Journal of Wireless Networking and Communications 2017, 7(1): 1-8

DOI: 10.5923/j.jwnc.20170701.01

Mimo Optimization Techniques and Their Application

in Maximizing Throughput for 3GPP HSPA+
Olabode B. Idowu-Bismark*, Augustus E. Ibhaze, A. A. Atayero

Electrical & Information Engineering Dept Covenant University, Ota, Nigeria

Abstract The world is going digital by the day and the need for digital data communication such as high definition video
streaming, web browsing and machine to machine communication due to internet of things (IoT) are on the increase
continually, such applications require high data rate to deliver them and without increase in bandwidth which is an expensive
resource constant. Also in order for 3G cellular system operators to compete favorably in the mobile data market with new
competing technologies such as WIMAX, LTE etc especially in the future explosive IoT market and to preserve their
investment a cost effective way is to utilize the Multiple Input and Multiple Output antenna system otherwise called MIMO to
obtain the needed high data rates. In this paper we looked at the various data rate optimization techniques of MIMO and their
applications in 3G networks. Using a drive test we conducted investigation and compare the impact of MIMO on the
throughput of legacy Rel. 99, HSDPA and HSPA+ Rel. 9. The result of our investigation shows that while Rel. 99 offered a
0.403Mbps in the uplink, HSUPA offered 1.335Mbps with HSPA+ showing a throughput of 10.950Mbps. Likewise
throughput in the downlink is 0.244Mbps, 1.041Mbps and 35.370Mbps for Rel. 99, HSDPA and HSPA+ respectively
showing a progressive increase as the number of transmit antenna increases at the base station.
Keywords Diversity, HSPA+, MIMO, Spatial Multiplexing

experience. In section 2 of this paper, we looked at the

1. Introduction MIMO system and its various optimization techniques. We
considered MIMO evolution in 3GPP HSPA/HSPA+ in
It is pertinent that since 3G cellular systems cannot be section 3, while the result of our drive test were presented in
done away with and must compete in the mobile data section 4. We concluded in section 5.
industry with the new technologies that has inherent capacity
for higher data rate, we must look for a solution to high data
rate demand that uses the 3G cellular infrastructure [1]. One 2. MIMO System
of such solutions is the improvement of performance and
spectral efficiency using MIMO. 2.1. MIMO Basic Models
MIMO takes advantage of the space dimension to enhance
network capacity, increase range and improve reliability.
MIMO does all these without additional bandwidth nor
increased transmit power. The application of higher order
modulation and MIMO in 3G cellular systems like the
HSDPA produces the Evolved HSPA+. High data rate
demand, high network capacity and coverage are the basic
challenges of every wireless cellular networks as users
demand for data usage soars, various technologies, methods
and optimization techniques have been designed for network
improvement and spectral efficiency. One way is to
maximize 3G network performance through MIMO
optimization techniques, thereby enabling it to offer higher Figure 1. Different Antenna Format
data rate, better quality of service and quality of end user
There are four basic antenna configuration models which
include Single Input Single Output (SISO) where the
* Corresponding author:
[email protected] (Olabode B. Idowu-Bismark) transmitter and receiver of the radio system have only one
Published online at http://journal.sapub.org/jwnc antenna each, next is the Single Input Multiple Output
Copyright © 2017 Scientific & Academic Publishing. All Rights Reserved (SIMO) where the receiver has multiple antennas but the
2 Olabode B. Idowu-Bismark et al.: Mimo Optimization Techniques and
Their Application in Maximizing Throughput for 3GPP HSPA+

transmitter has only one antenna followed by Multiple Input Where r and s are the received and transmit vectors, while
Single Output (MISO), here the transmitter has multiple H and n are the channel matrix and the noise vector
antennas and the receiver has only one antenna. Finally respectively as shown in Figure 3 below. [11]
Multiple Input Multiple Output (MIMO) which means that
both the transmitter and receiver have multiple antennas.
This is as shown below in Fig 1.

