The mobile packet core – a central part of the overall mobile cellular network – has a long history of evolution. Through the years, its architecture has been drastically changed to meet the demand coming from the fast growth in the number of devices as well as the introduction of new types of applications and services. On one hand, the number of new devices and subscribers keeps increasing at an unprecedented rate, which can still give rise to scalability issues in the mobile packet core if it is not properly managed. On the other hand, the introduction of new types of services brings with it a new set of requirements, such as low-latency and high reliability. The network softwarization is widely considered as a promising approach to address these two challenges.

This thesis focuses on enhancing the scalability of a softwarized mobile packet core network for 5G and beyond, and the communication latency it provides. Moreover, the thesis provides an extensive survey of existing softwarized mobile packet core network solutions, identifying important questions and gaps for future research. This thesis also explores the possibility of leveraging the network softwarization concept to design a softwarized mobile packet core network to support multicast and broadcast services. In order to tackle the scalability issue in a softwarized mobile packet core, the thesis proposes several dynamic and adaptive load-balancing algorithms to efficiently manage the traffic load in both the control and the user plane. These load-balancing algorithms take into account different factors such as the current load of virtualized network functions and the estimation of the communication latency. On the latency aspect, the thesis studies the feasibility of using a softwarized mobile packet core for delivering time-critical messages in a smart-grid environment, and proposes several deployable communication solutions to support the study.