Date of Award
Doctor of Philosophy (PhD)
Electrical & Computer Engineering
Frederic (Rick) D. McKenzie
The proliferation of intelligent wireless devices is remarkable. To address phenomenal traffic growth, a key objective of next-generation wireless networks such as 5G is to provide significantly larger bandwidth. To this end, the millimeter wave (mmWave) band (20 GHz -300 GHz) has been identified as a promising candidate for 5G and WiFi networks to support user data rates of multi-gigabits per second. However, path loss at mmWave is significantly higher than today's cellular bands. Fortunately, this higher path loss can be compensated through the antenna beamforming technique-a transmitter focuses a signal towards a specific direction to achieve high signal gain at the receiver. In the beamforming mmWave network, two fundamental challenges are network topology control and user association and scheduling. This dissertation proposes solutions to address these two challenges. We also study a spectrum sensing scheme which is important for spectrum sharing in next-generation wireless networks.
Due to beamforming, the network topology control in mmWave networks, i.e., how to determine the number of beams for each base station and the beam coverage, is a great challenge. We present a novel framework to solve this problem, termed Beamforming Oriented tOpology coNtrol (BOON). The objective is to reduce total downlink transmit power of base stations in order to provide coverage of all users with a minimum quality of service. BOON smartly groups nearby user equipment into clusters to dramatically reduce interference between beams and base stations so that we can significantly reduce transmit power from the base station. We have found that on average BOON uses only 10%, 32%, and 25% transmit power of three state-of-the-art schemes in the literature.
Another fundamental problem in the mmWave network is the user association and traffic scheduling, i.e., associating users to base stations, and scheduling transmission of user traffic over time slots. This is because base station has a limited power budget and users have very diverse traffic, and also require some minimum quality of service. User association is challenging because it generally does not rely on the user distance to surrounding base stations but depends on if a user is covered by a beam. We develop a novel framework for user association and scheduling in multi-base station mmWave networks, termed the clustering Based dOwnlink user assOciation Scheduling, beamforming with power allocaTion (BOOST). The objective is to reduce the downlink network transmission time of all users' traffic. On average, BOOST reduces the transmission time by 37%, 30%, and 26% compared with the three state-of-the-art user scheduling schemes in the literature.
At last, we present a wavelet transform based spectrum sensing scheme that can simultaneously sense multiple subbands, even without knowing how the subbands are divided, i.e., their boundaries. It can adaptively detect all active subband signals and, thus, discover the residual spectrum that can be used by unlicensed devices.
"Topology Control, Scheduling, and Spectrum Sensing in 5G Networks"
(2020). Doctor of Philosophy (PhD), Dissertation, Electrical/Computer Engineering, Old Dominion University, DOI: 10.25777/5jdz-tm09