Summary:
In this
report author has described about
the methods of applying Multiple-Input Multiple-Output (MIMO) antenna systems,
with Time Division Multiplexed (TDM) shared channel on the WCDMA forward link. Author has made it clear in this report
about use of cellular system-level simulations with and without MIMO.
Stimulation methods used by author interpret multiple channel models with some
other cell inferences.
He also proposed methods to computer SINR at output
of linear MMME receiver, and maps it with help of turbo code results. Wide
range of receivers, antennas and transmitters are used to improve system
capacity and present results. Author mentions that use of two transmit with
four receiver antenna, increases capacity up to three times. In this way result
of three sector systems can easily compete with old system with six sector
systems.
Cellular systems and their
working:
According
to report prepared by author MIMO
application in used from last many years, but recently bunch of people realized
that it is important to remove interferences from this technology. Large number
of publications is working with MIMO technology and they are also being
explored with problem of interference. These publications are well aware with
the fact that CDMA cellular system are limited for interference, therefore they
are need of methods which can best handle this issue. As a solution to this
issue, author explained used of flat
fading channels, which are not working for WCDMA yet.
This
report helps readers to know about concepts of applying MIMO with WCDMA forward
links. Stimulation parameters which describe that path loss exponent is 37.6
dBIdecade, Site-to-site Correlation is 0.5, Shadow Fading Standard deviation is
8 dB, Antenna Pattern, mobile distribution and multiple channel model, are
similar to TDM. This stimulation is based on abstract model but soon it will be
observed in real system as well. Besides MIMO, author also discuss about 3GPP parameters which are responsible for
high speed in shared channels.
Author has
covered detail description of cellular system in this report, while explaining
detailed functioning of all schemes. He has provided readers with working of
the MIMO transmit schemes, with operations of receiver used in stream to
compute results and analysis of data. In the end he gives ease to all by
writing summarized conclusion of all his findings.
Cellular system details:
Author has used complete information
about cellular system and its working. He has mentioned that it consisted of 19
three sector or six sector base stations, whose mobile movement is not tracked
with help of sector. For each snapshot, mobile is located within are of central
hexagon. Different mobile locations are stimulated around us and these are
assumed as base stations used to transmit maximum power. Author further explains about selection methods of serving sector.
Stimulation parameters which are described earlier help in selecting serving sector.
He says to select these sectors; each available path is sorted after which
strongest power receiver is selected.
Besides this author further used Thermal noise
power (AWGN) to se 8 dB bellow median signals at hexagon corner, this can insure
to eliminate interference on limited conditions. ITU
multipath (tapped delay-line) channel models, describe in report are also
important to consider. In this model pedestrian A and Vehicular A channels are
discussed. For all these stimulations multiple rays are distributed
independently and identically and Doppler spread is also not mentioned here.
Doppler spread is not part of this model which makes this channel a
quasistatic, from where channel move from fixed slots. This paper further
explain 4x4 system and multi antenna configurations, which are presented in
model as MxN denoting M transmit and N receive antennas.
Transmitter details:
Further
in this report author provide us
with transmitter details, to develop deep understanding for readers about
transmitted signal structure and packet scheduling algorithms.
i)
Transmission Schemes:
Author
has used transmission schemes to utilize transmit antennas at open-loop and
closed-loop for WCDMA schemes. Use of multi transmitter with double antenna
serves two purposes in cellular system with help of block diagram. This diagram spits the information in individual
data streams, and each of those streams is encoded and modulated independently,
to deliver it to best coder to experience extreme channel conditions.
After this Walsh spreading is applied to
assign a scrambling sequence to antenna signals. In the end, each of them is filtered
and finally transmitted to their accurate location in base station power. Two
possible ways used in diagram help in setting antenna array. First method is to
always set V equal to the identity matrix, second says to adapt V based on
feedback from the mobile, where V is comprised of the eigenvectors of the
channel matrix. Both of them provide us results in form of Dual-Stream.
ii)
Packet Scheduling:
Other
transmission method that author has
explained here is pocket scheduling. Under this method a round ribbon pocket
scheduling is assumed to deliver each mobile with accurate amount of queue. Author mentioned, it as an
inappropriate method because it allocate higher throughputs to users which are
near the base stations. For purpose of justification new packets are added with
old one, to enhance the service. This improves capacity for users, which are
away from base station.
Receiver details:
As
receivers, author has used SMIR
estimation procedures, antenna array weight adaptation and quantization, link
adaption and throughput estimation methods in this report. Author has focused
linear MIMSE receivers only in this report.
i)
SNIR Estimation
He told
us that SMIR is computed through
output of linear MMSE receiver with a 2x2 system. The goal of the MMSE receiver is to estimate the
transmitted symbols by filtering (linear combining) the received samples. Purpose
of computing receiver filter coefficients is to bring error between the actual transmitted
symbols and estimated symbols to least.
ii)
Antenna array weight adaption and
quantization
He further says, to estimate capacity of
single stream transmission and dual stream transmission, SINR stream is
selected. When SNIR is high or low, two different streams transmit data. If
SNIR is high then two data streams are executed while with low SNIR one stream
is executed. He has used standardized codes for TXAA description in his report.
