic models and
UE data rate limitations. Also, the performance compar-
ison of DL CA with a single carrier of the same band-
width (BW) is analyzed.
The paper is organized as follows. We discuss strate-
gies for DL scheduling in section 2. Section 3 outlines
simulation assumptions. In section 4, we provide the si-
mulation results. Finally, some conclusions are given in
section 5.
2. CA Scheduling and Rate Capping
Methods
One centralized scheduler serving CA UEs and non-CA
UEs of all aggregated cells can potentially offer the op-
timum performance. The frequency diversity over all
aggregated cells can be exploited in scheduling of CA
UEs. However, the challenge of this centralized schedul-
ing method is the implementation complexity increased
with the number of aggregated cells. In addition to lack
of scalability, this method might be not feasible in future
inter-eNB carrier aggregation scenarios.
Independent schedulers per aggregated cell represent
the simple and scalable extension of single carrier sche-
duling. This option is expected to have worse perfor-
mance compared to the centralized scheduling principle
because the frequency diversity is exploited separately
within each cell. Furthermore, the fairness between UEs

M. CHMIEL ET AL.
2
can only be achieved and controlled on a cell basis;
therefore, this so lution is capa ble of neither achi eving nor
controlling throughput fairness between CA UEs and
non-CA UEs. Also, it shall be noted that in fact the
scheduling for CA cannot be fully independent per cell
because there are UE data rate limits which shall not be
exceeded when allocating resource s on multip le cells to a
CA UE. Such data rate limits are, for example, the 3GPP
defined peak data rate of a given UE category [11] or the
amount of UE data available for transmission in the buf-
fer.
The distributed and coordinated schedulers per cell ca n
achieve better performance for Carrier Aggregation
compared to independent schedulers [10], the reason
being that distributed schedulers can exploit frequency
diver sity o ver all aggre gated cel ls in a si milar way a s the
centralized scheduler. In this solution, each cell has its
own scheduler; however, as opposed to the independent
schedulers, the coordinated schedulers in aggregated
cells communicate with each other for the purpose of
optimizing scheduling metric calculation. In [9], it is
shown that distributed and coordinated schedulers are
optimal fro m the utility maxi mizatio n po int o f vie w. This
scheduling method can use the same or similar schedul-
ing metric calculation as the centralized scheduling with
the difference that the computation is distributed. The
performance of distributed and coordinated schedulers
for CA is on a par with centralized scheduling for
full-buffer traffic and without considering UE data rate
limits. However, if real-life effects like non-full-buffer
traffic and finite UE data rate limits (e.g. the peak data
rate) kick in, the performance of distributed and coordi-
nated scheduling depends also on the rate capping me-
thod used to fulfill the CA UE da ta rate limits.
In this paper, we consider two methods for rate cap-
ping for CA UEs:
1) Static 50/50: the amount of data in the buffer and
the peak data rate are divided equally to active serving
cells.
2) Dynamic: the amount of data in the buffer and the
peak data rate are divided to active serving cells propor-
tionally to the UE throughput achieved on each of the
active cells. Additio nally, t he d ivision of da ta in the buf-
fer might be adjusted if all data assigned to a given cell is
drained in a TTI.
Another relevant topic is the performance comparison
of distributed and coordinated CA scheduling with the
performance of single-carrier scheduling in the same
bandwidth. This comparison is impacted by higher pro-
tocol overhead of CA because separate T ranspor t Blocks
(TBs) are generated per each scheduled cell. On the other
hand, a single cell of a b andwidth equal to the sum o f the
bandwidths of the aggregated cells will have a worse
Channel State Information (CSI) and Resource Block
Group (RBG) granularity.
3. Simulation Assumptions
A hexagonal regular cell layout in an urban deployment
scenario with 500 m Inter-Site Di stance ( ISD) wa s simu-
lated with frequency reuse 1. The deployment area com-
prises 21 cells placed in a wrap-around model assuming a
Typical Urban (TU) channel model. A pathloss model for
small ce lls with P L slope o f 37.6 dB per decade was used.
Additional penetration loss of 20 dB for indoor coverage
was taken into consideration [12]. Basic configuration
parameters such as the pathloss model and antenna dia-
gram were selected in accordance to [12].
The number of users within the simulation area was
kept constant. Slow-moving subscribers were assumed.
During the simulation run, a UE can change its serving
cell by handover based on measurements (ha ndover mar-
gin 3 dB). The simulation model includes non-adaptive
Hybrid Automatic Repeat Request (HARQ) with Chase
Combining. The essential simulation parameters are
listed in Table 1.