2.2. Channel Capacity for MIMO

For the purpose of simplicity, MIMO channel capacity can
be represented by the approximated formula below.
𝐶𝐶 ≈ m∙log (1+SNR) (1)
Where m is the number of antennas in either the
transmitter or receiver, whichever one is smaller. From the
equation, we could see that MIMO system offers spectral
efficiency that increases linearly with the increase in the
number of antennas, where multiple parallel channels were
Figure 3. MIMO Matrix
created to deliver more data traffic streams simultaneously.
Figure 2 below shows SISO, SIMO, MISO, and MIMO The channel state information is represented by H, the
systems for 2-antenna and 4-antenna spectral efficiency channel matrix where each of its entries is a distortion
comparison cases. It is observed that significant spectral coefficient acting on the transmitted signal phase and
efficiency improvement can be achieved using MIMO amplitude in time-domain. The H matrix is constructed by
system. [2] the receiver, the transmitter does not have an idea of what the
channel looks like, this state is called open loop system
where the transmitter is transmitting blindly. In the event that
there is a path for the receiver to transmits this matrix
information to the transmitter, the transmitter would then be
able to see the status of the signals at the receiver and might
make adjustments in the powers allocated to its antennas
based on this information, this is called closed loop system
[4].
The entire process above can be called channel estimation.
Channel estimation is a very important part of MIMO
performance enhancement. Through channel estimation, we
are able to measure time delays, attenuation, and phase (as a
function of frequency) of each path between transmit and
receive antennas in order to maximize our signal recovery
and hence performance.

2.4. MIMO Optimization

The aim of MIMO Optimization is the achievement of the
Figure 2. Spectral efficiency comparison graph. [3] highest possible throughput and connectivity within the
network and within a given environment by taking advantage
2.3. MIMO Channel Model of the multipath potentials of that environment.
In MIMO systems, the transmitter sends multiple streams
2.4.1. Hindrance to MIMO Optimization
of signals by multiple transmit antennas. The transmit
streams of signals go through a matrix channel which All interference limited networks such as the 3G cellular
consists of all N transmit and N received paths between the systems including HSDPA and HSPA+ have a major
Nt transmitting antennas and Nr receiving antennas. The hindrance in the application of MIMO to optimize and
receiving multiple antennas on getting the received signal increase throughput [5]. Under interference limited networks,
vectors decodes them into the original information. A all users transmit on the same frequency, therefore internal
narrowband flat fading MIMO system is thus modeled with interference generated by the system is the most important
Equation (2) below. The variables are matrices of complex factor that limits the system capacity and determines call
numbers rather than scalar numbers. quality. For maximum capacity to be achieved, the signal to
r = Hs + n (2) noise ratio (SNR) of all users must be at the minimum level
required for acceptable channel performance. Interference
 Journal of Wireless Networking and Communications 2017, 7(1): 1-8 3

decreases the SNR of the network and reduces system the information may be retrieved from the channel or
capacity and throughput, [5]. Interference in cellular systems channels with the best signal to noise ratio (SNR).
includes self–cell interference, other–cell interference and There are two types of spatial diversity:
co–antenna interference [5]. Thus the major hindrance to the 1. Transmit diversity: where more than one transmit
application of MIMO in CDMA 3G cellular networks is antenna is used to send coded information to one
interference management. receive antenna.
2. Receive diversity: where one transmit antenna send
2.5. Types of MIMO Techniques
coded information to more than one receive antennas.
2.5.1. Spatial Multiplexing We lay more emphasis on transmit diversity as a result of
its wide spread application in 3G cellular network.
Spatial multiplexing using M transmitting and N receiving
antennas allows the capacity of a MIMO network system 2.5.2.1. Transmit Diversity
grows linearly with the minimum of M or N without the use Due to the surge for data usage and multimedia services
of extra bandwidth or transmit power. MIMO works as a particularly in the downlink, improving the downlink
combination of SIMO and MISO otherwise called receive capacity has thus become one of the challenges facing the
diversity and transmit diversity respectively (where signals evolution of the 3G technology. Transmit diversity (TD) is
from different paths were combined together to form a therefore one of the contributing technology of MIMO for
stronger single signal with higher SNR), resulting in even addressing this problem in 3G WCDMA and CDMA2000.
greater SNR gain than either of them, it does this by sharing while TD provide diversity gain, channel state feedback can
the total SNR between its multiple data streams with each of be incorporated to provide antenna gain in which case we
them having a lower power level. The result is that each data will have either Open loop TD (without feedback) or Closed
stream will have a lower SNR than would be possible if it loop TD (with feedback). [6] One of the worst types of
were a single data stream. As each data stream recombines channel conditions are slow changing flat fading channels as
multipath signals to increase its SNR, diminishing returns this type of channel condition can result in the received
sets-in and each data stream is limited or capped by the value signal been below the background noise level thereby
of the original data stream SNR. Therefore, each of the making communication unreliable.
multiple data stream may be capable of convening nearly as Transmit diversity can enhance the performance of the
much data as a single original stream. [14]. receiver in the face of flat fading. TD reduces the effect of
This spatial multiplexing technique is very important in a fading by offering multiple independent copies of the signal
high SNR situation where the system is degree-of-freedom at the receiver where the probability of all copies of the
limited rather than power limited. Quoting Foschini G.J, signals been in a fade simultaneously is very small. [7].
Zheng. L et al (2003) stated that the capacity of a channel Transmit diversity is achieved by employing large antenna
with M transmit and N receive antennas with independent spacing of the order of several carrier wavelength between
and identical distributed Rayleigh faded gains between each the transmit antennas leading to uncorrelated fading of the
antenna pair is given by transmit signals. TD has the advantage that the multiple
C (SNR) = min {M, N} log SNR + δ antennas and the complex coding are implemented at the
From the above, it is shown that the degree of freedom is transmitter (Base station) which has enough space and
the minimum of M and N and the channel capacity increases complexity to accommodate them rather than at the receiver
linearly with M and N. Figure 4 below shows the block (Mobile station), TD is thus used at the downlink of 3G
diagram of a spatial multiplexing technique. WCDMA systems [8].

