Prof. Navrati Saxena's Professional Research Experience:

 

My prolonged research experience goes hand-in-hand with my overall academic career and spans across different Universities across different countries in Europe, Asia and North America. My research is primarily focused on different facets of mobile wireless networks, including 5G/6G wireless, Device to Device (D2D) communications, Social Networking, Power Savings, Internet of Things (IoT), UAVs (drones), Contact Tracing, and Connected Cars. 
 

A. Assistant Professor - SJSU, USA, Fall 2020 ~

Now and over the next couple of years, my research agenda is made up of four major components, evolving around next generation “Beyond 5th Generation” (B5G) and 6th Generation (6G) wireless networks:

 

• Optimizing Next Generation B5G and 6G Wireless using Machine Learning Methods

The existing 5G wireless architecture lacks sufficient flexibility and intelligence to efficiently handle stringent and diverse demands. As a result, the evolution towards B5G and 6G wireless calls for an architectural transformation required to support service heterogeneity, coordination of heterogenous wireless technologies, and on-demand service deployment.  Intelligence is rapidly becoming a necessity for the deployment, optimization, and operation of wireless networks beyond 5G. This is primarily because of the increasing complexity of 5G wireless networks in response to the need to handle demanding service requirements. Therefore, there is an impending need to leverage emerging learning techniques to embed intelligence in every layer of the RAN architecture. The telecom industry has already demonstrated the use of Machine Learning (ML) in many trial deployments and the defacto global wireless standard bodies (e.g., 3GPP, Open RAN) have also decided to embrace ML in wireless networks. Maintaining a synergy with global wireless revolution, my future research is also focused on using ML to learn and predict the complicated wireless channel characteristics, positions of mobile devices and mobile’s energy consumption patterns. Using efficient learning and prediction techniques, my research will aid the next generation devices with reduced computation complexity and energy (battery) consumption. 

 

• Un-licensed bands and Non-Terrestrial Networks for Ubiquitous Coverage

The increasing demand of high data rate and connectivity has urged the wireless vendors to think of exploring unlicensed (free) bands. The unlicensed band is typically used by many other devices, like Wi-Fi and even microwave-oven, resulting in the reduced probability of the availability of channel for data transmission. The mobile device has to keep its radio circuitry “on” and wait until it gets access to the unlicensed channel again. This process escalates the energy expense of the battery-constrained mobile devices. With my research colleagues, I would like to explore efficient algorithms and statistical analysis to model and analyze the mobile devices’ activity to reduce its energy consumption, thus extending its battery life. 

 

Similarly, wireless connectivity is also slowly embracing the non-Terrestrial (satellite) networks (NTN) to provide universal coverage, even in the extreme geographical areas and in the developing countries, having low population density. Existing wireless infrastructure fails to provide such connectivity as the high installation costs of new wireless infrastructure cannot match the revenue generated from areas with low population density, especially in developing and under-developed countries. Hence, such a new wireless network coverage, which is inclusive even to the so-called “un-reachable”, needs major research efforts. In this aspect I would like to develop new algorithms and simulation to study the movement patterns and analyze the connectivity for both Low Earth Orbit (LEO), Geosynchronous (GEO) satellites.

 

• Efficient inter-vehicular communications for Safety and Infotainment

The number of automated vehicles is expected to rise steadily over the current and next decade, eventually making the vision of “connected cars” a reality. Transmission of emergency safety messages, while providing the infotainment solutions is a key challenge in inter-vehicular communications. Automated vehicles will generate decentralized environmental notification messages in response to emergency events, like accident or hazards on the road. These notification messages need to be immediately and reliably transmitted to the neighboring vehicles. In my work, I plan to consider the co-existence of high-priority notification messages with low-priority background and infotainment traffic and explore the associated impact. The notification messages could be broadcast using vehicle-to-vehicle wireless links, while the background traffic is sent using vehicle-to-infrastructure (road-side network towers). I would like to analyze average end-to-end delay, average packet reception ratio for notification messages, as well as capturing average background traffic user throughput using detailed queuing analysis and realistic vehicular simulation.

 

• Contact Tracing using Internet of Things (IoT) for fighting against Pandemic

The outbreak of this pandemic has disrupted our lives in numerous ways, crippling the healthcare systems of even the most advanced countries. One reason behind this was the large fraction of asymptomatic infection, creating a large number of reasonably healthy disease carriers, making it easier for the virus to reach the more vulnerable population. This makes the contact tracing and infection tracking indispensable for containing the spread of this pandemic. The emergence of Internet of Things (IoT) is gradually getting popularity in healthcare applications. This IoT devices has the potential to become an ideal contact/infection tracing because of its ubiquity. Inspired by this, we have recently introduced a new IoT-based framework for contact and infection tracing, which specifically incorporates symptom-based detection that has been ignored on existing contact tracing models. The ability of this framework to meaningfully merge real-time symptom information from the IoT devices provides a fast and efficient way of tracking the pandemic spread. As the next step of my research, I would like to embed intelligence inside our IoT-based contact tracing to make it more efficient and generic enough to fight against any future variants or other pandemic.