According to him 2x2 system with two eigenvectors is used for quantization and
feedback of DS-TXAA. It creates a matrix, where knowledge on one eigenvector
direct signals to the next eigenvector. In this stream data keep on moving from
one to another to boost signal power. Two factors always work together in this
report and used model; they are always fed back using the TXAA method. Author
described the formulation in simple words as a uniform phase shift of all the
elements of an antenna array will still produce the same antenna pattern.
Author has also used 4x2 and 4x4
systems here, in these systems additional bits are required to judge other
antenna array weights. Here four transmit antenna can support only four
orthogonal antenna array weight accurately. A good proposal has been given here
by him to made standards body to extend the codebooks to four antennas. These
codebooks are not available yet to test in real, so author is using assumed
array weights for all four antennas in the stream. Author also assumed perfect channel and received signal covariance matrix
estimation, and reported that mobile never use to have proper knowledge about
this array used by different sectors, they always follow their own adaption
feedback to know about powerful coding or base station. When data is
transmitted mobile use its own knowledge for selection purpose.
iii)
Link Adaption and Throughput Estimation
Link
adaption and throughput estimation as also written in this report by author;
that helps in computing receiver’s performance with turbo code and a line MIMSE
receiver. Here curves are obtained
instead of bounds to estimate performance with turbo codes; this is reason of
semi-analytical results monitored in this stimulation. These curves describe
about turbo code performance in AWGN. To
compute performance more accurately, noise ratio and signal to interface is
computed, best combination of coder and modulator is selected, packet error
rate is computed and mentioned, to remove packet error automatic repeat request
is forwarded and new PER value is processed for best SINR results.
Figure 3 in this report is also a graph
that represents AWGN throughput curves with Chase combining. This graph is
generated assuming a 1x1 system in AWGN. Coder used in this curve was fixed to
a limited type with low ARQ levels. The selected SNR switch points are the
intersection of the throughput curves, which determine average throughput
accurately.
Figure 4 shows Link adaptation simulation
results for Pedestrian A and Vehicular A channel models. In this curve,
throughput is low for Vehicular A channel model. But MIME equalizer is not
completely able to remove all interferences because multi-paths are used here.
With that in this curve throughput of single stream TXAA is lower at high SNR
because the dual stream technique allows for a higher peak data rate by
simultaneously transmitting two data streams.
System Stimulation Results:
Author
further describe stimulated results that are computed from three sector and six
sector schemes with various MIMO and non-MIMO sectoring.
i)
Three Sectors
Table IV in this report contains the
results for a system with three sectors per cell. Data rate for this scheme is
reported by author in megabits for each antenna configuration purpose, and
transmission scheme for multiple channels. ITU Pedestrian A single path
channels were used with ITU Vehicular A multipath channels. This table also
shows an average comparison between capacities of Pedestrian-A and Vehicular-A
channel. If antenna is attached with both receiver and transmitter of MIMI
system, then it greater results are obvious to obtain with high system
capacities.
In Table IV author has shown results that a
2x2 system can double the capacity of a 1x1 system and a 2x4 system can triple
the capacity of a 1x1 system. With that dual stream TXAA provides the best
performance with two transmit antenna system as well. 2x2 system improves its
ability to transmit second stream with 10% more (TXAA). And it is gained 30%
with system of 2x4.
ii)
Six Sectors
In this table author has also described
about six sector system result. He has computed them with 6 sector base
station, and results are compared with three sector results after normalizing
the capacities by the number of sectors per base station. Making sectors is an
efficient technique to use when you want to remove interruption between signals
of base stations. Here with 2x2 system 40% gain is received over 4x2 systems. Author
suggested using three sectors per base station to create better downlinks with
all the antennas.
Conclusion:
At the
end author is giving summary of all his findings that he has collected. He
mentioned about use of MIMO methods which provides highest capacity to Dual-Stream
TXAA. According to calculations, 2x2 MIMO dual stream, transmission increases
the average throughput over single stream transmission by about 10%. While
using 2x4 MIMO dual streams, transmission increases the average throughput over
single stream transmission by about 30%. Further in the results author describe
about use of six sector base station with two antennas per sector, which gave
throughput equal to four antennas per sector. All results computed by author
focus on Rayleigh fading channel models, therefore for any other model these
results are not feasible to use.
Pros:
1. Author in this repot has
mentioned details and procedure about cellular system accurately
2. Author has used professional
skills for gathering accurate details, to provide ease to its readers
3. This system has increased the
average throughput over single stream transmission by about 10%.
4. Results used in this report by
author are compiled by experts
5. Use of figures and graphs by
author has elaborated this report
Cons:
1. Author has used mathematical
calculations and terms in this report, which make it tough for non professional
readers
2. Results computed for this report are
just focusing Rayleigh fading channel model
3. Author has not provided ease to readers
who wanted to take advantage from effects of Rician fading
4. This system is still observed as
abstract model, real finding are not examined yet.
5. Author has over elaborated some
facts in this report, which can misguide readers.
No comments:
Post a Comment