Content uploaded by Vanlin Sathya
Author content
All content in this area was uploaded by Vanlin Sathya on Aug 25, 2018
Content may be subject to copyright.
Enhanced Distributed Resource Allocation and
Interference Management in LTE Femtocell
Networks
Vanlin Sathya, Harsha Vardhan Gudivada, Hemanth Narayanam, Bala Murali Krishna and Bheemarjuna Reddy Tamma
Department of Computer Science and Engineering
Indian Institute of Technology Hyderabad, India
Email: [cs11p1003, cs09b011, cs09b024, bala, tbr]@iith.ac.in
Abstract—Femto cells have been integrated into 4G Long
Term Evolution (LTE) cellular network architecture to efficiently
address the coverage and capacity issues faced in indoors and at
hotspots. Though spectral efficiency increases through frequency
reuse one at Femtos, it could lead to co-tier interference and
cause higher interference for cell edge User Equipments (UEs).
This problem is more severe in enterprise and hotspot Femto
deployments due to dense placement of Femtos. Existing co-tier
interference management techniques do not solve this problem
completely. Hence, in this paper, we propose a Variable Radius
(VR) algorithm which dynamically increases or decreases the cell
edge/non-cell edge region of Femtos and efficiently allocates the
radio resources among cell edge/non-cell edge region of Femtos so
that the co-tier interference between neighboring Femtos can be
avoided. We implemented the proposed VR algorithm on top of
Proportional Fair (PF) scheduling algorithm in NS-3 simulator.
In our experiments, for 90 UEs the proposed technique (VR +
PF) achieved 29% and 38% improvement in average throughput
for static and mobile scenarios, respectively when compared to
classic PF algorithm without any interference management.
Index Terms—LTE; Femto Cells; Interference Management;
Spectral efficiency
I. INTRODUCTION
Due to popularity of smart phones and tablets, there is
an exponential increase in the demand for higher data rates.
To provide higher data rates, 3GPP proposed 4G Long Term
Evolution (LTE) standard. As per traffic statistics given by
Huawei and Nokia-Siemens [1], [2], 60% of the voice and
video traffic in cellular networks come from indoor environ-
ments. The indoor users typically get low data rates because
of poor cellular network coverage inside buildings. Into the
LTE standard [3], Femto Base Stations (Home eNB/Enterprise
eNB) are introduced to provide good coverage and high data
rates for the indoor User Equipments (UEs). These Femtos are
installed by end users who have broadband wire-line Internet
connections. In enterprise Femto networks, a large number of
Femtos are deployed in places such as office buildings and
hotspot areas. These Femtos can serve 15 to 25 UEs and
have coverage of 60 to 70 meters [4]. LTE system comprising
of legacy macro BSs and Femto BSs is called as two-tier
Heterogeneous Network (HetNet).
Different types of Femto access are defined namely open,
closed and hybrid. In open access, all UEs of a given cellular
(mobile) network operator are allowed to connect to Femto
BS, but in closed access, only authorized UEs are allowed
to connect. In hybrid access, both authorized UEs and a
limited number of other UEs can connect in a prioritized
manner. In case of enterprise networks and hotspots, Femtos
are commonly configured for open access. Fig. 1 shows the
architecture of LTE Femtocell network, where Femtos are
connected to broadband network (Internet) and then eventually
connected to Femto Gateway (F-GW) via S1 interface.
P−GW
Internet
X2
E−eNB
E−eNB
E−eNB
E−eNBX2
MME/S−GW
S1
F−GW
S1
X2
S1
X2
X2
S1
S1
S1
S1
Fig. 1. Architecture of Enterprise LTE Femtocell Network
Interference results in packet loss and low data rates [5].
Two types of interference is possible between macro and
Femto BSs of two-tier LTE HetNets namely cross-tier in-
terference and co-tier interference [5]. Cross-tier interference
occurs between macro and Femto BSs. It occurs especially
when same bandwidth (RB) is allocated to the UEs of both
macro and Femto BSs. Co-tier interference occurs when all
Femto BSs (also true for macro BSs) share the same spectrum
resources through frequency reuse one. Due to high UE den-
sity, enterprise Femto BSs experience high co-tier and cross-
2013 IEEE 9th International Conference on Wireless and Mobile Computing, Networking and Communications (WiMob)
978-1-4799-0428-0/13/$31.00 ©2013 IEEE 553
tier interference when compared to Femto BSs used in home
environments. In [6], three cross-tier interference management
schemes are proposed. First scheme divides spectrum between
macros and Femtos, but as number of Femtos increases the
spectrum allocated to macros decreases considerably. Second
scheme allocates the whole spectrum to both macros and Fem-
tos which can lead to high interference. In third scheme, some
part of the spectrum is shared by Femtos and macros. The
remaining spectrum is divided between macros and Femtos.
