Mobility and Resource Management for 5G Small Cell Networks

Abstract

Today, the deployment of moving and fixed small cells is becoming popular. However, to achieve the Quality of Service (QoS) required by the ever-increasing services in telecommunications, many issues must be considered. The main issues are user association, resource, and mobility management of the network. In our thesis, we focus on handling those issues separately and then combine them to enhance the QoS provided to each User Equipment (UE). In the first part, a novel Chance-Based Deferred Acceptance Matching (CBDAM) algorithm is proposed. A matching game-based user association scheme is introduced for solving the assignment problem. The new proposed approach maximizes the number of users admitted in the system (handoff and new user association request) and the total system through put by assigning each user (macro or vehicular) to the stations based on negotiations between eNBs and UEs. It also significantly decreases the handover failure rate for vehicular users (VUEs). In the second part, another aspect is investigated in order to enhance the small cell network performance, which is the mobility load balancing technique. A Utility-Based Mobility Load Balancing (UMLB) algorithm is introduced to balance the load across a small-cell network by considering the operator utility and the user utility for the handover process. The operator utility is calculated for each potential handover based on the load of the neighbouring small cells. Whereas, the user utility calculation is based on the sigmoid function by considering different criteria. Also, we presented a new term named load balancing efficiency factor (LBEF). The LBEF considers a load of neighbouring cells and the edge-UEs for each overloaded cell. This factor specifies the sequence of overloaded cells for the UMLB algorithm operation. In the third part, the problem of increasing number of handovers, which are performed in the small cell networks as a result of adopting a Mobility Load Balancing (MLB) algorithm, is studied. The handover decision of UEs is based on the classification of the candidate UEs within an overloaded cell. Some handovers are completely avoided; others are made to the macro cell to minimize the frequent handovers. The UMLB-HO aims to minimize the standard deviation with a minimum number of handovers compared to UMLB algorithm. Finally, the proposed solutions are evaluated using extensive simulations, and a comparison with related works found in the literature is introduced.

Date	27 Apr 2020
Original language	American English
Awarding Institution	École de technologie supérieure
Supervisor	Michel Kadoch (Supervisor)