Designing an Advanced Mobility Management and Utilization Framework for Enabling mmWave Multi-Band Ultra-Dense Cellular Networks of Future

Title: Designing an Advanced Mobility Management and Utilization Framework for Enabling mmWave Multi-Band Ultra-Dense Cellular Networks of Future.
Funding: This project is funded by the National Science Foundation of USA, under Award No. 1718956.
Duration: Three years (October 1, 2017, to September 30, 2020)
Funding Amount: $ 500,000.00

The problem addressed by the project

Providing connectivity on the go i.e., supporting mobile users is the key reason the mobile cellular networks exist. To support user mobility cellular networks are required to have the capability for seamlessly executing handovers from one cell to another cell. Each handover incurs a certain cost for the network as well as the users. The most significant of these costs include:

  1. Signalling overhead
  2. Throughput loss due to low SINR before and during the handover
  3. Latency
  4. UE battery drainage in cell discovery and measurements, and
  5. Risk of call drop due to handover failure

As the number of handovers increases in a network, all of these costs increase. The number of handovers in a given network compound mainly with the following three factors:

  1. Shrinking cell sizes.
  2. Increase in the number of mobile users compared to static users.
  3. Increase in the average speed of the users e.g. due to highly mobile devices needing connectivity such as those manifested by autonomous cars, bullet trains and UAVs etc.

As all of the above three factors are on the rise, the costs associated with handovers are expected to increase substantially in emerging networks. Maintaining resource efficiency and quality of experience expected from emerging networks vis-a-vis 5G and beyond is a challenge and requires a rethinking of the current handover process in cellular networks.

Another factor that makes current mobility management and handover protocols inadequate for emerging networks is the advent of mmWave cells. Low-frequency bands allow graceful overlap among cells to provide enough time margin for Handover to be conducted. In contrast, mmWave cell coverage may end abruptly, giving little to no time for executing handover.

These challenges together require a paradigm shift in mobility management in emerging and future cellular networks i.e. 5G and 6G.

Project Goals

The overarching goal of this project is to design, develop and evaluate an Advanced Mobility Management Framework for emerging and future cellular networks. This framework is aimed to act as a key enabler for seamless, resource-efficient and Quality of Experience aware mobility management in emerging ultra-dense multi-band networks. The framework is also aimed to act as a key enabler for Ultra Reliable Low Latency Connectivity (URLLC), particularly for use cases that involve mobility such as autonomous cars, UAVs, health-care applications on the go and bullet trains etc. The main idea in the proposed project is to develop a machine learning-based framework to transform mobility management from being a reactive process to a proactive process. To realize this framework project has the following three interlinked goals:

  1. The first goal in this project is to develop robust machine learning-based models to predict certain attributes of user mobility in cellular networks. These attributes include but are not limited to candidate cell and frequencies for next handover and future traffic hotspots and cell loads.
  2. The second goal is to exploit these machines learning-based mobility and load prediction models for developing novel algorithms, protocols and solutions for proactive mobility robust optimization (P-MRO) and proactive mobility load balancing (P-MLB). This goal also includes using these prediction models to develop other proactive SON functions such as proactive energy efficiency SON.
  3. The third goal is to validate the accuracy limits of the machine learning-based prediction models. This goal also includes the analysis of the performance bounds of the proactive SON solution that leverage these models.

Active Members

PI: Dr. Ali Imran (Principal Investigator)

Co-PI: Dr. Hazem Refai

GRA: Asad Zaidi (Ph.D. student)

GRA: Marvin Manalastas (Masters graduate/Ph.D. student)

GRA: Waseem Raza (Ph.D. student)

Fulbright Ph.D. Scholar: Aneeqa Ijaz (Ph.D. student)

Fulbright Ph.D. Scholar: Muhammad Umar Bin Farooq (Ph.D. student)

Past Members

Post Doctoral Fellow: Dr. Hasan Farooq

GRA: Hasan Farooq (Ph.D. graduate)

GRA: Azar Taufique (Ph.D. graduate)

Academic Collaborators

Prof. Muhmmad Imran, Rural 5G Hub, University of Glasgow, UK.

Prof. Rahim Tafazolli, 5GIC, University of Surrey, UK.

Industry Collaborators

US Cellular

AT&T

Fujitsu

Integration of project outcomes into teaching

The project outcomes are being adapted into following two courses being taught by PI.

  1. "Emerging Topics in 5G and Beyond" Offered in Sring 2018, 2019 and 2020.
  2. "Telecommunication Technologies" Offered in Fall 2017, 2018, 2019.

