I am interested in almost all aspects of Wireless Communications,
including:
next-generation mobile communication systems, multiple access
techniques, cellular networks, diversity techniques, wideband
CDMA, radio resource management, multi-tier cellular design,
power control, handoff and mobility, wireless local loop,
packet radio networks (CDMA Aloha), interference management
and control, capacity and performance analysis, channel
modeling, software radio, and CDMA multi-antenna systems.
I have a special interest in a project on CDMA multi-antenna systems.
This project is expected to be funded by grants from NSERC and/or
Centers of Excellence.
General Background on Multi-Antenna Systems:
In a multi-antenna system transmission and reception are maintained
through multiple antenna elements (AE's). It is clear, even
intuitively, that such a system will exceed the performance of the
conventional type which uses only a single antenna. However, the
formation of different antenna architectures (that is, different
ways of arranging and networking the AE's), and the analysis of the
improvements that they potentially offer, are non-trivial issues.
Future wireless systems will be required to offer higher capacity,
and even more importantly, substantially higher transmission rates
(in future wireless systems, data traffic is expected to far exceed
the voice type). Therefore, small modifications in the modulation
or coding schemes will most likely be insufficient to attain the
necessary performance improvements in future systems; more fundamental
changes which will yield manifold capacity and rate increases have
to be considered. For instance, a novel antenna architecture, which
would offer substantial performance benefits at the system level,
may be a good candidate for utilization in such future systems. To
this end, the general objective of this project is to utilize
antennas in novel ways so as to achieve performance benefits at the
system level in future wireless networks.
In this research, analytical approaches (based on theory) as well as
computer simulations will be used. Note that our approach is from
systems point of view; therefore, the students are not required to
know much about electromagnetics. All of the below research topics
have potentially significant practical applications.
-
No./Title: YANIKOMEROGLU - 1 CDMA Sectorized Distributed Antenna
Level: Ph.D. and Master's
Synopsis:
CDMA sectorized distributed antenna (SDA) is a recently proposed
multi-antenna system. It has been shown that in an SDA system with
L AE's, in the limit, a remarkable L-fold increase in the reverse
link capacity and/or transmission rate can be achieved. This
remarkable potential gain of the SDA system over the conventional
schemes is achieved through the exploitation of the space dimension,
i.e., through the distributed nature of the AE's. The main objective
of this research is to study the performance of the novel CDMA SDA
system. This will eventually yield a better understanding of the
multi-antenna systems in general, and also will help in developing a
perspective which will enable us to interrelate different types of
multi-antenna systems.
-
No./Title: YANIKOMEROGLU - 2 Optimal Power Control in CDMA SDA System
Level: Ph.D. and Master's
Synopsis:
The high performance achievements of the SDA system are a result of
spatial interference management. In order for the interference
experienced at the CS to be kept to a minimal level, each wireless
user should transmit just at the right power level. Therefore, it is
apparent that a fine reverse link power control scheme, based on
SIR-balancing, is essential for the proper operation of the SDA system.
The objective of this research is to develop efficient power control
schemes for the proper operation of the SDA system.
-
No./Title: YANIKOMEROGLU - 3 Suboptimal Power Control in CDMA SDA System
Level: Ph.D. and Master's
Synopsis:
Due to the joint decoding nature of the SDA system, the set of equations
to be solved in the power control problem become nonlinear, and this
results in the solutions for SIR-balancing algorithms, given in the
literature, becoming inapplicable. Since no closed-form solution exists
for this nonlinear set of equations, iterative solutions should be
considered. However, because of the size of the problem, the converge
speeds of the iterative solutions of the optimal power control algorithms
may be intolerably slow. Consequently, possible suboptimal power control
schemes, which are potentially easier to compute and yet do not incur
significant performance penalties, should be investigated.
-
No./Title: YANIKOMEROGLU - 4 Spatial Interference Analysis
Level: Preferably Ph.D. (consult professor)
Synopsis:
The remarkable performance increase that SDA system offers is valid
only if the interference received by different AE's are uncorrelated.
Therefore, in a CDMA multi-antenna system with L AE's, the existence
of a combiner with L branches, does not automatically guarantee an
L-fold capacity or rate increase. Consequently, in the SDA system,
it is critical to investigate the presence of correlated interference
in the branches of the combiner. In this research, we are interested
in finding the optimal number of AE's, as well as their optimal
locations, for a given set of system parameters, traffic distribution,
processing capability, and service region.
-
No./Title: YANIKOMEROGLU - 5 Forward Link of the CDMA SDA System
Level: Master's
Synopsis:
In the reverse link of an SDA system, although a user's signal is
picked up by all the AE's, having separate feeders for each AE
prevents the accumulation of multiple access interference. This is
not the case, however, for the forward link. If, in the forward
link, a user's signal is transmitted by all the AE's, then the
system will operate as a conventional distributed antenna type.
Clearly, this would result in quite a large difference in the
forward and reverse link capacities. It should be noted that higher
capacity in the reverse link is never a real ``problem"; in order
to equalize the capacities in both links, more resources (such as
bandwidth), from the given pool, can be allocated for the forward
link. Nevertheless, remedies for this situation should be
investigated. For instance, in the forward link, the signal for a
particular user could be transmitted through the AE which has the
highest SIR in the reverse link for that particular user
(alternatively, a few AE's can be utilized). Therefore, there
would be selection diversity in the forward link, and thus, the
benefits of the macro diversity would still be valid.
-
No./Title: YANIKOMEROGLU - 6 Diversity Combining Techniques for
CDMA SDA Systems
Level: Master's
Synopsis:
In the SDA system, the maximal ratio combining scheme is used at
the CS. In order for the maximal ratio combining scheme to maximize
the output SIR, the interference components at the branches of the
combiner must be uncorrelated. If this is the case, then the output
SIR becomes the algebraic sum of the branch SIR's. If, however, the
interference components are correlated, then the output SIR will be
less than the sum of the branch SIR's. In fact, in the case of
significant correlation between the noise components of different
branches, the maximal ratio combining scheme will cease to be the
preferred linear combining technique. In order to demonstrate this
fact, let us consider a hypothetical case, where the signal
components at the branches of the combiner are different but the
noise components are identical. In such a case, the best linear
combining technique would be the selection combining; that is,
selecting the branch with the highest SIR. In the limiting case of
identical signal and interference components at the branches, there
is no gain at all from combining, since amplification does not
increase the SIR. In this research, the choice of diversity
combining technique, in the case of intensive correlation, will be
investigated. Linear (such as equal gain and selection) as well as
nonlinear (square-law) combining techniques will be considered.
-
No./Title: YANIKOMEROGLU - 7 Network Architectures for CDMA SDA
Systems
Level: Master's
Synopsis:
The detailed investigation of the issues related to network
architecture constitute one of the major research items in the SDA
system. In an SDA system, the AE's can be connected to the CS via
microwave, coaxial, fiber-optic, or hybrid coaxial/fiber-optic
links. The choice depends on a host of factors, such as the size
and nature of the coverage region, and deployment cost. However,
these points should be critically investigated by considering the
issues concerning implementation, link budget, feasibility,
scalability, and cost. In addition, the distribution of intelligence
in the system is another issue in need of further study. The
centralized structure of the SDA system (with AE's containing minimal
hardware and a CS where all signal-specific processing is performed)
is conceptually very attractive; however, the other possible ways of
distributing the processing in the system should also be considered
for the potential benefits they may offer in deployment.
Page