 

B.     Associate / Assistant / Research Professor – SKKU, Korea, Dec 2006 – Feb. 2019

 

• Massive Internet of Things (IoT) using 3GPP 5G NR Systems: This project is focused on automated communications between IoT devices and Smart Grids, using 3GPP 5G wireless technology. We developed 5G-enabled gateways, which gather user data from thousands of IoT devices and use data compression over 5G NR links to efficiently deliver the data in the uplink. The project is verified with experimental prototypes/ simulation and published in IEEE Communications Magazine.
• Optimal Demand Response in Smart Grids using 5G NR MBMS:  Efficient 5G small-cell planning and optimal multicast communications can significantly improve the Demand Response programs in Smart Grids (SG). Actual SG data is collected across Korea’s Electric Power Corporation (KEPCo) and the algorithms are verified using simulation. The work is published in IEEE Transactions on Industrial Informatics.
• Energy Efficiency for Green 3GPP 5G NR Systems: In this research project, we have looked into the design and analysis of new, cloud-based virtualized 5G wireless networks for reducing the energy dissipation. We have also proposed new battery-aware scheduling to prolong the residual energy in smart phones. The works are published in famous journals, like IEEE Journal on Selected Areas in Communications, popular IEEE Transactions on Mobile Computing, Springer Multimedia Tools & Applications and Wiley Transactions on Emerging Technologies.
• Directional Discontinuous Reception in 3GPP 5G NR Systems: In this work we introduce new Directional-Discontinuous Reception (DDRX) for directional air interface in mmWave enabled 5G communications. It emphasizes the importance of beam searching for alignment of directional beams between UE and 5G gNB, after every sleep cycle. Beam searching, reduces the effective sleep time but it is inevitable. We propose three new DDRX mechanisms: Integrated DDRX, Standalone DDRX and Cooperative DDRX (C-DDRX) to limit the impediments on power saving. The work is published in famous IEEE Trans. on Mobile Computing. 
• Video over Next Generation Wireless Networks: With the popularity of social networking, like Facebook, Twitter etc., video is in increasingly becoming a dominating application. It is expected that next generation wireless networks need to efficiently deliver video streaming with some precise Quality of Experience (QoE). In this project, we have pointed out that depending on transmitter and receivers’ locations, wireless operators and video service providers could cooperate to optimize the video delivery architecture and performance. The work in published in famous IEEE Communications Magazine – the best magazine in communication society.
• Self-Organization in 5G/4G Networks: Self-Organizing Networks (SON) aim at auto-configuration and management of large-scale 4G/5G wireless networks. In this project, we identified that optimal configuration of networks is computationally hard (NP-complete) and propose near-optimal solutions based on biologically inspired algorithms. The work is implemented in testbed and published in Springer Wireless Networks journal and Hindawi Wireless Communications and Mobile Computing journal.
• QoS-provisioning in Multimedia Sensor Networks: Emerging multimedia sensor networks are going to be increasingly deployed for applications like video surveillance and tele-medicine. We have pointed out that QoS routing with multiple QoS constraints (like delay and bandwidth) is an NP-hard problem and proposed new QoS-routing protocol for these multimedia sensor networks. The work is published in Elsevier Computer Networks, and Hindawi Wireless Communications and Mobile Computing. 
   

C.     Research Experience at Amity University, India.

 

• Research Coordinator and lead for joint-research on Next Generation Networks (NGN) / Multimedia Services with Siemens Public Communication Networks Limited (SPNCL).
• Coordinator – Research & Technical Programs in ASCS
  • Wireless multicasting, mobility tracking and wireless sensor networks (with high-quality International conference and Journal papers)
  • Guiding Research Projects jointly with IIT-Delhi: Actively involved with IIT-Delhi’s wireless research group for joint research project in MAC-Physical Layer interaction/modeling.
     

D.     Visiting Research Associate, University of Texas Arlington, USA July 2003 – March 2005

 

• Dynamic hybrid scheduling for heterogeneous, asymmetric, wireless environment.
• Probabilistic combinations of consecutive push and pull algorithms for adaptive hybrid scheduling.
• Differentiated QoS based on service classification of clients to reduce the churn rate and increase the efficiency of practical hybrid scheduling in wireless data networks.
• The role of subjective parameters like clients' impatience and its effects (anomalous behavior) in heterogeneous hybrid scheduling.
• Optimal, online hybrid scheduling over multiple channels to minimize overall waiting time.

 

E.     Ph. D. Researcher, University of Trento, Italy: Nov 2001 – Mar 2005

 

Summary of PhD Dissertation, entitled: “A Framework for Dynamic Hybrid Scheduling Strategies in Heterogeneous Asymmetric Environments,"

A new framework for hybrid scheduling strategy in heterogeneous, asymmetric, wireless environments. The framework effectively combines push and pull scheduling to minimize the overall waiting time perceived by the clients. The entire set of clients is divided into different priority-classes and the scheduling scheme attempts to optimize the waiting time and blocking of highest priority clients, thus reducing the churning rate and increasing the profit of wireless service providers. The dissertation also considers practical, yet subjective measures, like client's impatience and removes the possible anomalies arising in the system. The work is also extended to incorporate the hybrid scheduling over multiple channels.