ASSH: Designing Agile and Scalable Self-Healing Functionalities for Ultra Dense Future Cellular Networks

  • Cell densification is emerging as one of the key means to gain 1000x target capacity gain in the fifth generation (5G) cellular systems. This emerging reality poses a major challenge for the cellular industry: How to manage the susceptibility of an ultra-dense, extremely complex 5G network to a potentially high cell outage rate?
    Definition: A complete (or partial) cell outage is a scenario when either Base Station (BS) hardware and/or software malfunctions or when one or more cell parameters become misconfigured during network operations. Partial outage refers to scenarios when the cell continues to operate but its performance degrades below its typical level. In this proposal, the term cell outage refers to both partial and complete cell outages. The rate of outages is intrinsically proportional to cell density, and complexity of hardware and software that constitute the radio access network. Both of these factors have been consistently on rise from 1G to 4G, and trend is expected to continue for 5G. Currently,cellular carriers in the US alone currently spend over $15 billion annually to manage cell outages. In current cellular networks, drive tests or hardware fault alarms are employed for detecting cell outage; transitory cell outage compensation in the affected area is accomplished with makeshift cell-on-wheel. Such semi-manual approaches to cell outage management have proven inadequate and highly inefficient even for today’s operators, making them unfeasible for sustaining future cellular networks marked by ultra-dense cell deployment and mounting operational complexity. If no intervening measures are taken, cell outage management will be a primary challenge for future cellular networks, such as 5G.
    The ASSH project aims to address this challenge by developing an Advance Cell Outage Management (ACOM) framework for automating cell outage detection and compensation in future ultra-dense, heterogeneous cellular networks, thereby equipping them with fully self-healing functionality. ACOM integrates three novel schemes: 1) Autonomous highly agile, Macro Cell Outage Detection (MOD); 2) Autonomous Small Cell Outage Detection (SOD); and 3) Autonomous Heterogeneous Cell Outage Compensation (HOC). ACOM will provide solution for not only complete outages, which are easy to detect, but partial outages i.e. sleeping cells. Sleeping cells refer to scenarios where cells remain ON, but certain of its KPIs fall below the typical level.
    A large number of technical challenges are anticipated in development of ACOM. These challenges will be addressed by leveraging analytical tools from machine learning, big data analytics, optimization, chaos theory and game theory paradigms, by building on our past experience in this domain gained e.g. under QSON project.
    If this work sounds interesting, contact PI Ali Imran for collaboration opportunities.

    • Prof. Ali Imran (Principal Investigator- University of Oklahoma)
    • Hasan Farooq (PhD candidate - University of Oklahoma)
    • Ahmad Asghar (PhD candidate - University of Oklahoma)

  • Intellectual Merit

    • Patent
    • 1. Ali Imran, Ahmad Asghar, Hasan Farooq, “Method For Enhancement Of Capacity And User Quality Of Service In Mobile Cellular Networks”. PCT 62/681,320 (Pending).

    • Journal
    • 1. A. Asghar, H. Farooq and A. Imran, "Self-Healing in Emerging Cellular Networks: Review, Challenges, and Research Directions," in IEEE Communications Surveys & Tutorials, vol. 20, no. 3, pp. 1682-1709, third quarter 2018. doi: 10.1109/COMST.2018.2825786.

      2. Ahmad Asghar, Hasan Farooq, Ali Imran “On Concurrent Optimization of Coverage, Capacity and Load Balance in HetNets through Joint Self-Organization of Soft and Hard Cell Association Parameters”, in the press for publication in IEEE Transactions on Vehicular Technologies, 2018 doi: 10.1109/TVT.2018.2846655.

      3. Azar Taufique, M. Abdelrahim, Ali Imran, Rahim Tafazolli, “On Analytical Modelling for Mobility Signalling in Ultra-dense HetNets," in the press for publication in IEEE Transactions on Vehicular Technology, 2018. doi: 10.1109/TVT.2018.2864627.

      4. Anwar Said, Syed Waqas Haider Shah, Hassan Farooq, Adnan Noor Mian, Ali Imran, Jon Crowcrof, “Proactive Caching at the Edge Leveraging Influential User Detection in Cellular D2D Networks, accepted in open access Journal on Future Internet. 10(10), 93; https://doi.org/10.3390/fi10100093.

