Wireless Communications Group

A Novel Approach to 5G Handover

Researchers: Dr. Muhieddin Amer & Dr. Omar Abdul Latif

The evolution of mobile networks towards 5G and Ultra Dense Networks (UDNs) brings new complexities to the process of handovers, which is the transfer of an ongoing connection from one base station (gNB) to another as users move from one point to another. These challenges include frequent handovers or a “ping pong” effect, increased signaling overhead, and issues with interoperability, especially with the adoption of mmWave frequencies. To address these issues, our project explores a Novel Approach to 5G handover optimization using artificial intelligence to predict the movement of users and ensure seamless and efficient transitions from gNBs, ultimately improving the user experience in next-generation networks. The implementation considers other performance-related factors such load balancing across neighboring cells and subjecting the cell-edge throughput to some desired minimum value.

Channel Estimation Framework for 6G Intelligent Reflecting Surface-Enabled MIMO 

Researchers: Dr. Muhieddin Amer, Dr. Omar Abdul Latif, and Nazia Begum

The Intelligent Reflecting Surface (IRS) serves as a technology enabling passive manipulation of wave properties like amplitude, frequency, phase, and polarization through reflection. This technology is poised to revolutionize wireless communication by enhancing spectrum and energy efficiency while demanding minimal energy consumption. However, in scenarios where an IRS assists a base station (BS) with multiple antennas and user equipment (UE) with a single antenna, obtaining Instantaneous Channel State Information (I-CSI) for every link at the IRS poses challenges due to the numerous reflective elements and passive operation of the IRS. This imposes additional burdens on the system, necessitating the integration of radiofrequency chains into the IRS system. To address this, the paper proposes a three-phase pilot-based channel estimation framework for uplink multiuser communications, utilizing modular redundancy to reduce the time needed for channel estimation. The framework leverages IRS to achieve this goal by estimating the direct channels between UEs and BSs, as well as the reflected channels between a single UEs, IRS, and BS.

Implementation of Network Slicing and IoT Integration in 5G

Researchers: Dr. Muhieddin Amer, Dr. Omar Abdul Latif, Dr. Andres Kwasinski, and Nazia Begum

This $30k funded-research project explores the use of 5G testbed system to implement network slicing and IoT integration. Creating private network is a key promise of 5G and Beyond 5G (B5G) systems where several networks are virtually multiplexed on the same network physical infrastructure. The project will focus on the allocation of the network resources that realizes the technology of network slicing. With network slicing, the communication resources in a cellular network (e.g. bandwidth, timeslots, and transmit power) are distributed among different groups of wireless connections, each associated with a type of cellular traffic, so that each connection meets the performance required by each type of traffic. Machine learning/artificial intelligence techniques play a crucial role in two aspects of this problem: (1) they enable the different decision-making elements in the network to autonomously  gain  awareness  of the  network  condition,  resource  availability,  and traffic requirements, and (2) it allows the same decision-making elements to learn to make the resource allocation decisions at the heart of the network slicing technology.

Mitigating Intercell Interference (ICI) in 5G Cellular Networks

Researchers: Dr. Omar Abdul Latif

In modern cellular networks, intercell interference poses a significant challenge, degrading overall network performance and user experience. This paper presents a novel approach to mitigating intercell interference through the use of artificial intelligence (AI) to dynamically control base station transmission power. Our proposed system employs machine learning algorithms to analyze real-time network conditions and predict optimal transmission power levels for each base station. By continuously adjusting power levels, the system minimizes interference and maximizes spectral efficiency. Simulation results demonstrate significant improvements in key performance metrics, including signal-to-interference-plus-noise ratio (SINR), data throughput, and energy efficiency, compared to traditional static and rule-based power control methods. Our findings indicate that AI-driven power control can effectively reduce intercell interference, enhancing network capacity and user satisfaction. The proposed solution is particularly beneficial in dense urban environments where interference is more pronounced. This research paves the way for more intelligent and adaptive network management strategies, highlighting the potential of Artificial Intelligence (AI) to revolutionize future wireless communication systems.

