Ανάλυση και βελτιστοποίηση της επίδοσης ασύρματων δικτύων επόμενης γενιάς
Performance analysis and optimization of next generation wireless networks
Keywords5G Vehicular Cloud Computing (5G-VCC) ; Handover ; Long Term Evolution (LTE) ; Medium Access Control (MAC) ; Mobile Edge Computing (MEC) ; Scheduling ; Software Defined Networking (SDN) ; Software Defined Vehicular Architectures (SDN-V) ; Ultra Dense Networking (UDN) ; Vehicle to Infrastructure (V2I) communication ; Vehicle to Pedestrian (V2P) communication ; Vehicle to Vehicle (V2V) communication ; Vehicles using Cloud (VuC) ; Vehicles using Fog (VuF) ; Vehicular Cloud (VC) ; WiMAX ; Wireless Access for Vehicular Environment (WAVE) ; 5G wireless networks ; Cloud computing ; Fog computing ; Mobility management ; Resource manipulation ; Vehicular networks
The Fifth Generation (5G) networks, including the 5G Vehicular Cloud Computing (5G-VCC) systems, have evolved rapidly offering multiple services to users. The operating principles of vehicular networks, Cloud Computing (CC), Fog Computing (FC), Mobile Edge Computing (MEC) and Software Defined Networks (SDN) are applied to 5G infrastructures. In a 5G-VCC system, the vehicles are equipped with On-Board Units (OBUs) which communicate with each other as well as with Road Side Units (RSUs). Each RSU interacts with a Cloud infrastructure which offers vehicular services with strict Quality of Service (QoS) requirements, including Driver Assistance (DA), Passengers Entertainment and Information (PEnI) and Medical (MED) services. Dense deployments of 5G access networks are also implemented, called Ultra Dense Networks (UDNs), aiming to support high data rates produced by an increased number of vehicular users. In this environment, heterogeneous technologies are used to transfer the network services to vehicles. Optimal manipulation of the communication resources is required, while at the same time vehicular users should always obtain connectivity to the most appropriate network access technology, in order the constraints of the vehicular services to be satisfied. In this thesis, existing schemes for resource allocation as well as for mobility management are studied, while novel solutions are proposed for each topic. Initially, the theoretical background of the 5G wireless networks and 5G-VCC systems is described. Subsequently, the available delivery models for providing cloud services in 5G infrastructures are mentioned, while the available 5G-VCC architectures and the communication models are presented. Regarding the topic of the manipulation of the available network resources in 5G-VCC systems, the available Medium Access Control schemes (MAC schemes) are studied, while at the same time resource scheduling algorithms implemented at the MAC layer of 5G-VCC systems are described. Subsequently, two novel solutions are proposed to optimize the resource allocation process in modern networks. The first one is called FLS Advanced (FLSA) and optimizes the resource allocation for real-time services. Additionally, the second one is called FLS Advanced - Cross Carrier (FLSA-CC) and is specialized for LTE-Advanced (LTE-A) access networks where carrier aggregation is applied in order to provide widened bandwidth to the users. The evaluation of the proposed algorithms is performed through extended experiments. Simulation results show that the proposed algorithms outperform existing solutions. The mobility management requirements arising at the 5G-VCC systems are also studied. As a result of this study, initially three methods for calculating the significance of the criteria affecting the mobility management are proposed, namely the Trapezoidal Fuzzy Analytical Network Process (TF-ANP), the Trapezoidal Fuzzy Adaptive Analytical Network Process (TF-AANP) and the Pentagonal Fuzzy Analytical Network Process (PF-ANP). Subsequently, three network selection algorithms are proposed, namely the Trapezoidal Fuzzy TOPSIS (TFT), the Trapezoidal Fuzzy TOPSIS with Adaptive Criteria Weights (TFT-ACW) and the Pentagonal Fuzzy TOPSIS (PFT). Following this study and regarding the research results of the implemented algorithms, a mobility management scheme is proposed. This scheme takes into consideration the Signal to Noise plus Interference (SINR) parameters as well as the velocity of the vehicular users to evaluate the necessity of performing a handover. Then, it applies the PFT algorithm to select the most appropriate network for the user. Experimental results showed that the proposed scheme outperforms the existing solutions described in the research literature.