4. Simulation Results
4.1. Carrier Aggregation and Single Ce ll without
Physical Downlink Control Channel
In this section, the performances of downlink intra-band
CA and single-carrier operation are analyzed without
Physical Downlink Control Channel (PDCCH) overhead.
To evaluate the cell throughput performance, the same
number of UEs (12) per cell scheduler will be set with
full-buffer traffic. Other simulation parameters are listed
in Table 2.
Table 1. Parameters of system simulation model.
Parameters Settings
Wrap around la yo ut
7 sites with 3 cells/site
Propagation scenario Macro 1 (ISD 500 m) [12]
Carriers frequency 1 Int r a-band: 2 G H z
Carriers frequency 2
Inter -band: 850 MH z and 2 G Hz
System bandwidth
CA: 2*10 MHz
Fast fading model According to [13]
Indoor penetration loss 20 dB (according to [12])
Traffic mod el 1 Full buffer [14]
Traffic mod el 2 Constant Bi t Rate
UE receiver 2 RX (maximum ratio combining)
UE speed 3 km/h
Scheduler Proportion al Fair
eNodeB pow er o f cell 40 W
Transmission mode Closed loop MIMO, 2TX
CQI reporting mod e Mode-3
Block Error Rate target 10%

M. CHMIEL ET AL.
3
Figure 1 shows the average cell throughput norma-
lized to 10 MHz bandwidth.
CA of two 10 MHz cells uses s maller Resource Block
Group (RBG) and Channel Quality Indicator (CQI) gra-
nularity compared to a 20 MHz cell. From the simulation
result we see that the modified 20 MHz simulation has
+3.54% higher cell throughput compared to normal 20
MHz s imulatio n with worse granular it y.
Considering the protocol overhead of additional tran-
sport blocks and no frequency diversity exploration
across a ggregated ce lls, the CA with independent scheduler
has -2.32% loss on cell throughput compared to a single
20 MHz cell.
The CA with distributed scheduler is capable to have
inter-scheduler communications. It recovers some of the
freq uency dive rsit y gai n from larger b andwidt h. The d is-
tributed scheduler brings +2.93% higher average cell
throughput compared to independent sc hedulers.
Fro m simula tio n wit hout PDCCH, CA wit h distributed
schedulers using two aggregated 10 MHz cells can
achieve performance similar to a si ngle 20 MHz cell.
Table 2. Settings for CA and single carrier simulations
without PDCCH.
Parameters Single
20MHz
Modified
Single
20MHza
CA
2x10MHz
Optio n 1
CA
2x10MHz
Optio n 2
Num ber of cells 21 21 42 42
Number of UEs
per all cells 252 non-CA UEs 252 non-CA UEs 126 CA UEs 126 CA UEs
RBG size 4 PRBs 3 P RBs 3 PRBs 3 PRBs
CQI resolution 4 PRBs 3 P RBs 3 PRBs 3 PRBs
Scheduler Centralized Centralized Independent Distributed
PDCCH Disabled Disabled Disabled Di sabled
aThe RBG and CQI resolut ion gr anularity is increased in simulation, but not
possible by the 3 GP P spec ifica t ion according to [15].
Figure 1. Cell throug hput of CA and single carrier w ithout
PDCCH.
4.2. Carrier Aggregation and Single Cell with
Load-Adaptive PDCCH
In this section, the performances of downlink intra-band
CA and a single carrier are analyzed with the modeling
of load-adaptive PDCCH. To focus on UE throughput,
the same number of UEs (126) in simulation area will be
set with the full buffer traffic model. Other simulation
parameters are listed in Table 3.
Figure 2 shows the average and 5%-ile of UE thro ugh-
put.
With the PDCCH considered, the performance gap
bet wee n C A a nd a s ingle carrier b ecomes larger. There is
-6.98% loss on average UE throughput and -11.39% on
5%-ile UE throughput.
Figure 3 shows the utilization of PDCCH s ymbols and
the utilization of Control Cha nnel Elements (CCEs).
In CA, scheduling of the additional bandwidth requires
additional PDCCH assignments. The higher number of
orthogonal frequency-division multiplexing (OFDM)
symbols for PDCCH reduce s the number of OFDM
symbols available for data transmission.
Table 3 . Settings for CA and single car rier si mul ati ons w i th
loa d adaptive PDCCH.
