| dc.description.abstract | The sky-rocketing growth of multimedia infotainment applications and broadband-hungry
mobile devices exacerbate the stringent demand for ultra high data rate and more spectrum resources. Along with it, the unbalanced temporal and geographical variations
of spectrum usage further inspires those spectral-efficient networks, namely, cognitive
radio and heterogeneous cellular networks (HCNs). This thesis focuses on the system
design and performance enhancement of cognitive radio (CR) and HCNs. Three different
aspects of performance improvement are considered, including link reliability of cognitive
radio networks (CNs), security enhancement of CNs, and energy efficiency improvement
of CNs and HCNs.
First, generalized selection combining (GSC) is proposed as an effective receiver design
for interference reduction and reliability improvement of CNs with outdated CSI. A uni-
ed way for deriving the distribution of received signal-to-noise ratio (SNR) is developed
in underlay spectrum sharing networks subject to interference from the primary trans-
mitter (PU-Tx) to the secondary receiver (SU-Rx), maximum transmit power constraint
at the secondary transmitter (SU-Tx), and peak interference power constraint at the
PU receiver (PU-Rx), is developed. Second, transmit antenna selection with receive
generalized selection combining (TAS/GSC) in multi-antenna relay-aided communica-
tion is introduced in CNs under Rayleigh fading and Nakagami-m fading. Based on
newly derived complex statistical properties of channel power gain of TAS/GSC, exact
ergodic capacity and high SNR ergodic capacity are derived over Nakagami-m fading.
Third, beamforming and arti cial noise generation (BF&AN) is introduced as a robust
scheme to enhance the secure transmission of large-scale spectrum sharing networks
with multiple randomly located eavesdroppers (Eves) modeled as homogeneous Poisson
Point Process (PPP). Stochastic geometry is applied to model and analyze the impact of
i
BF&AN on this complex network. Optimal power allocation factor for BF&AN which
maximizes the average secrecy rate is further studied under the outage probability con-
straint of primary network. Fourth, a new wireless energy harvesting protocol is proposed
for underlay cognitive relay networks with the energy-constrained SU-Txs. Exact and
asymptotic outage probability, delay-sensitive throughput, and delay-tolerant through-
put are derived to explore the tradeoff between the energy harvested from the PU-Txs
and the interference caused by the PU-Txs. Fifth, a harvest-then-transmit protocol is
proposed in K-tier HCNs with randomly located multiple-antenna base stations (BSs)
and single antenna mobile terminals (MTs) modeled as homogeneous PPP. The average
received power at MT, the uplink (UL) outage probability, and the UL average ergodic
rate are derived to demonstrate the intrinsic relationship between the energy harvested
from BSs in the downlink (DL) and the MT performance in the UL. Throughout the
thesis, it is shown that link reliability, secrecy performance, and energy efficiency of
CNs and HCNs can be signi cantly leveraged by taking advantage of multiple antennas,
relays, and wireless energy harvesting. | en_US |