But, this scheme is efficient only if UEs count is low.
In this paper, we propose a Variable Radius (VR) algorithm
which dynamically increases or decreases the cell edge region
of Femtos logically and efficiently allocates radio Resource
Blocks (RBs) among cell edge/non-cell edge region of Femtos
so that the co-tier interference between neighboring Femtos
can be avoided in enterprise deployments. Rest of the paper
is organized as follows: Section II describes the related work.
Proposed VR algorithm is discussed in Section III. The sim-
ulation methodology and results are presented in Section IV.
Finally, Section V contains concluding remarks.
II. RELATED WORK
In this section, we review existing works addressing the
interference issues due to incorporation of Femtos into LTE
systems. In Release 8 [7], X2 interface avoids the interference
at the cell edge of two neighboring macro BSs. In this case,
eNBs share the information of RBs assigned to cell edge
UEs.In Release 11 [7], X2 interface is introduced between
Femtos of enterprise femtocell networks to avoid interference
and directly route the data and signaling messages among
Femtos, thereby reducing the load on Mobility Management
Entity (MME) of LTE core network and offers better coordi-
nation among Femtos. Two types of interference is possible in
two-tier cellular network i.e cross-tier and co-tier. Cross-tier
interference can be avoided by dividing the spectrum between
macro and Femto cells orthogonally [8], [9]. In their schemes
resources are shared between Femtos in a distributed manner
by using F-ALOHA scheme, which introduces slotting and
contention amongst Femtos. But, in this scheme spectrum can
not be reused unlike proposed VR algorithm.
Two types of frequency reuse techniques can be applied
to reduce co-tier interference. Fractional Frequency Reuse
(FFR) [10] has frequency reuse three, which means that only
one third of the spectrum is used in a particular cell and
therefore leads to inefficient usage of spectrum resource. The
other approach is Soft Frequency Reuse (SFR) [11], [12].
In SFR, the cell area is divided logically into two regions
based on spectrum allocation: an inner region where major
portion of spectrum is available and a cell edge area where a
small fraction of the spectrum is available. Since the Shannon
capacity at cell edge may be very low, it can be increased
by allocating higher power carriers to UEs in this region,
where as lower power carriers are allocated to UEs in the inner
region. But, SFR was studied only for macros. To improve the
spectrum efficiency and throughput for the indoor UEs, SFR
technique can also be adapted to enterprise Femto networks.
But the drawback of implementing SFR in Femtos is that
it can lead to high interference due to overlap of coverage
regions of Femtos. Hence, we propose an efficient interfer-
ence management technique (VR: Variable Radius algorithm)
which dynamically increases or decreases the width of cell
edge region inside the Femto coverage area to overcome the
drawback of SFR for Femtos.
III. PROPOSED WORK
In this work, we consider a two-tier HetNet comprising of
macro and Femto BSs in a LTE system. Inside the enterprise
buildings we assume that, a large number of Femtos are
deployed and configured for open access. We rely on Position
Reference Signal (PRS) [13] to get the positions of UEs
inside the buildings without GPS. We also assume that the
available spectrum is divided between macros and Femtos to
avoid cross-tier interference. But, co-tier interference can exist
among Femtos due to reuse factor one and overlap of coverage
regions. To reduce this co-tier interference in enterprise Femto
networks, two logical regions namely inner and outer region
are assumed inside the Femto coverage area as shown in
Fig. 2. The radius of inner region (and hence the width of
outer region) changes dynamically. These regions are created
logically due to changes in the power transmitted and the
average CQI values, but only created virtually by the Femtos
in the proposed VR algorithm.
F
Outer region (Cell Edge)
Inner region
Fig. 2. Regions inside Femto coverage area
Every Femto communicates about the RBs allocated to its
cell edge UEs with neighboring Femtos through X2 interface.