2.5.3. Space Time Codes

Space Time Coding is a method of coding that brings
together various techniques for obtaining spatial
multiplexing gains and diversity gains for a communication
link. There are several techniques grouped under space time
coding which are used to achieve spatial multiplexing and
diversity gains. Space time code can be divided or classified
into two. Space Time Block Code (STBC) and Space Time
Figure 4. Spatial Multiplexing Trellis Code (STTC). An application of the STBC is in space
time transmit diversity (STTD). This STTD is used in UMTS
2.5.2. Spatial Diversity
3G cellular systems mandatorily for the mobile station but
Diversity is the ability to use different ''channels'' to not compulsory for the base station, the improved signal to
convey the same information stream from the transmit noise ratio provided by STTD is used by the mobile station
antenna to the receive antenna when the different channels for cell edge performance [9]. STTC on the other hand works
fade in a statistically independent fashion in such a way that by combining modulation and trellis coding to transmit
4 Olabode B. Idowu-Bismark et al.: Mimo Optimization Techniques and
Their Application in Maximizing Throughput for 3GPP HSPA+

information over multiple transmits antennas and MIMO in 1999 and run through Rel.4, Rel.5, Rel.6 and Rel.7. Rel.8
channels. upward is actually considered as LTE, as the evolution path
of UMTS/WCDMA is towards LTE. The cellular network
2.5.4. Spatial Multiplexing Vs Spatial Diversity system is divided into two parts; the radio network part and
Comparatively SM outperform SD where high signal the core network part. Comparatively, the Rel.99 and Rel.4
scattering environment exist and the SNR is relatively high core network did not differ much from each other [11]. As
like near the base station whereas when the SNR is weak, we move from Rel.4 to Rel.5, there was a tremendous change
like at cell edge, SD outperforms SM because SD will both in the core network and in the UTRAN where IP
improve the SNR and help offer increased data rate [5] thus multimedia subsystem was introduced in the core network
an ideal wireless cellular system will incorporate both SM while IP transmission technology and the High Speed Data
and SD with the system calculating the optimal crossover Packet Access (HSDPA) technology were introduced in the
point between the two in such a way that the system UTRAN. Rel.5 thus brought with it the introduction of smart
dynamically switches between the two to offer the necessary antenna in the UTRAN. Although not a full fledge MIMO
coverage (offered by SD) or capacity gain (offered by SM) system, it has a great impact on the throughput. The MIMO
depending on channel condition. [10] technology was actually introduced in Rel.7
Comparing how the various Releases were able to enhance
throughput with the use of MIMO technology, we show
3. MIMO Evolution in 3GPP HSPA+ below in table 1 and 2 the expected data rates for different
Release 7 to 9 releases of HSPA+ in both downlink and uplink respectively
in relation to the rank of MIMO and multiple carrier
The 3GPP UMTS/WCDMA standard started with Rel.99 frequency used.

