More Like this

1. Deep Neural Network Based Dynamic Resource Reallocation of BBU Pools in 5G C-RAN ROADM Networks

   Mo, Weiyang; Gutterman, Craig L.; Li, Yao; Zussman, Gil; Kilper, Daniel C. (January 2018, Optical Fiber Conference)

   An LSTM network is developed to predict BBU pool traffic in 5G C-RAN ROADM networks. 5G throughput improvement and resource savings are observed with resource reallocation by reconfiguring the optical network 30 minutes in advance. OCIS codes: (060.4256) Networks, network optimization; (060.0060) Fiber optics and optical communications 

   more » « less
2. Machine Learning-based Spectrum Resource Assignment and End-to-End Path Reconfiguration for Flexible Optical Networks

   https://doi.org/10.1109/ICTON62926.2024.10648205  

   Hirota, Yusuke; Goto, Yuta; Furukawa, Hideaki (July 2024, IEEE)

   Optical resource management and path reconfiguration have become increasingly paramount in the era of dynamic and flexible optical networks. This paper presents machine learning-based optical resource assignment for highly efficient resource utilization in Spatial Division Multiplexing (SDM) networks. We also demonstrate a network orchestration with a Whitebox-based optical network and failure prediction in an SDM network. 

   more » « less
3. DeepCare: Deep Learning-Based Smart Healthcare Framework using 5G Assisted Network Slicing

   https://doi.org/10.1109/ANTS56424.2022.10227802  

   Basu, Deborsi; Krishnakumar, Vikram; Ghosh, Uttam; Datta, Raja (August 2023, IEEE)

   5G and beyond communication networks require satisfying very low latency standards, high reliability, high- speed user connectivity, more security, improved capacity and better service demands. Augmenting such a wide range of KPIs (Key Performance Indicators) needs a smart, intelligent and programmable solution for TSPs (Telecommunication Service Providers). Resource availability and quality sustainability are challenging parameters in a heterogeneous 5G environment. Programmable Dynamic Network Slicing (PDNS) is a key technology enabling parameter that can allow multiple tenants to bring their versatile applications simultaneously over shared physical infrastructure. Latest emerging technologies like virtualized Software- Defined Networks (vSDN) and Artificial Intelligence (AI) play a pivotal supporting role in solving the above-mentioned constraints. Using the PDNS framework, we have proposed a novel slice backup algorithm leveraging Deep Learning (DL) neural network to orchestrate network latency and load efficiently. Our model has been trained using the available KPIs and incoming traffic is analyzed. The proposed solution performs stable load balancing between shared slices even if certain extreme conditions (slice unavailability) through intelligent resource allocation. The framework withstands service outage and always select the most suitable slice as a backup. Our results show latency-aware resource distribution for better network stability. 

   more » « less
4. Helix: A RAN Slicing Based Scheduling Framework for Massive MIMO Networks

   An, Qing; Pandey, Divyanshu; Doost-Mohammady, Rahman; Sabharwal, Ashutosh; Shakkottai, Arinivas (July 2024, arXivorg)

   An important aspect of 5G networks is the development of Radio Access Network (RAN) slicing, a concept wherein the virtualized infrastructure of wireless networks is subdivided into slices (or enterprises), tailored to fulfill specific use-cases. A key focus in this context is the efficient radio resource allocation to meet various enterprises’ service-level agreements (SLAs). In this work, we introduce Helix: a channel-aware and SLAaware RAN slicing framework for massive multiple input multiple output (MIMO) networks where resource allocation extends to incorporate the spatial dimension available through beamforming. Essentially, the same time-frequency resource block (RB) can be shared across multiple users through multiple antennas. Notably, certain enterprises, particularly those operating critical infrastructure, necessitate dedicated RB allocation, denoted as private networks, to ensure security. Conversely, some enterprises would allow resource sharing with others in the public network to maintain network performance while minimizing capital expenditure. Building upon this understanding, Helix comprises scheduling schemes under both scenarios: where different slices share the same set of RBs, and where they require exclusivity of allocated RBs. We validate the efficacy of our proposed schedulers through simulation by utilizing a channel data set collected from a real-world massive MIMO testbed. Our assessments demonstrate that resource sharing across slices using our approach can lead up to 60.9% reduction in RB usage compared to other approaches. Moreover, our proposed schedulers exhibit significantly enhanced operational efficiency, with significantly faster running time compared to exhaustive greedy approaches while meeting the stringent 5G sub-millisecond-level latency requirement. 

   more » « less
5. Helix: A RAN Slicing Based Scheduling Framework for Massive MIMO Networks

   https://doi.org/10.1145/3696399  

   An, Qing; Pandey, Divyanshu; Doost-Mohammady, Rahman; Sabharwal, Ashutosh; Shakkottai, Srinivas (December 2024, Proceedings of the ACM on Networking)

   An important aspect of 5G networks is the development of Radio Access Network (RAN) slicing, a concept wherein the virtualized infrastructure of wireless networks is subdivided into slices (or enterprises), tailored to fulfill specific use-cases. A key focus in this context is the efficient radio resource allocation to meet various enterprises' service-level agreements (SLAs). In this work, we introduce Helix: a channel-aware and SLA-aware RAN slicing framework for massive multiple input multiple output (MIMO) networks where resource allocation extends to incorporate the spatial dimension available through beamforming. Essentially, the same time-frequency resource block (RB) can be shared across multiple users through multiple antennas. Notably, certain enterprises, particularly those operating critical infrastructure, necessitate dedicated RB allocation, denoted as private networks, to ensure security. Conversely, some enterprises would allow resource sharing with others in the public network to maintain network performance while minimizing capital expenditure. Building upon this understanding, Helix comprises scheduling schemes under both scenarios: where different slices share the same set of RBs, and where they require exclusivity of allocated RBs. We validate the efficacy of our proposed schedulers through simulation by utilizing a channel data set collected from a real-world massive MIMO testbed. Our assessments demonstrate that resource sharing across slices using our approach can lead up to 60.9% reduction in RB usage compared to other approaches. Moreover, our proposed schedulers exhibit significantly enhanced operational efficiency, with significantly faster running time compared to exhaustive greedy approaches while meeting the stringent 5G sub-millisecond-level latency requirement. 

   more » « less