Femto allocates RBs to its UEs in outer region such that same
RBs are not allocated in the outer regions of its neighboring
Femtos, thus avoiding the interference. Such an allocation is
known as restricted RB allocation. But, since the delay to get
the required number of RBs increases, the average throughput
for cell edge UEs decrease. For the UEs of inner region, there
is no such restriction on RB allocation, unlike cell edge users.
Any free RB can be allocated to them.
Let us consider an enterprise Femto deployment scenario
with six Femtos namely F1-F6 and randomly placed UEs as
shown in Fig. 3 for describing the proposed VR algorithm.
The two scenarios of it are given below.
Interference Scenario 1: Initially the width of outer region
is zero. In this case, interference occurs if the cell edge
UEs in overlapping regions of neighboring Femtos use the
554
same RBs. This results in decrement of data rate and CQI
due to poor signal-to-interference-to-noise ratio (SINR) value.
According to 3GPP TS [36.301], the CQI values vary from 1
to 15. Active UEs provide CQI feedback to Femto at regular
intervals. Femto transmits data with higher modulation scheme
like 64-QAM if the UE has higher CQI value. In consummate
circumstances caused by very high interference, CQI value
becomes zero and the UE may not able to transmit any data.
To reduce this interference, the radius of the inner region
is decreased which inturn increases the width of the outer
region as shown in Fig. 3. To determine the average CQI
of an UE at a distance dfrom the Femto center, firstly an
inner and an outer circle are drawn with the radius as (d-
δ) and (d+δ), respectively as shown in Fig. 4. The width
of the resultant strip is 2δ(in our experiments, δis taken as
0.5 m). Secondly, the average of the CQI of all UEs within
this strip is calculated and this value is assigned to the UE
at distance d. The average CQI is calculated and assigned
similarly to every UE at any distance from the Femto within
the radius of inner region. Thirdly, the average CQI of all
UEs is sorted in increasing order. Fourthly, the first UE whose
average CQI value is greater than a threshold CQI value is
identified (in our experiments, the threshold CQI is set as
4). The distance of this identified UE from the Femto is the
threshold distance and is named γ. Finally, bisection method
is used to calculate the mean of the inner region radius rand
the threshold distance γas r=(r+γ)/2.
This mean value (r) is the radius of the inner region. By
using X2 interface the interference is avoided in the outer
region by exchanging signaling messages between neighboring
Femtos for restricted RB allocation. Bisection method is used
in general to find the roots of a polynomial. Here Bisection
method is used to find the approximate radius value for which
the average CQI value at the given radius is equivalent to
threshold value.
Using mean value helps us to decrease the effective inner
radius from R (cell radius) to r. The advantage of calculating
the mean over considering the threshold distance γ, as radius,
is that when the threshold distance is very small, a large
number of UEs reside in outer region which may lead to
unfair allocation of RBs to the cell edge UEs, as very less
amount of RBs are catered to cell edge UEs, due to restricted
allocation. The threshold CQI value for contraction is the
CQI used for indoor data traffic handover i.e., less than -3 dB
in terms of SINR. [14] gives the mapping of SINR to CQI.
Interference Scenario 2: When the number of UEs in the
outer region increases drastically, due to restricted RB allo-
cation, many RB requests from UEs of the outer region may
not get satisfied. This leads to dramatic decrease in the system
throughput. In order to overcome this problem, the inner region
has to be expanded to accommodate the excess UEs of the
outer region as shown in Fig. 5.
