Optimization of Handover, Survivability, Multi-Connectivity and Secure Slicing in 5G Cellular Networks using Matrix Exponential Models and Machine Learning

This works proposes optimization of cellular handovers, cellular network survivability modeling, multi-connectivity and secure network slicing using matrix exponentials and machine learning techniques. We propose matrix exponential (ME) modeling of handover arrivals with the potential to much more accurately characterize arrivals and prioritize resource allocation for handovers, especially handovers for emergency or public safety needs. With the use of a ‘B’ matrix for representing a handover arrival, we have a rich set of dimensions to model system handover behavior. We can study multiple parameters and the interactions between system events along with the user mobility, which would trigger a handoff in any given scenario. Additionally, unlike any traditional handover improvement scheme, we develop a ‘Deep-Mobility’ model by implementing a deep learning neural network (DLNN) to manage network mobility, utilizing in-network deep learning and prediction. We use the radio and the network key performance indicators (KPIs) to train our model to analyze network traffic and handover requirements. Cellular network design must incorporate disaster response, recovery and repair scenarios. Requirements for high reliability and low latency often fail to incorporate network survivability for mission critical and emergency services. Our Matrix Exponential (ME) model shows how survivable networks can be designed based on controlling numbers of crews, times taken for individual repair stages, and the balance between fast and slow repairs. Transient and the steady state representations of system repair models, namely, fast and slow repairs for networks consisting of multiple repair crews have been analyzed. Failures are exponentially modeled as per common practice, but ME distributions describe the more complex recovery processes. In some mission critical communications, the availability requirements may exceed five or even six nines (99.9999%). To meet such a critical requirement and minimize the impact of mobility during handover, a Fade Duration Outage Probability (FDOP) based multiple radio link connectivity handover method has been proposed. By applying such a method, a high degree of availability can be achieved by utilizing two or more uncorrelated links based on minimum FDOP values. Packet duplication (PD) via multi-connectivity is a method of compensating for lost packets on a wireless channel. Utilizing two or more uncorrelated links, a high degree of availability can be attained with this strategy. However, complete packet duplication is inefficient and frequently unnecessary. We provide a novel adaptive fractional packet duplication (A-FPD) mechanism for enabling and disabling packet duplication based on a variety of parameters. We have developed a ‘DeepSlice’ model by implementing Deep Learning (DL) Neural Network to manage network load efficiency and network availability, utilizing in-network deep learning and prediction. Our Neural Network based ‘Secure5G’ Network Slicing model will proactively detect and eliminate threats based on incoming connections before they infest the 5G core network elements. These will enable the network operators to sell network slicing as-a-service to serve diverse services efficiently over a single infrastructure with higher level of security and reliability.

Name thesis:degree_name

Ph.D. (Doctor of Philosophy)

Level thesis:degree_level

Doctoral

Discipline thesis:degree_discipline

Telecommunications and Computer Networking (UMKC)

Grantor

University of Missouri--Kansas City

Year dc:date.issued

2022

Author dc:creator

• Paropkari, Rahul Arun

Advisor dc:contributor.advisor

• Beard, Cory

Handle dc:identifier.uri

https://hdl.handle.net/10355/93878

OAI identifier oai:identifier

oai:mospace.umsystem.edu:10355/93878