Parameters Single
20 MHz Modified Single 20 MHza CA 2x10 MHz Optio n 2
Num ber of cells 21 21 42
Number of UE s
per all cells 126 non-CA UEs
126 no n-CA UEs 126 CA UEs
RBG size 4 PRBs 3 PRBs 3 PRBs
CQI resolution 4 PRBs 3 P RBs 3 PRBs
Scheduler Centralized Centralized Distributed
PDCCH Adaptive Ada ptive Adaptive
aThe RBG and CQI resolut ion gr anularity is increased in simulation, but not
possible by the 3 GP P spec ifica t ion according to [15].
Figure 2. UE throughput of CA and single carrier with load-
adaptive PDCCH.

M. CHMIEL ET AL.
4
4.3. Rate Capping on UE Buffer
The problem of rate limitation of CA UEs is described in
section 2. There are several solutions to rate capping due
to the UE buffer, which are investigated in this section;
rate capping due to the peak data rate is investigated in
section 4.4. Table 4 lists the solutions to rate capping
due to the UE RLC buffer which are compared in our
simulation
The simulation assumption for RLC buffer rate cap-
ping simulations can be found in Table 1 with the second
traffic model – Constant Bit Rate (CBR) and inter-band
CA. Each user has a 1Mb/s CBR service. The simulation
resul ts are shown in Figure 4.
The simulation results are analyzed in terms of number
of allocated Physical Resource Blocks (PRBs) per cell
for the CBR service. From the results it can be seen that
the ideal mode is the most efficient method for each
number users. T he dynamic mode is superio r to the static
mode with 50%-50% split.
4.4. Rate Capping on Peak Data Rate
Table 5 shows the solutions to rate capping due to the
UE peak data rate which are co mpared in our simulation.
The peak data rate of the UE in our simulations is
51.024 Mbps which is based on UE category 2 according
to [11].
Figure 3. PDCCH symbols of CA and single carrier.
Table 4. Solutions to RLC buffer rate capping.
Solution Description
Ide a l C lose to genie-aided
Static 50% of the RLC buffer allocated to the PCell and
remaining to the SCell statically
Dyna m ic X% RLC buffer allocated for PCell and (100-X)%
for SCell dynamically
The simulation assumption for rate capping due to the
UE peak data rate can be seen in Table 1 with the first
traffic model – Full Buffer and inter-band CA. Figure 5
shows the simu l atio n results.
The simulation results are analyzed in terms of aver-
age user throughput for the full-buffer service. From the
results it can be seen that the difference between different
mode s in hi gher numb er o f user s is ver y sma ll. H o wever,
in case of a very low number of users such as 1 or 2, the
ideal mode is superior to other modes while the d ynamic
mode is slightly better than the static mode with 50%-
50% split.
0
2
4
6
8
10
12
14
16
18
2
3
4
5
6
Num of Allocated PRB Per Cell
CL-MIMO, CBR, rate capping-ideal mode
CL-MIMO, CBR, rate capping-dynamic mode
CL-MIMO, CBR, rate capping-static 0.5 mode
Total number of users
Figure 4. Simulation re sults of RLC buffer rate caping.
Table 5. So lutions to pe a k data r ate cap pi ng .
Solution Description
Ide a l C lose to genie-aided
Static 50% of UE Peak Data Rate allocated to the PCell and
remainin g to the SCell s tat ic ally
Dyna m ic X% UE Peak Data Rate allocat ed for PC ell and
(100-X) % for SC ell dynamicall y
0
5000
10000
15000
20000
25000
30000
1
2
3
4
5
Average User Througphut (kbps)
CL-MIMO, Full Buffer, rate capping-ideal mode
CL-MIMO, Full Buffer, rate capping-dynamic mode
CL-MIMO, Full Buffer, rate capping-static mode 0.5
Total number of users
Figure 5 . Simulatio n Resul ts of R ate Cap ping on P eak Dat a
Rate.

M. CHMIEL ET AL.
5
5. Conclusions
In this paper, we discussed and simulated rate capping
methods for LTE-A DL Carrier Aggregation scheduling.
Such rate capping is required due to non -full-buffer traf-
fic and other practical UE data rate limits when indepen-
dent or distributed /coord inated scheduling is us ed fo r C A.
Also, we provided simulation results comparing the per-
formance of CA s cheduli ng methods with a singl e cell .
Base d o n the a nal ysis a nd t he simul at ions i n thi s p ape r,
we draw the following conclusions. Aggregated cells can
have spectral efficiency similar to a single cell of the
same bandwidth assuming the same number of full- buf-
fer UEs per cell. Distributed and coordinated schedulers
provide better performance for DL CA compared to in-
dependent schedulers. The analyzed rate capping solu-
tions provide good performance; the dynamic solution
outperforms the static solutio n.