Table 1. Expected downlink peak data rate. [11]

HSPA+ P1 HSPA+ P2 HSPA+ P3

Bandwidth
(3GPP R7) (3GPP R8) (3GPP R9/R10)
SISO 21.09 SISO 21.09 SISO 21.09
5MHz
2x2 MIMO 27.95 2x2 MIMO 42.19 2x2 MIMO 42.19
SISO 42.19 SISO 42.19
10MHz N/A 2x2 MIMO 84.38 2x2 MIMO 84.38
4x4 MIMO 168.76 4x4 MIMO 168.76
SISO 84.38
20MHz N/A N/A 2x2 MIMO 168.76
4x4 MIMO 337.52

Table 2. Expected uplink peak data rate. [11]

HSPA+ P1 HSPA+ P2 HSPA+ P3

Bandwidth
(3GPP R7) (3GPP R8) (3GPP R9/R10)
SISO 17.247 SISO 17.247

5MHz 11.498 2x2 MIMO 34.494 2x2 MIMO 34.494

N/A N/A 4x4 MIMO 68.988

SISO 34.494 SISO 34.494

10MHz N/A 2x2 MIMO 68.988 2x2 MIMO 68.988

N/A N/A 4x4 MIMO 137.976

SISO 68.988

20MHz N/A N/A 2x2 MIMO 137.976

4x4 MIMO 275.952

 Journal of Wireless Networking and Communications 2017, 7(1): 1-8 5

Rel.8 came out with further enhancement for HSPA+ such measurement results tally with our observed indications
as a dual carrier operation in the downlink and simultaneous from section 4.4 and 4.5 above that our throughput values
operation of MIMO and 64QAM modulation which push the will be in the fair region/range due to channel conditions as
data rate to 42.2Mbps in the downlink while latency was revealed by the collected RF parameters (FTP RSCP and Tx
further improved upon. In Rel.9, dual carrier operation was Pwr).
combined with MIMO capabilities to push downlink data
Table 3. HSDPA and Rel.99 DL Throughput comparison table
rate to 84.4Mbps while the dual cell HSUPA feature supports
two carrier frequencies in the uplink leading to 23Mbps data THROUGHPUT ANALYSIS
rate. The latest Rel.10 specification uses four carriers
HSDPA that enables a single user access to four carrier Activity/process Sector 1 Sector 2 Sector 3
frequencies of 5MHz each leading to 20MHz bandwidth R99 DL Throughput 201.83 403 125.95
which is what is available with LTE and WiMAX. [12] HSDPA Throughput 1117.6 860.09 1145.38

4. Result and Analysis

Having completed our drive test experiment, we
proceeded here to analyze the reports obtained. Two
experiments were carried out, one for 3G WCDMA/HSDPA
and one for 3.75G HSPA+.

4.1. WCDMA Rel.99 and HSDPA Rel.5 Single Cell

Functional Test Report

Antenna Configuration Figure 5. HSDPA and Rel.99 DL Throughput Comparison chart

Activity/Proces Sector 1 Sector 2 Sector 3 Tabulating the downlink throughput measured for the
ATR ATR ATR three sectors of R.99 and HSDPA, we obtain the comparison
Antenna Model table (Table 3) and graph (Figure 5) above. The graph
451703 451703 451703
Antenna Type (2/4/6 ports) 2 2 2
revealed a drastic improvement of HSDPA downlink
throughput over that of R.99 with an average of over 400%
Antenna Height (m) 18 18 18
justifying the inclusion of DTxAA antenna system in
Antenna Azimuth (+/- 5°) 0 140 240 HSDPA for throughput improvement as compared to the use
Antenna Electrical of TxAA in R.99.
1 0 1
Down-tilt
4.1.2. UL Throughput Analysis
Total Down-tilt (+/- 1°) 1 0 1
Table 4. HSUPA and Rel.99 UL Throughput Comparison table
4.1.1. DL Throughput Analysis THROUGHPUT ANALYSIS
From the result of our experimental test, we see that R.99
sector 1 downlink throughput is low at 201.83Kbps, however Activity/Proc ess Sector 1 Sector 2 Sector 3
this can be explained from the fair (yellow) values of the cell R99 UL Throughput 403.2 403.2 403.2
RSSI (Received Signal Strength Indicator) and the SIR HSUPA UL Throughput 1414.72 1337.62 1251.84
(Signal to Interference Ratio) radio parameters, telling us
that the channel condition is not too good. Looking closely,
we also see that the neighboring cells (U07081, U07082 and
U07083) have strong RSCP (Received Signal Code Power)
radio parameter values (Green) which is good for handover
as they took over most of the traffic, thus negatively
affecting our R.99 sector 1 DL throughput.
The same applies to the HSDPA sector 1 throughput at not
too good value of 1,117Kbps. However, looking at HSDPA
sector 2, the SNR value is really bad (Red) at 15.20 which
gives the sector throughput value at 880.09Kbps showing
and proofing to us that signal to noise ratio is a major factor
(enhances or hindrance) to the effect of MIMO in improving
throughput and data rate. The above direct throughput Figure 6. HSUPA and Rel.99 UL Throughput Comparison Chart
6 Olabode B. Idowu-Bismark et al.: Mimo Optimization Techniques and
Their Application in Maximizing Throughput for 3GPP HSPA+

