Author ORCID Identifier


Document Type


Date of Award


Degree Name

Doctor of Philosophy in Electrical Engineering - (Ph.D.)


Electrical and Computer Engineering

First Advisor

Nirwan Ansari

Second Advisor

Ali N. Akansu

Third Advisor

Abdallah Khreishah

Fourth Advisor

Roberto Rojas-Cessa

Fifth Advisor

Cristian Borcea


In recent years, fifth-generation New Radio (5G NR) has attracted much attention owing to its potential in enhancing mobile access networks and enabling better support for heterogeneous services and applications. Network slicing has garnered substantial focus as it promises to offer a higher degree of isolation between subscribers with diverse quality-of-service requirements. Integrating 5G NR technologies, specifically the mmWave waveform and numerology schemes, with network slicing can unlock unparalleled performance so crucial to meeting the demands of high throughput and sub-millisecond latency constraints.

While conceding that optimizing next-generation access network performance is extremely important, it needs to be acknowledged that doing so for the core network is equally as significant. This is majorly due to the numerous core network functions that execute control tasks to establish end-to-end user sessions and route access network traffic. Consequently, the core network has a significant impact on the quality-of-experience of the radio access network customers. Currently, the core network lacks true end-to-end slicing isolation and reliability, and thus there is a dire need to examine more stringent configurations that offer the required levels of slicing isolation for the envisioned networking landscape.

Considering the factors mentioned above, a sequential approach is adopted starting with the radio access network and progressing to the core network. First, to maximize the downlink average spectral efficiency of an enhanced mobile broadband slice in a time division duplex radio access network while meeting the quality-of-service requirements, an optimization problem is formulated to determine the duplex ratio, numerology scheme, power, and bandwidth allocation. Subsequently, to minimize the uplink transmission power of an ultra-reliable low latency communications slice while satisfying the quality-of-service constraints, a second optimization problem is formulated to determine the above-mentioned parameters and allocations. Because 5G NR supports dual-band transmissions, it also facilitates the usage of different numerology schemes and duplex ratios across bands simultaneously. Both problems, being mixed-integer non-linear programming problems, are relaxed into their respective convex equivalents and subsequently solved.

Next, shifting attention to aerial networks, a priority-based 5G NR unmanned aerial vehicle network (UAV) is considered where the enhanced mobile broadband and ultra-reliable low latency communications services are considered as best-effort and high-priority slices, correspondingly. Following the application of a band access policy, an optimization problem is formulated. The goal is to minimize the downlink quality-of-service gap for the best-effort service, while still meeting the quality-of-service constraints of the high-priority service. This involves the allocation of transmission power and assignment of resource blocks. Given that this problem is a mixed-integer nonlinear programming problem, a low-complexity algorithm, PREDICT, i.e., PRiority BasED Resource AllocatIon in Adaptive SliCed NeTwork, which considers the channel quality on each individual resource block over both bands, is designed to solve the problem with a more accurate accounting for high-frequency channel conditions.

Transitioning to minimizing the operational latency of the core network, an integer linear programming problem is formulated to instantiate network function instances, assign them to core network servers, assign slices and users to network function instances, and allocate computational resources while maintaining virtual network function isolation and physical separation of the core network control and user planes. The actor-critic method is employed to solve this problem for three proposed core network operation configurations, each offering an added degree of reliability and isolation over the default configuration that is currently standardized by the 3GPP.

Looking ahead to potential future research directions, optimizing carrier aggregation-based resource allocation across triple-band sliced access networks emerges as a promising avenue. Additionally, the integration of coordinated multi-point techniques with carrier aggregation in multi-UAV NR aerial networks is especially challenging. The introduction of added carrier frequencies and channel bandwidths, while enhancing flexibility and robustness, complicates band-slice assignments and user-UAV associations. Another layer of intriguing yet complex research involves optimizing handovers in high-mobility UAV networks, where both users and UAVs are mobile. UAV trajectory planning, which is already NP-hard even in static-user scenarios, becomes even more intricate to obtain optimal solutions in high-mobility user cases.



To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.