Depending upon the fail ratio (FR) the radius increases,
where FR is defined as, FR =RejectedRequests(RR)
AcceptedRequests(AR), where
RR is the number of unsatisfied requests coming from outer
Outer Region (−−−) : High TX Power
Inner Region(white) : Low TX power
***
*
*
*
*
*
**
*
*
**
**
**
*
**
*
*
*
*
***
*
*
*
*
*
*
**
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
**
F1
F4
F6
F7
F3
*: User
x2
x2
x2 F2
*x2
x2
x2
x2
x2
F5 *
x2
x2
: E−eNB
Fig. 3. Reducing the inner regions of Femtos
Fig. 4. Calculating Avg CQI value in the strip
***
*
*
*
*
*
**
*
*
**
**
**
*
**
*
*
*
*
***
*
*
*
*
*
*
**
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
**
F1
F4
F6
F7
F3
: E−eNB
*: User
x2
x2
x2 F2
*x2
x2
x2
x2
x2
F5 *
x2
x2
Fig. 5. Increasing the Inner regions of Femtos
region due to restricted RB allocation and AR is the number
of requests coming from outer region that can be satisfied in
some subframe. Due to unavailability of RBs certain requests
cannot be satisfied in a particular subframe and these requests
555
are excluded in AR. The radius of inner region will remain the
same if FR is less than or equal to the threshold value and this
value can be set by the network operator. If FR is greater than
threshold value, the radius of inner region is increased. The
radius is incremented by δsuch that RR−(AR/2) unsatisfied
UEs from outer region are brought into the inner region. Thus,
the excess UEs of outer region are brought into the inner
region and the FR reduces below FR Threshold. Hence,
the UE load in the outer region reduces and the throughput
increases. The proposed VR algorithm (refer Algorithm 1) will
therefore reduces the interference efficiently in a large scale
deployment of Femto networks.
Algorithm 1 Variable Radius Algorithm
Input CQI Threshold :Handover CQI threshold
Input FR Threshold :Threshold Fail Ratio
Input R:Radius of Femto
0: r←R{Initialize Radius of Inner Region}
while true do
CQI ←CalculateCQIInnerRegion(); {Calculates
average CQI for a given inner region }
if ( CQI <CQI Threshold )then
DecreaseRadius ←true;
else
DecreaseRadius ←false;
end if
FR ←Cal culateF RU EsOuterReg ion(); {Calculates
Fail Ratio of UEs in cell edge region }
if (FR>FR Threshold )then
IncreaseRadius ←true;
else
IncreaseRadius ←false;
end if
if ((IncreaseRadius) && (DecreaseRadius)) || ((!In-
creaseRadius) && (!DecreaseRadius)) then
Continue;
else
if ( DecreaseRadius && !IncreaseRadius ) then
CQI array ←Sort(CQI inner region)
γ←Search(CQI array){finds threshold distance
γof the first UE whose AVG CQI along circumfer-
ence of circle with radius d>CQI Threshold }
r←(r+γ)/2;{(where δis the width of the region
containing users whose AVG CQI <CQI T hreshold)}
PFScheduling(); {Proportional Fair Algorithm}
else
r←r+δ;{(where δwill bring RR-(AR/2)
unsatisfied users of outer region nearest to the boundary
between inner and outer regions into inner region)}
PFScheduling();
end if
end if
end while
end
TABLE I
SIMULATION PARAMETERS
Parameters Values
Number of Femto cells 6
Number of UEs per Femto 10, 15
UE Deployment Random
Femto coverage range 70 m
Femto BandWidth 5 MHz (25 RBs)
Duplexing Mode FDD
Scheduling Algorithm PF, VR+PF
Simulated Traffic Downlink (Video)
Mobility of Mobile UEs 1m/s
Mobility of Static UEs 0.1m/s
Mobility Model Building Mobility Model
Application Data Rate 4 Mbps
Frame Duration 10 ms
TTI 1ms
1020
1040
1060
1080
1100
1120
1140
1160
1180
1200
1220
900 950 1000 1050 1100 1150 1200 1250
Position in Y-axis
Position in X-axis
F3
F5
F6
F1
F4
F2
Femto(F)
Users
Fig. 6. Positions of six Femtos with 90 UEs
IV. SIMULATION METHODOLOGY AND RESULTS
In NS-3 simulator six apartment buildings scenario is cre-
ated and in each apartment one Femto is placed randomly.
Simulation parameters are given in the Table 1. The VR
algorithm is implemented in NS-3 on top of the Proportional
Fair (PF) scheduling algorithm to ensure fairness to all the
UEs. We modified the building mobility model in NS-3 to
introduce limited mobility for indoor UEs. We restrict the users
from entering into the other room, as we are not dealing with
handovers in this work. In real life, even static users will have
some mobility. In order to replicate the same scenario in the
simulator, we assigned the mobility rate as 0.1 m/s even for
the static users. Each UE has single downlink flow from its
connected Femto. The CQI Threshold is the CQI value used
for indoor data traffic handover. It varies between 4 and 6 and
it is less than -3db in terms of SINR. The FR Threshold
is set as 0.5. The metrics used for performance evaluation
are area spectrum efficiency in b/s/hz/m*m and throughput in
Mbps. The results shown in this work are the averaged values
after running simulations for 10 different seed values. Fig. 6
shows the positions of 90 indoor UEs (6 Femtos and 15 UEs
in each Femto).