End-to-End Network Slicing using Hypergraph Theory

Researchers: Dr. Muhieddin Amer, Dr. Omar Abdul Latif, Dr. Andres Kwasinski

Network  slicing is based on the concept of network virtualization and, when it reaches maturity, is expected to result in complete softwarization of 5G, Beyond-5G (B5G) and 6G networks. This means that future networks will only need minimal physical infrastructure upgrades (mostly in the frontend of the network). Network slicing is identified as one of the key enablers of next generation wireless mobile networks due to its ability to multiplex virtualized and independent architectures on the same physical network infrastructure . The virtual architectures instantiated through network slicing can be tailored to the technical requirements of specific verticals or applications. However, there is still the challenge of providing type -specific mechanism to generate and provision the virtual networks (i.e. network slices) that are tailor-made for specific applications. This challenge is currently an active research topic in the field of wireless communication networks. In this research work, three end-to-end network slicing provisioning frameworks are proposed and investigated.

End-to-End Network Slicing using SNN and SVM

Researchers: Dr. Muhieddin Amer, Dr. Omar Abdul Latif, Dr. Andres Kwasinski

The advent of 5G has reinforced network slicing as a transformative mechanism to deliver customized, end-to-end virtual networks over shared physical infrastructure. However, efficiently provisioning and managing these slices in real-time remains a significant challenge due to the dynamic and heterogeneous nature of 5G service requirements. This research work proposes a hybrid intelligent framework that leverages Spiking Neural Networks (SNN) and Support Vector Machines (SVM) to optimize and automate end-to-end network slicing provisioning. SNNs are employed to model temporal network traffic patterns and respond to asynchronous events with low latency, while SVMs are utilized for accurate classification of service types and prediction of optimal resource allocations. The integration of these machine learning models enables adaptive, scalable, and predictive slice management, reducing operational complexity and enhancing Quality of Service (QoS). Results demonstrate the effectiveness of the proposed approach in improving resource utilization and provisioning speed, making it a viable solution for intelligent 5G network orchestration.

PAPR Reduction in 5G OFDM Systems

Researchers: Dr. Muhieddin Amer, Naziya Begum

5G networks employ multicarrier modulations (MCM)s such as filtered-orthogonal frequency division multiplexing (F-OFDM) and universal filtered orthogonal frequency division multiplexing (UF-OFDM) as a solution to overcome the challenges of high data rates and spectral efficiency [1]. However, MCMs have high peak to average power ratio (PAPR) which drives the power amplifier (PA) in the linear region resulting in the reduced efficiency. To overcome this problem PAPR should be reduced [1-4]. In this paper, precoding based PAPR reduction techniques such as Discrete Fourier Transform (DFT), Discrete Cosine Transform DCT) and Zadoff-Chu Transform (ZCT) are implemented using MATLAB for F-OFDM and UF-OFDM systems. Comparison analysis shows that Zadoff-chu Transform precoding technique for PAPR reduction gives better results. Hence, ZCT precoding is proposed for both F-OFDM and UF-OFDM systems. Simulation results show that proposed technique lowers down the power spectral density (PSD) tails at the PA output, reduces PAPR and instantaneous to average power ratio (IAPR) and conserves the bit error rate (BER) in the AWGN channel.

5G in Autonomous Driving

Researchers: Dr. Muhieddin Amer, Mahmoud Wadi

This project discusses the evolution of autonomous driving supported with a 5G network mainly focusing on key enabler technologies such as enhanced mobile broadband (eMBB), massive machine type communication (mMTC), and ultra-reliable low latency communications (URLLC). Autonomous driving (AD) is involved in safety applications so the communication system has to be reliable while exchanging communication in real-time with low latency. Next, the system architecture of AV will be discussed in depth followed by a brief about vehicular ad-hoc networks and their contribution towards building a reliable system of Autonomous driving. Different challenges will be discussed related to AV and future research directions will be provided.

5G Heterogeneous Network Planning Using AI

Researchers: Dr. Muhieddin Amer, Jonathan Chellappa

Cellular networks suffer from limitations due to capacity and from RF coverage especially in higher frequency bands. To address limitations, network densification (deployment of Small Cells (SC) or Femto cells inside Macro cells) are being implemented. This hybrid combination of networks is referred to as Heterogenous Networks (HetNets). HetNets deployment increases the complexity involved in the configuration and management of the network as well as issues such as SC interference which introduces the need for Self-Optimizing Networks (SON) that are both adaptable and autonomous while being able to continuously improve their performances independently. 5G network while boasting tremendous gains in data rates and latency improvement, require service providers to implement a wide variety of techniques to meet these demands. Implementation of AI techniques that can process this large volume of data can significantly improve the efficiency in managing such a vast interconnected network and this project aims to discuss and implement machine learning AI algorithms that can predict and improve network throughput and user Quality of service (QoS) focusing primarily on planning and orchestrating HetNet clustering solutions.