6. Acknowledgements
The authors of this paper would like to thank Hans
Kroener and Harald Steinhaus for their valuable com-
ments and help.
REFERENCES
[1] S. Parkvall, E. Dahlman, A. Furuskar, A. Kangas, M.
Olsson, S. Wanstedt and K. Zangi, “LTE-Advanced -
Evolving LTE towards IMT-Advanced,” IEEE 68th Ve-
hicular Technology Conference VTC 2008-Fall, 21-24
September 2008, pp. 1-5. 10.1109/VETECF.2008.313
[2] S. Parkvall, A. Furuskar and E. Dahlman, “Evolution of
LTE toward IMT-advanced,” IEEE Communications
Magazine, Vol. 49, February 2011, pp. 84-91.
doi:10.1109/MCOM.2011.5706315
[3] S. Parkvall, E. Englund, A. Furuskar, E. Dahlman, T.
Jönsson and A. Paravati, “LTE Evolution towards IMT-
advanced and Commercial Network Performance,” IEEE
International Conference on Communication Systems
(ICCS), 17-19 November 2010, pp. 151-155.
doi:10.1109/ICCS.2010.5686370
[4] R Ratasuk, D. Tolli and A. Ghosh, “Carrier Aggregation
in LTE-Advanced,” IEEE 71st Vehicular Technology
Conference (VTC 2010-Spring), 16-19 May 2010, pp.
1-5. doi:10.1109/VETECS.2010.5493902
[5] Z. K. Shen; A. Papasakellariou, J. Montojo, D. Gersten-
berger and F. L. Xu, “Overview of 3GPP LTE-advanced
Carrier Aggregation for 4G Wireless Communications,”
IEEE Communications Magazine, Vol. 50, Feruary 2012,
pp. 122-130. doi:10.1109/MCOM.2012.6146491
[6] K. I. Pedersen, F. Freder iksen, C. Rosa, H. Nguyen, L. G.
U. Garcia and Y. Y. Wang, “Carrier Aggregation for
LTE-advanced: Functionality and Performance Aspects,”
IEEE Communications Magazine, Vol. 49, 2011, pp.
89-95. doi:10.1109/MCOM.2011.5783991
[7] I. S hayea, M. Ismail an d R. Nordin, “Cap acity Evaluation
of Carrier Aggregation Techniques in LTE-Advanced
system,” Int ernational Conference on Computer and
Communication Engineering (ICCCE), 2012, pp.
99-103. doi:10.1109/ICCCE.2012.6271160
[8] H. Wang, C. Rosa and K. Pedersen, “Performance of
Uplink Carrier Aggregation in LTE-Advanced Systems,”
IEEE 72nd Vehicular Technology Conference Fall (VTC
2010-Fall), 6-9 Septe mber 2010, pp. 1-5.
doi:10.1109/VETECF.2010.5594543
[9] Y. Y. Wang, K. I. Pedersen, T. B. Sorensen, P. E. Mo-
gensen, “Utility Maximization in LTE-Advanced S yste ms
with Carrier Aggregation,” IEEE 73rd Vehicular Tech-
nology Conference (VTC Spring), 15-18 May 2011, pp.
1-5. doi:10.1109/VETECS.2011.5956494
[10] Y. Y. Wang, K. I. Pedersen, T. B. Sorensen and P. E.
Mogensen, “Carrier Load Balancing and Packet Schedul-
ing for Multi-carrier Systems,” IEEE Transactions on
Wireless Communications, Vol. 9, May 2010, pp.
1780-1789. doi:10.1109/10.1109/TWC.2010.05.091310
[11] 3GPP TS 36.306 v10.9.0 “Evolved Universal Terrestrial
Radio Access (E-UTRA); User Equipment (UE) radio
access cap abilities (Release 1 0),” TSG RAN.
[12] 3GPP TR 25.814 v7.1.0, “Physical Layer Aspects for
Evolved Universal Terrestrial Radio Access (UTRA)
(Release 7),” TSG RAN.
[13] ETSI TR 101 112 v3.2.0, “Procedures for the Choice of
Radio Transmission Technologies of the UMTS (UMTS
30.03 version 3.2.0),” SMG.
[14] 3GPP R1-070674, “LTE Physical Layer Frame work for
Performance Verification,” Orange, China Mobile, KPN,
NTT DoCoMo, Sprint, T-Mobile, Vodafone, Telecom
Italia, 3GPP TSG RAN1#48.
[15] 3GPP TS 36.213 v10.9.0 “Evolved Universal Terrestrial
Radio Access (E-UTRA); Physical layer procedures (Re-
lease 10 ),” TSG RAN.