556
0
0.2
0.4
0.6
0.8
1
0 0.5 1 1.5 2 2.5 3
CDF interms of Users
Throughput in Mbps
PF With No VR & No Mobility 60 users
VR+PF With No Mobility 60 users
FFR With No Mobility 60 user
Fig. 7. Throughput for 60 static UEs inside the building
0
0.2
0.4
0.6
0.8
1
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6
CDF interms of Users
Throughput in Mbps
PF With No VR & No Mobility 90 users
VR+PF With No Mobility 90 users
FFR With No Mobility 90 user
Fig. 8. Throughput of 90 static UEs inside the building
0
0.2
0.4
0.6
0.8
1
0 0.5 1 1.5 2 2.5
CDF interms of Users
Throughput in Mbps
PF With No VR & Mobility 60 users
VR+PF With Mobility 60 users
FFR With Mobility 60 user
Fig. 9. Throughput of 60 mobile UEs inside the building
1. Throughput Results: In Figs. 7 and 8, average throughput
of VR+PF algorithm is compared against classic PF scheduling
and FFR for 60 and 90 static UEs (i.e., one flow per UE),
respectively. Average throughput for 60 static UEs is increased
by 27% when VR algorithm is employed in PF. For 90 static
0
0.2
0.4
0.6
0.8
1
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8
CDF interms of Users
Throughput in Mbps
PF With No VR & Mobility 90 users
VR+PF With Mobility 90 users
FFR With Mobility 90 user
Fig. 10. Throughput of 90 mobile UEs inside the building
0
2e-06
4e-06
6e-06
8e-06
1e-05
1.2e-05
1.4e-05
1.6e-05
1.8e-05
1 2 3 4 5 6
Area Spectral Efficiency in b/s/hz/m*m
Femto ID
VR+PF With No Mobility 60 users
PF With No VR & No Mobility 60 users
FFR With No Mobility 60 user
Fig. 11. Area Spectrum Efficiency of Femtos with 60 static UEs
indoor UEs, the average throughput is increased by 29%
when VR algorithm is employed in PF. In Figs. 9 and 10,
achieved throughput of VR+PF algorithm is compared against
PF and FFR for 60 and 90 mobile UEs, respectively. Average
throughput for 60 mobile UEs is increased by 37% when
VR+PF algorithm is used. For 90 UEs the average throughput
is increased by 38% when VR+PF algorithm is used. Since
the inner region radius changes dynamically more number of
UEs can be served by the inner region and thus it increases
the average throughput. Bisection method makes sure that UEs
who are supposed to be in the outer region will come inside
the inner region, even though they have interference with
neighboring Femtos. It is observed that proposed VR algorithm
also performs better in mobile scenarios because UEs mobility
there is enough potential for interference management and load
balancing in outer regions and the average CQI values of UEs
with high mobility vary at much faster rate when compared
to UEs with low mobility.
2. Area Spectrum Efficiency Results: In Figs. 11 and 12,
area spectral efficiency of VR+PF, PF and FFR are compared
for 60 and 90 static UEs, respectively. In Figs. 13 and 14,
area spectral efficiency of VR+PF and PF are compared for
557
1e-06
2e-06
3e-06
4e-06
5e-06
6e-06
7e-06
8e-06
9e-06
1 2 3 4 5 6
Area Spectral Efficiency in b/s/hz/m*m
Femto ID
VR+PF With No Mobility 90 users
PF With No VR & No Mobility 90 users
FFR With No Mobility 90 user
Fig. 12. Area Spectrum Efficiency of Femtos with 90 static UEs
0
2e-06
4e-06
6e-06
8e-06
1e-05
1.2e-05
1.4e-05
1.6e-05
1.8e-05
1 2 3 4 5 6
Area Spectral Efficiency in b/s/hz/m*m
Femto ID
VR+PF With Mobility 60 users
PF With No VR & Mobility 60 users
FFR With Mobility 60 user
Fig. 13. Area Spectrum Efficiency of Femtos with 60 mobile UEs
1e-06
2e-06
3e-06
4e-06
5e-06
6e-06
7e-06
8e-06
9e-06
1e-05
1.1e-05
1 2 3 4 5 6
Area Spectral Efficiency in b/s/hz/m*m
Femto ID
VR+PF With Mobility 90 users
PF With No VR & Mobility 90 users
FFR With Mobility 90 user
Fig. 14. Area Spectrum Efficiency of Femtos with 90 mobile UEs
60 and 90 mobile UEs, respectively. In order to be more
precise, area spectral efficiency of each of six Femto is plotted
separately in the graphs. Area spectral efficiency of Femtos
has increased considerably as the interference is avoided in
the outer overlapping regions of Femtos by restricted RB
allocation with the help of communication over X2 interface.