      5. U. S. Hashmi, S. A. R. Zaidi and A. Imran, "User-Centric Cloud RAN: An Analytical Framework for Optimizing Area Spectral and Energy Efficiency," in IEEE Access, vol. 6, pp. 19859-19875, 2018. doi: 10.1109/ACCESS.2018.2820898.

      6. A. Zoha, A. Saeed, H. Farooq, A. Rizwan, A. Imran and M. A. Imran, "Leveraging Intelligence from Network CDR Data for Interference-Aware Energy Consumption Minimization," in IEEE Transactions on Mobile Computing, vol. 17, no. 7, pp. 1569-1582, 1 July 2018. doi: 10.1109/TMC.2017.2773609.

      7. Onireti, O., Imran, A. & Imran, M., “Coverage and Rate Analysis in the Uplink of Millimeter Wave Cellular Networks with Fractional Power Control” in EURASIP Journal on Wireless Communications and Networking (2018): 195. https://doi.org/10.1186/s13638-018-1208-0.

      8. H. Farooq, A. Asghar and A. Imran, "Mobility Prediction based Autonomous Proactive Energy Saving (AURORA) Framework for Emerging Ultra-Dense Networks," in IEEE Transactions on Green Communications and Networking. doi: 10.1109/TGCN.2018.2858011.

      9. S. Bassoy; H. Farooq; M. A. Imran; A. Imran, "Coordinated Multi-Point Clustering Schemes: A Survey," in IEEE Communications Surveys & Tutorials , vol.PP, no.99, pp.1-1, 2017.


    • Conference Papers
    • 1. U. S. Hashmi, S. A. R. Zaidi, A. Darbandi and A. Imran, "On the Efficiency Tradeoffs in User-Centric Cloud RAN," IEEE International Conference on Communications (ICC 2018), Kansas City, MO, 2018, pp. 1-7. doi: 10.1109/ICC.2018.8422228.
      2. Ahmad Asghar, Hasan Farooq, Ali Imran, “Novel Load-Aware Cell Association for Simultaneous Network Capacity and User QoS Optimization in Emerging HetNets”, IEEE 28th Annual International Symposium on Personal, Indoor, and Mobile Radio Communications (PIMRC), Montreal, QC, 2017, pp. 1-7. doi: 10.1109/PIMRC.2017.8292402.
      3. Sinasi Cetinkaya, Umair Hashmi, Ali Imran, “What User-Cell Association Algorithms Will Perform Best in mmWave Massive MIMO Ultra-Dense HetNets?”, IEEE 28th Annual International Symposium on Personal, Indoor, and Mobile Radio Communications (PIMRC), Montreal, QC, 2017, pp. 1-7. doi: 10.1109/PIMRC.2017.8292248.
      4.Yash Kumar, Hasan Farooq, Ali Imran, "Fault Prediction and Reliability Analysis in a Real Cellular Network" accepted in IWCMC 2017 Conference to be held june 26-30, 2017, Valencia, Spain.
      5. Umair Hashmi, Arsalan Darbandi, Ali Imran, “Enabling Proactive Self-Healing by Data Mining Network Failure Logs”, accepted in 2017 International Conference on Computing, Networking and Communications (ICNC).

    • Thesis
    • 1. The data analysis of CDR to enable proactive self-healing has also resulted in an MS thesis that was successfully defended in May 2017.
      2. Haneya Naeem Qureshi “Coverage Analysis of Unmanned Aerial Vehicles Based Netwrok”, Dec 2017.

    • Book Chapter
    • 1. Book chapter titled “ Continuous-Time Markov Chain-Based Reliability Analysis for Future Cellular Networks”, in book titled “ Big Data Applications in Telecommunications Industry. IGI Publisher, Feb, 2017.
      2. Book chapter titled “Spectral Efficiency Self-Optimization through Dynamic User Clustering and Beam Steering”, in book titled “ Big Data Applications in Telecommunications Industry. IGI Publisher, Feb, 2017.