K-12 Outreach Program

  • PI is running a K-12 outreach program with Booker T. Washington high school, Tulsa. The majority of students in this school belong to underrepresented communities in STEM.
  • Through this program several high school students have participated in research being carried out as part of this project and other NSF funded projects.
    • For the details of this K-12 outreach program and how to participate see:
    Internship opportunities for K-12

Tutorials

  1. Ali Imran, Muhammad Ali Imran, Moving Towards Zero-Touch Automation, A Key Enabler for 6G: The Challenges & Opportunities a half day tutorial at IEEE Wireless Communications and Networking Conference, May 25-28, 2020.
  2. Ali Imran, Muhammad Ali Imran, "LEAP for IoT- LEAN, Elastic, Agile and Proactive (LEAP) Wireless Networks for Enabling IoT", a half day tutorial at 2017 IEEE PIMRC, Montreal, 8-13 Oct 2017.
  3. Ali Imran, "Can mmWave, massive MIMO and densification suffice to meet capacity and QoS requirements in emerging IoT?", half day seminar at International Conference on Communications Technologies (ComTech-2017), Pakistan, 19-21 October, 2017.
  4. Ali Imran "CDSA and BSON: The Two Key Enablers of Lean, Elastic and Proactive Wireless Networks Needed for Future IoT", a half day tutorial to be delivered at 2016 IEEE 3rd World Forum on Internet of Things, Washington DC, 6-8 Dec, 2017.

Keynotes/Invited Talks

  1. Ali Imran "Leveraging AI for Zero-Touch Automation in 6G: How to Address the Training Data Sparsity/Scarcity Challenge?" keynote at IEEE 2nd International workshop on Data Driven Intelligence for Networks and Systems workshop at Infocom 2020, July 6-9, Toronto, Canada.
  2. Ali Imran "Towards Zero Touch Automation in Emerging Wireless Networks", keynote at 13th IEEE International Conference on Open Source System and Technologies 17-19 Dec, 2019.
  3. Ali Imran "Addressing the Hyper Parameterization Challenge in AI for Wireless Networks", panel talk at, IEEE Globecom, Dec, 9-13-2019
  4. Ali Imran "What AI Can Deliver for 5G and Beyond", keynote at 5G NA, May 8th-9th , 2019, Denver, USA
  5. Ali Imran "Role of AI in Emerging Networks: 5G and Beyond", keynote at 12th IEEE International Conference on Open Source Systems and Technologies, 19-21, Dec, 2018 (IEEE ICOSST 2018), Lahore, Pakistan.
  6. Ali Imran "AI for RAN automation: do's and don'ts for industry", plenary talk at RAN USA, Dec 3, 2018, Silicon Valley.
  7. Ali Imran "Challenges in Use of AI for Network Automation and How to Address These Challenges", invited talk at Fujitsu Laboratories, Europe, July 11, 2018, London, UK.
  8. Ali Imran "Leap Towards Zero Touch RAN Automation", invited talk, at Telekom Austria HQ, Jun 25, 2018, Vienna, Austria
  9. Ali Imran "How AI Will Transform the Future of RAN", invited talk at AT&T Campus, Silicon Valley , April 17, 2018, San Romano, CA, USA.
  10. Ali Imran "Future of Open Source Software Defined Big Data Enabled RAN", keynote at 11th international conference on open source system and technologies, Dec 18-20, 2017, Lahore, Pakistan.
  11. Ali Imran "Next Generation Artificial Intelligence Based RAN", invited talk in an industry panel at RAN USA, Dec 4, 2017, Silicon Valley, USA.

International Collaboration Opportunities

  • Project students gained international collaboration experience by working with collaborators at:
    1. 5GIC, Surrey, UK.
    2. The University of Glasgow. UK.
    3. The University of Leads, UK.

Conference Attendance by Students

  • Project students have attended and presented their work at following conferences
    1. IEEE 17th Annual Consumer Communications and Networking Conference (CCNC), Las Vegas, NV, Jan 2020.
    2. IEEE ICC, Kansas, USA, 20-24 May, 2018.

Industry Internships

  • Project students have interned with the following industry collaborators:
    1. New T-Mobile, CT in Spring 2020.
    2. Sprint, CT in Fall 2019.
    3. MobileComm, Dallas in Summer 2019.
    4. Bell Labs, NJ in Summer 2018.
    5. Phazr, Dallas in Fall 2017.
Home Publications Projects Team Open Positions Outreach Contact © 2019-20 AI4Networks