V. CONCLUSIONS AND FUTURE WORK
The proposed VR algorithm dynamically increases or de-
creases the radius of inner regions to avoid co-tier interference
among Femto BSs. All Femtos need not increase/decrease
their inner region radius by same amount at the same time
as VR algorithm depends on the user count and overlap with
neighboring Femtos. In VR, using FR the radius of inner
region is increased. We intend to determine the optimal value
of FR in our future work. We also have to define a function to
vary δbased on UE density. Also, the proposed VR needs to
be modified to include the cases of handovers between Femtos.
ACKNOWLEDGMENT
This work was supported by the Deity, Govt of India (Grant
No. 13(6)/2010CC&BT).
REFERENCES
[1] S. Nielsen, “Lte evolving towards local area in release 12 and beyond.”
Future Radio in 3GPP, Nokia Corporation, 2012.
[2] “Views on rel-12 and onwards for lte and umts.” Future Radio in 3GPP,
Huawei Technologies, 2012.
[3] “Small Cells Forum.” http://smallcellforum.org/smallcellforum.
[4] “Enterprise multi-femtocell deployment guidelines.” Qualcomm Inc.,
2011.
[5] N. Saquib, E. Hossain, L. B. Le, and D. I. Kim, “Interference man-
agement in ofdma femtocell networks: Issues and approaches,” IEEE
Wireless Communications, vol. 19, no. 3, pp. 86–95, 2012.
[6] M. Andrews, V. Capdevielle, A. Feki, and P. Gupta, “Autonomous
spectrum sharing for mixed lte femto and macro cells deployments,” in
IEEE INFOCOM Conference on Computer Communications Workshops,
2010, pp. 1–5, 2010.
[7] “3GPP Releases.” http://www.3gpp.org/releases.
[8] V. Chandrasekhar and J. Andrews, “Spectrum allocation in tiered cellular
networks,” IEEE Transactions on Communications, vol. 57, no. 10,
pp. 3059–3068, 2009.
[9] J. Yoon, M. Y. Arslan, K. Sundaresan, S. V. Krishnamurthy, and
S. Banerjee, “A distributed resource management framework for inter-
ference mitigation in ofdma femtocell networks,” in ACM MOBIHOC
2012, pp. 233–242, 2012.
[10] P. Lee, T. Lee, J. Jeong, and J. Shin, “Interference management in lte
femtocell systems using fractional frequency reuse,” in The 12th Inter-
national Conference on Advanced Communication Technology (ICACT
2010), vol. 2, pp. 1047–1051, IEEE, 2010.
[11] Y. Yu, E. Dutkiewicz, X. Huang, M. Mueck, and G. Fang, “Performance
analysis of soft frequency reuse for inter-cell interference coordination
in lte networks,” in International Symposium on Communications and
Information Technologies (ISCIT 2010), pp. 504–509, 2010.
[12] M. Al-Shalash, F. Khafizov, and Z. Chao, “Interference constrained soft
frequency reuse for uplink icic in lte networks,” in IEEE PIMRC 2010,
pp. 1882–1887, 2010.
[13] J. A. del Peral-Rosado, J. A. L´
opez-Salcedo, G. Seco-Granados,
F. Zanier, and M. Crisci, “Joint channel and time delay estimation for
lte positioning reference signals,” in 6th ESA Workshop on NAVITEC
2012, pp. 1–8, 2012.
[14] M. T. Kawser, N. I. B. Hamid, M. N. Hasan, M. S. Alam, and M. M.
Rahman, “Downlink snr to cqi mapping for different multiple antenna
techniques in lte,” in International Conference on Future Information
Technology (ICFIT 2010), 2010.
558