    Broader Impact
    • Course Delivered by PI
    • 1. Course titled “Emerging Topics in LTE-A and 5G”. This course was taught by PI in Spring 2018 and was taken by all students working on the project.
      2. Courses on cellular system advance concepts including machine learning, Big data analytics were arranged for project students in Fall 2017.
      3. Courses on machine learning, Big data analytics and stochastic processes were arranged for project GRA in Fall 2016.

    • Tutorials
    • 1. PI presented a half-day tutorial at IEEE PIMRC, Oct 2017, Montreal Canada. This forum was used to disseminate this project's outcomes along with the outcomes of PI’s other NSF funded projects.
      2. PI gave a presentation at the 5GIC university of Surrey, during his NSF IRES funded visit to 5GIC in July 2018. 5GIC hosts over 150 researchers working on cellular communications and hosts a consortium of over 20+ wireless industries.
      3. PI gave a presentation at the University of Glasgow, during his NSF IRES funded visit to the UK in July 2018.
      4. PI organized a tutorial for delivery at IEEE WFIoT, Washington DC, Dec 2016.
      5. PI organized a tutorial for delivery at IEEE ICC, Paris, May 2017.
      6. PI gave a presentation at 5GIC university of Surrey, during his NSF IRES funded visit to 5GIC in July 2017.
      7. 10) PI gave a presentation at university of Glasgow, during his NSF IRES funded visit to UK in July 2017.

    • Keynotes/Invited Talks
    • 1. “Network automation: fundamental challenges, solution approaches and opportunities”, invited talk at Summer School Sponsored by IEEE ComSoc and HEC at LUMS, Aug 9, 2018, Lahore, Pakistan.
      2. “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.
      3. “ Leap Towards Zero Touch RAN Automation”, invited talk, at Telekom Austria HQ, Jun 25, 2018, Vienna, Austria.
      4. “On the role of AI in 5G and beyond”, invited plenary talk at 5G North America, May 14-16, 2018, Austin Texas.
      5. “How AI will transform the Future of RAN”, invited talk at AT&T Campus, April 17, 2018, San Romano, CA.
      6. “Towards next generation AI Enabled SON”, invited seminar at T-Mobile HQ, April 1, 2018, Seattle, WA.
      7. “Future of open source software-defined Big Data-Enabled RAN”, keynote at the 11th international conference on open source system and technologies, Dec 18-20, 2017, Lahore, Pakistan, http://icosst.kics.edu.pk/2017/
      8. “Next Generation Artificial Intelligence Based RAN”, invited talk in an industry panel at RAN USA, Dec 4, 2017, Silicon Valley, USA https://ran-usaevent.com/speakers/.
      9. "Big Data Empowered Self Organizing Networks, the game changing paradigm for enabling 5G", at the International Conference on Communications Technologies (ComTech-2017), 19-21 April 2017.
      10. PI delivered a half day seminar at International Conference on Communications Technologies (ComTech-2017), 19-21 April 2017


    • K-12 Outreach Program
    • 1. PI is running and to recruit new K-12 students to work on the REU component of this project, in Feb 2018, PI gave a presentation at Booker T Washington High School, Tulsa.

    • International Collaboration Opportunities
    • 1. To give project GRAs an experience of international collaboration, they were introduced with PI’s collaborators at 5GIC, Surrey; the University of Glasgow and the University of Leads UK, via emails and video conference sessions.

    • Conference Attendance
    • 1. One GRA involved in the project was sent to present her paper and attend IEEE PIMRC held in Bologna, Italy, Sep 2018. 2. Two of the GRAs were sent to present their paper and attend IEEE International Conference on Computing, Networking and Communications (ICNC), Silicon Valley, CA, Jan 2017.

    • Internship
    • 1. Internships were arranged for two GRAs involved in the project with Bell Labs, NJ in Summer 2018.
      2. An internship was arranged for one of the GRAs with AT&T Big Data foundry, Dallas in Spring 2017.

    • Mentoring
    • 1. Project students were provided one to one mentoring by PIs through twice a week meetings for close guidance to conduct the proposed research.
      2. One of the project participant students who was eligible for participation in PI’s ongoing IRES project was sent to 5GIC, Surrey, the UK for summer 2018.

Thank you for visiting AI4Networks.com

4502 East 41st Street Building 4, Tulsa, OK 74135

+1-(918)-660-3348

info@bsonlab.com

Search