In the Uplink sector, section 4.1.3 above, tabulating the The combination of high SINR and high MCS values is
throughput measured for the three sectors of R.99 and expected to give us a high throughput value for the network
HSUPA, we obtain the comparison table (Table 4) and graph which is reflected in the measured throughput values for
(Figure 6) above showing a HSUPA higher throughput both UL and DL.
values of about 300% over R.99 UL throughput which is in Putting our result in a tabular form, we have:
agreement with the theoretical throughput analysis of
Table 5. HSPA+ Throughput table analysis
HSUPA and R.99 UL.
S/N Mode UL (Mbps) DL (Mbps)
Activity/Process Sector 1 Sector 2 Sector 3 1 TCP 10.95 35.37
R99 PS Setup 2 UDP 10.95 30.91
5/0 5/0 5/0
Success / Failure 3 Peak Rate 18.63 40.35
HSUPA Setup
5/0 5/0 5/0 The above table shows an average value for the TCP and
Success / Failure
HSDPA Setup
UDP UL/DL values with a channel peak rate for the UL and
5/0 5/0 5/0 DL at 18.63Mbps and 40.35Mbps respectively.
Success / Failure
HSDPA Through put (Kbps) 1,117.6 860.0 1,145.3 4.2.1. Coverage Test
R99 Throughput (DL) (Kbps) 201.8 403.0 125.9 The coverage test was performed to determine the
throughput across the entire network, particularly to
HSUPA Through put (Kbps) 1,414.7 1,337.6 1,251.8
determine the throughput values at the cell edge thereby
R99 Throughput (UL) (Kbps) 403.2 403.2 403.2 determining the limit of the network before handover takes
place.
Latency Average / Maximum (ms) 72/73 80/82 78/80
4.2.2. DL Throughput Plot
4.2. Throughput and Capacity Test Analysis for HSPA+
The HSPA+ Rel.9 drive test experiment for throughput Type DL Throughput DT Plot
measurement was carried out for a 20MHz 4 transmit and 4
receive (4T4R) antenna system for both throughput and
capacity test as well as for coverage test to determine the
throughput at the cell edge of the network where spatial
diversity is used (and not spatial multiplexing that offers
higher throughput values).
The HSPA+ throughput and capacity test experiment,
gave us an uplink throughput in UDP (User Datagram
protocol) and TCP (Transmission Control Protocol) of
10.95Mbps and 10.95Mbps respectively while the
corresponding values in the downlink are 30.91Mbps and
35.37Mbps respectively.
Looking at the outdoor coverage RF parameter of the our
result with various measurement values to justify the above
high throughput values, we see that the RSRP has values for
each sector of the network as -85.5, -83.75 and -98.67
indicating near eNodeB and mid cell location where MIMO
uses spatial multiplexing as against spatial diversity for cell
edge. DL Throughput Drive Test Bin Drive Test
Also, at near cell site (near eNodeB), the SINR is usually Mbps Count Percentage (%)
high which is used by MIMO spatial multiplexing to [INF to 20] 2079 41.06
produce high throughput. Thus the reflected high values of
[20 to 10] 2116 41.81
the measured DL SINR at 24, 21 and 22 (all above the
parameter high range value of 20 for excellent performance) [10 to 5] 634 12.52
is justified. Finally, the above high SINR is collaborated by [5 to 2] 181 3.52
high MCS values (between 17 to 44 corresponding to a
[2 to 1] 27 0.58
modulation order of 6bits/symbol for 64QAM in the MCS
index vs. modulation order table) for higher order [1 to -INF] 25 0.49
modulation scheme. Average DL Throughput Received: < 18.72> Mbps
 Journal of Wireless Networking and Communications 2017, 7(1): 1-8 7

