Banca de DEFESA: Yarisley Peña Llerena
Uma banca de DEFESA de DOUTORADO foi cadastrada pelo programa.STUDENT : Yarisley Peña Llerena
DATE: 20/12/2022
TIME: 08:30
LOCAL: Virtual
TITLE:
Controller Placement and Radio Resource Allocation for D2D Communications in 5G Wireless Networks
KEY WORDS:
Controller Placement; Radio Resource Allocation; D2D Communications; 5G Wireless Networks
PAGES: 165
BIG AREA: Engenharias
AREA: Engenharia Elétrica
SUBÁREA: Telecomunicações
SPECIALTY: Sistemas deTelecomunicações
SUMMARY:
Next generation cellular networks promise faster data transmission, higher bandwidth, and
lower latency and the Device-to-Device (D2D) communication technology helps make such promises
a reality, due to proximity services (ProSe) of communication devices. The advantages of such
communication include higher throughput, enhanced data rate, lower latency and energy consumption, fairness, and improved spectral efficiency. This research focuses on the influence of
the management of D2D communications on a cellular network. Two problems, namely Controller Placement Problem (CPP) and Radio Resource Allocation (RRA) were studied.
The technique proposed for solving the former involves the obtaining of number of D2D communications to be managed and then number of controllers required, their physical location in
the cellular infrastructure, and the best evolved NodeB (eNB) assignment for them. The problem
was modeled as an optimization problem one and Artificial Bee Colony (ABC) and Ant Colony
System with External Memory (ACS-EM) meta-heuristics solved it. They were compared with
Ant Colony System (ACS) and Particle Swarm Optimization (PSO) algorithms, and the analysis
revealed the computational complexities of ABC and PSO are lower than of ACS-EM and ACS.
However, ABC and ACS-EM showed better performance in solving the problem, with characteristics that enable a more efficient exploration of the search space, thus avoiding sub-optimal
solutions. The best results were obtained with ABC, followed by ACS-EM, then ACS, and lastly
PSO.
Regarding the RRA problem, two solution methods were proposed towards obtaining the
number of D2D connections admitted in the network. In the first RRA solution method, the optimization problem is solved by the Social-aware RRA Artificial Bee Colony (SA-RRA-ABC),
considering the social relationship between users, a feedback scheme, and maximization of the
system throughput. A selected-NM Maximum Distance Ratio (MDR) q-bit feedback scheme
designed reduces feedback overhead, since each D2D receiver sends only q-bit feedback Channel State Information (CSI) among N cellular User Equipment (CUE) and M D2D pairs with
the largest MDR metric. SA-RRA-ABC was validated through simulations and compared with
Greedy Resource Allocation Algorithm (GRAA) and Social-aware Greedy Resource Allocation
Algorithm (SA-GRAA). The simulation results showed its better performance. The selected-NM
q-feedback model proposed can achieve performance close to that of the full CSI model with
lower overhead.
The second RRA solution method deals with the optimization problem in a D2D cellular
network that offers Ultra-reliable and Low Latency Communication (URLLC) services for the
sending short packets directly to their destination, thus maximizing the network energy efficiency.
The problem is solved by three bioinspired algorithms, namely ABC, ACS-EM, and PSO, which
take into account interference when cellular and D2D users use a same radio resource. The metaheuristics were compared with a greedy heuristic and an exhaustive search algorithm and the
analysis revealed the computational complexity of Greedy is the lowest and those of ABC and
PSO are lower than that of ACS-EM. However, ABC showed better performance in solving the
problem, followed by ACS-EM, then PSO, and lastly Greedy heuristic.
BANKING MEMBERS:
Externo ao Programa - 2279048 - CARLOS HUMBERTO LLANOS QUINTERO
Externo à Instituição - LUIZ CARLOS PESSOA ALBINI - UFPR
Presidente - 2376576 - PAULO ROBERTO DE LIRA GONDIM
Externo à Instituição - RENATO MARIZ DE MORAES - UFPE