Towards throughput maximization in wireless networks : cognitive radio network and cellular network with D2D transmission / Doaa Mahmoud Khamis Mohammed

Supervisor: Tamer ElBatt

Thesis (M.A.)—Nile University, Egypt, 2017 .

"Includes bibliographical references"

Contents:
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Literature Review . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Motivation and Contributions . . . . . . . . . . . . . . . . . . . . . 5
2. Cognitive Radio Networks with a Dedicated Relay: Throughput and Delay
Trade-offs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.1 System Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.2 CRN with a Dedicated Relay under Perfect Sensing . . . . . . . . . 10
2.2.1 The arrival and service rates . . . . . . . . . . . . . . . . . 11
2.2.2 Lower and upper bounds on the arrival and service rates . . 13
2.2.3 Upper and lower bounds on the stability region . . . . . . . 14
2.3 CRN with a Dedicated Relay under Random Access Protocol . . . 16
2.3.1 The arrival and service rates . . . . . . . . . . . . . . . . . 17
2.3.2 Lower and upper bounds on the arrival and service rates . . 19
2.3.3 Upper and lower bounds on the stability region . . . . . . . 21
2.4 Baseline Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.4.1 No-relaying System . . . . . . . . . . . . . . . . . . . . . . . 22
2.4.2 Cooperative system where the SU acts as a relay . . . . . . 23
viii
2.5 Numerical Results . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.5.1 Stability Regions Comparisons . . . . . . . . . . . . . . . . 25
2.5.2 Average PU and SU Delays Under Perfect Sensing . . . . . 29
2.5.3 Average PU and SU Delays Under Random Access . . . . . 34
2.5.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
3. Cooperative D2D Communications in the Uplink of Cellular Networks
with Time and Power Division . . . . . . . . . . . . . . . . . . . . . . . . 40
3.1 System model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
3.2 Queueing Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
3.2.1 Average service rate of QCUE . . . . . . . . . . . . . . . . . 43
3.2.2 Average service rate of QDT . . . . . . . . . . . . . . . . . . 44
3.2.3 Average arrival and service rates of QDC . . . . . . . . . . . 46
3.2.4 Problem Formulation . . . . . . . . . . . . . . . . . . . . . . 47
3.3 Numerical Results . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
4. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
Bibliography . . . . . . .

Abstract:
In this work, we explore the best achievable throughput for two different wireless
networks. First, we characterize the stable throughput for the secondary user in
cognitive radio networks (CRN). Second, we investigate the case of device to device
(D2D) pair overlaying a cellular network in the uplink, where we optimize the division
of the power and time between the D2D and the cellular transmissions.
In the first part of this work, we characterize the stability region and quantify
the average packet delay for cognitive radio networks with a dedicated relay node,
under two different MAC protocols: perfect sensing and random access. We compare
the performance of our dedicated relay system to two baseline systems, namely norelaying
and secondary user-relaying. Our numerical results reveal that in terms of
the achievable stable throughput, the dedicated relay system outperforms the norelaying
system under the random access protocol. However, this is not the case
in the perfect sensing protocol as the performance of the dedicated relay system
compared to the no-relaying system depends on the primary user’s packet arrival
rate. Moreover, the secondary user-relaying system outperforms the dedicated relay
system. Nevertheless, the secondary user-relaying system has its own challenges in
terms of the standards allowing cooperation with the primary user in addition to the
security and privacy issues involved. In case of random access, delay simulations show
that our dedicated relay system has better primary user delay performance than the
iv
baseline systems at low arrival rates of the primary user packets. On the other hand,
at high arrival rates, the no-relaying system outperforms our dedicated relay system.
In case of perfect sensing and low channel access probabilities, our dedicated relay
system outperforms the no-relaying system due to the merits of cooperation. However,
the high probability of collision dominates as the access probabilities increase, thus,
the no-relaying system starts to have some performance gains.
In the second part of this work, we reflect on cooperative D2D communications as
a promising technology to improve the spectral efficiency in crowded communication
networks. We consider a transmitter-receiver pair, operating in the D2D transmission
mode, overlaying the cellular network. The D2D transmitter (DT) acts as a relay for
the undelivered packets of the cellular user equipment (CUE). We consider the case
in which the DT transmits its own data along with the CUE relayed data using
superposition coding in the uplink. We investigate how the time slot is split between
the cellular network transmission and the D2D transmission. Moreover, the optimal
approach to split the DT power between the transmission of the DT data and the
relayed CUE data is explored. Our main objective is to achieve the the maximum
D2D throughput. Towards this objective, we describe the system using a queuing
theoretic model and formulate an optimization problem to maximize the throughput
of the D2D link by allocating time and power for DT while satisfying the stability
conditions for the queues of the system. Finally, numerical results show the merits of
our system, with optimal time and power allocation, as compared to the constant time
or power allocation scenarios, in terms of the maximum achievable D2D throughput.

Text in English, abstracts in English.