4.2.3. UL Throughput Plot Drawing a comparison table for MIMO and MISO mode
for our test experiment result as shown below in table 7, we
see and is able to proof that MIMO offers a higher
Type UL Throughput DT Plot and Prediction Plot
throughput over transmit diversity.
Table 7. MIMO and MISO UL/DL throughput comparison table

S/N Mode UL (Mbps) DL (Mbps)

1 MIMO 10.95 35.37

2 MISO 4.13 18.72

From table 7 above we see MIMO offering about 200%

improvement over MISO, however we should note that
MIMO would have performed woefully at the cell edge if
applied rather than applying MISO thus the use of MISO
offers improvement of throughput at the cell edge (when
SNR/SINR is lower than what MIMO can cope with).

4.2.5. Final Comparison Analysis

In this paper, we have performed a test experiment to
determine the throughput of legacy 3G Rel.99, HSDPA and
Comments: Outdoor cell boundary of 350 meters obtain at -115 RSRP HSPA+ and in order to verify the impact of MIMO on the
[with 512 kbps] performance of HSPA+ over HSDPA and 3G Rel.99, using
the average values of the sector throughputs (average of
Drive Test Bin Drive Test sector 1, 2 and 3 throughput) for Rel.99 and HSDPA we
UL Throughput Mbps
Count Percentage (%) tabulated the result of the measured throughput for each
[INF to 20] 0 0
system so as to draw a comparison for our conclusion.

[20 to 10] 346 8.01 Table 8. Rel.99, HSDPA and HSPA+ Throughput Comparison Table

[10 to 5] 1206 27.93 REL.99 HSDPA HSPA+

S/N MODE
(Mbps) (Mbps) (Mbps)
[5 to 2] 1324 30.63
1 Uplink 0.403 1.335 10.95
[2 to 1] 448 10.37
2 Downlink 0.244 1.041 35.37
[1 to 0.5] 301 7.01
[0.5 to -INF] 693 16.06

Average UL Throughput Received: < 4.13> Mbps 40

35
4.2.4. Coverage Test Analysis 30
The DL throughput measurement at the cell edge were 25 Rel.99
20 Throughput
taken and the result tabulated in section 4.2.2 above with an
15
average DL throughput received value of 18.72Mbps. The HSDPA
10
same measurement was taken for the UL throughput and the 5 Throughput
result tabulated in section 4.2.3 with a throughput value of 0 HSPA+
4.13Mbps. According to section 2.5.4 (Spatial Multiplexing Throughput
Vs Spatial Diversity), we explained the use of SD at cell
edges where SNR is usually low. Thus in such situation as
we have now, SD in the form of MISO (Transmit Diversity)
is used rather than full MIMO. For our 4T4R antenna
system therefore, we have a 4T by 1R MISO system at the
cell edge of our network yielding a 18.72Mbps DL and Figure 7. Rel.99, HSDPA and HSPA+ Throughput Comparison Chart
4.13Mbps UL throughput. From the above table and chart, we are able to compare
Table 6. Cell edge throughput using MISO and see an improvement of the performance of HSPA+
throughput over both HSDPA and legacy 3G Rel.99.
S/N UL (Mbps) DL (Mbps)
Though the DL throughput of HSDPA is lower than its
1 4.13 18.72 HSUPA UL throughput due to channel conditions,
8 Olabode B. Idowu-Bismark et al.: Mimo Optimization Techniques and
Their Application in Maximizing Throughput for 3GPP HSPA+

nevertheless the UL/DL throughput values for channel condition ideal for their operation as we see in the
HSUPA/HSDPA are still far more superior than the values increased throughput provided by spatial diversity at the cell
obtained from Rel.99 thus supporting and proofing the edge where spatial multiplexing would not have been ideal
effect of DTxAA on HSDPA throughput performance over due to bad channel condition (low SINR). Our experimental
that of Rel.99 that uses TxAA antenna arrangement. test shows that though the above hold true, however the
The overriding factors for the above improvement of practical throughput achieved are far less from the
HSPA+ throughput performance over that of HSDPA and theoretical calculated figures due to channel conditions.
legacy 3G Rel.99 are the multiple bandwidth carrier and the
4x4 MIMO implementation as the HSPA+ (Rel.9) operate a
20MHz bandwidth thus multiplying HSDPA/Rel.99 5MHz
bandwidth by a factor of four which is suppose to translate
to a throughput improvement factor of four. Also while REFERENCES
Rel.99 uses a TxAA antenna arrangement, the HSDPA uses [1] Spatial Multiplexing in Cellular MIMO- CDMA Systems
a DTxAA which is twice that of Rel.99 but the HSPA+ uses with Linear Receivers: Outage Probability and Capacity by
a 4x4 antenna arrangement, offering and showing us a Wan Choi and Jeffrey G. Andrews, 2008.
quantum leap in the amount of throughput available to the [2] MIMO: From Theory to Reality by Ruifeng Wang, 2009.
end-user as displayed in the above chart.
[3] Multiple Input Multiple Output (MIMO) Systems: Basic
Principles, Algorithms and Networking Applications by
Harish Ganapathy, 2005.
5. Conclusions
[4] Tutorial 27- Finding MIMO by Charan Langton, Bernard
There are different techniques used for drive testing Sklar, 2011.
including - Network Benchmarking, Optimization &
[5] Overcoming interference in spatial multiplexing MIMO
Troubleshooting, as well as Service Quality Monitoring. The
cellular networks by Jeffrey G. Andrews, Wan Choi, and
result produced by each drive testing technique differs based Robert W. Heath Jr. IEEE Communication magazine Oct 20,
on the intended objectives. 2006.
While Optimization and Troubleshooting technique is
[6] Overview of Diversity Techniques in Wireless
more usually used for finding specific problems during the
Communication Systems by Hafeth Hourani, 2005.
rollout phases of new networks or to monitor specific
problems reported by consumers, the Service Quality [7] Transmit Diversity in 3G CDMA Systems by R. Thomas
Monitoring technique normally involves making test calls Derryberry, Steven D. Gray, D. Mihai Lonescu, Giridhar
Mandyam and Balaji Raghothaman. IEEE Communications
across the network to a fixed test unit to assess the relative Magazine, 2002.
quality of various services using Mean opinion score (MOS).
We focus our attention on the Network Benchmarking [8] Smart Antenna Implementation for MIMO by Suraya
technique where state of the art multi-channel tools are used Mubeen, A.M Prasad, A. Jhansi Rani. 2012. Journal of
Information Engineering and Application Vol. 2, No 4.
to measure several network technologies such as modulation,
MIMO and service types simultaneously to very high [9] Tutorial: Multiple Input and Multiple Output (MIMO)
accuracy [15]. This technique helps to provide comparable System Analysis by Jian-Jiun Ding, 2007.
information regarding competitive strengths and weaknesses [10] From Theory to Practice: An overview of MIMO space-time
of the network. Result from this technique, such as coded wireless system IEEE Journal on sel. Areas in
comparative data network speed analysis and data communications, Vol. 21 No 3, pp 281-302 by D. Gesbert, M.
throughput analysis, are useful in commercial campaigns. Shafi, D.Shiu, P.J Smith and A. Naguib April 2003.
Network Benchmarking technique is the only way mobile [11] HSPA+ Technology by ZTE Corporation, 2012.
network operators can collect accurate competitive data on
the true level of their own and their competitors technical [12] HSPA+ Technology Introduction, Application Note by Meik
performance and quality levels. Kottkamp, 2009. A Publication of Rohde & Schwarz.
From the result and analysis of our experiment, it is clear [13] MIMO- Future Wireless Communication by Pravin W. Raut,
that MIMO is efficacious in the provision of improved Badjate S.L April, 2013.
throughput as we move from legacy 3G Rel.99 through
[14] Maximizing LTE Performance Through MIMO Optimization
HSDPA to HSPA+ either across different frequency by PCTEL April, 2011.
bandwidth or within the same frequency bandwidth.
Also, each of its optimization technique such as spatial [15] Richa Budhiraja1, Jitendra Singh Jadon ''Study And
Implementation of Drive Test for Development of GSM
multiplexing and spatial diversity are effective in providing
Network'' International Journal of Engineering Trends and
increased data rate/throughput in their own right based on the Technology (IJETT) – Volume 4 Issue 10 - Oct 2013.