Major Goals of this Project
-
The significant advancement in wireless networking technologies as well as the proliferation of
mobile devices have enabled information-centric Internet of Things (IoT) systems, where timely information
updating is normally required or preferred by users. Age of Information (AoI) has recently been introduced to
quantify the freshness of the knowledge the controller has about the remote information sources.
While AoI is information-centric, in the present state of the art, the evaluation of information freshness has
been conducted mainly in a transmission-centric manner. This project aims to bridge the gap between transmission-level
information freshness and application-level decision making by revealing involved fundamental research issues,
developing generic modeling tools, and generating effective network protocols in some important practical scenarios.
Intellectual Merit
-
The proposed research is expected to contribute innovative information freshness optimization
techniques, with fundamental intellectual merits in four aspects.1) Our studies are to show that optimizing
information freshness from an application-oriented angle (that is, incorporating user query pattern,
correlations across different information sources, and cooperative decision among distributed agents)
requires new modeling/analysis techniques and will reveal new insights on how application-level factors
impact lower-level transmission scheduling. 2) As application-level decision making may not be
synchronized with lower-level information update, application-oriented freshness optimization does not
necessarily require minimizing AoI all the time. We will study effective decomposition techniques and
low-complexity algorithms for information freshness optimization, through predicting user query patterns
and incorporating delay-tolerant responses at the application level. 3) A fundamental obstacle to
information freshness optimization is the lack of a generic application-level abstraction, which is to be
studied in the presence of multiple correlated information sources. This study will demonstrate how the
intricacies raised by the correlations can be addressed efficiently through integrating domain knowledge
and deep reinforcement learning techniques. 4) Noticing the inadequacy of CSMA-based scheduling in
terms of freshness optimization reported by recent studies, we will design a decentralized TDMA-based
medium access control (MAC) protocol that facilitates timely information exchange in the scenario of
multi-agent decision making; the outcome of this study will contributed new insights on information
freshness oriented network protocol design.
Research Achievements
-
1) Research Activities in Year 1
Optimal sleeping strategy for both information freshness and energy efficiency --- This study is motivated by the balance between maintaining information freshness and mitigating energy consumption in Internet-of-Things (IoT) neworks. The freshness of data can be characterized by age of information (AoI). Intuitively, we may minimize AoI by scheduling a sensor to transmit every slot, but it may incur high or sometimes unnecessary cost of energy. We thus introduce sleeping mode and active mode for sensors. In the sleeping mode, the sensors will not send data to save energy. We theoretically analyzed how such an energy-efficient design impact AoI performance and derived the optimal sleeping strategy for the best AoI performance under a practical energy constraint.
Sleep-wake sensor scheduling for minimizing AoI penalty --- We consider a typical IoT application scenario where multiple sensors monitor some time varying physical processes and report measurements to a central base station through an unreliable wireless channel. Each sensor operates according to the sleep-wake pattern; the BS schedules the sensors for transmission according to a time-divided-multiple-access (TDMA) manner. The channel is unreliable where a scheduled transmission is successful with a probability. Our objective is to obtain an energy-efficient scheduling policy to minimize the system AoI averaged over all the sensors. A particular challenge is that the scheduling need to prioritize the handling of AoI increase due to unsuccessful transmissions for active sensors and that due to idling for sleeping sensors. From the perspective of energy saving, active sensors that have experienced more unsuccessful transmissions need to be scheduled with higher priority. The prioritization should also take the eagerness of each sensor for transmission (reflected by the duration of sleeping period). We successfully tacked this challenging scheduling problem by designing an innovative metric of AoI penalty.
2) Research Activities in Year 2
Adaptive messaging based on the age of information in vehicular ad hoc networks (VANETs) --- We propose an AoI based adaptive CAM messaging control algorithm. Our objective is to minimize an age-penalty function in a trajectory prediction application. In our design, each vehicle will compute a penalty score that integrates local penalty indicating whether the CAM messaging frequency is appropriate for its mobility status; and a neighboring penalty indicating the impact of network congestion on the trajectory prediction quality. The penalty score is used to adaptively control the CAM sending frequency. Practical simulations demonstrate that our adaptive messaging method can indeed control the CAM sending frequency at an appropriate level while balancing the network congestion level and driving safety requirements.
Minimizing the age of information for monitoring over a WiFi Network --- We model the tagged node as a first-come-first-serve (FCFS) queue over the MAC layer. The AoI queuing analysis tells that the AoI can be minimized by adjusting the traffic arrival rate (i.e., the sampling rate of the monitoring node) to meet an optimal utilization factor. However, the fundamental challenge here is that the service time over the MAC is correlated with the traffic arrival rate due to channel contentions. We thus develop an iterative algorithm that interweaves sampling control and MAC analysis for AoI minimization. In addition, we develop new techniques so that the MAC performance analysis can handle heterogeneous traffic models associated with the contending users. Simulation results demonstrate that our algorithm is very accurate and robust crossing a variety of networking scenarios.
A deep learning assisted approach for minimizing the age of information in a WiFi network --- In the iterative computation framework for AoI optimization, a critical component is to determine the accurate average packet service time of the tagged node given its traffic arrival rate and the background traffic. We develop a deep learning (DL) based approach to predict two critical values, the conditional collision probability of the tagged node and the successful transmission probabilities from other nodes, which are required for accurate MAC service time analysis. A particular original design is that our DL module only needs to use the data that is readily available at the AP and the tagged node, incurring no communication overhead with other nodes in the network. The DL module enables the tagged node to obtain the optimal sampling rate for minimized AoI in a full analytical manner.
3) Research Activities in Year 3
An analytical approach for minimizing the age of information in a practical CSMA network --- We propose a novel and general approach utilizing stochastic hybrid systems (SHS) for AoI analysis and minimization in CSMA based wireless networks. Specifically, we consider a practical yet general networking scenario where multiple nodes contend for transmission through a standard CSMA-based MAC protocol, and the tagged node under consideration uses a small transmission buffer for small AoI. We for the first time develop an SHS based analytical model for this finite-buffer transmission system over the CSMA MAC. Moreover, we develop a creative method to incorporate the collision probability into the SHS model, with background nodes having heterogeneous traffic arrival rates. Our model enables us to analytically find the optimal sampling rate to minimize the AoI of the tagged node in a wide range of practical networking scenarios. Our analysis reveals insights into buffer size impacts when jointly optimizing throughput and AoI. The SHS model is cast over an 802.11-based MAC to examine the performance, with comparison to ns-based simulation results.
AoI analysis and minimization in cell-free Industry IoT networks --- Cell-free massive multiple-input multiple-output (m-MIMO) architecture is a promising solution for Industrial Internet of Things (IIoT) because it not only provides massive connectivity but also eliminates the traditional cell edges. The existing studies on cell-free mMIMO IIoT networks mainly focus on capacity enhancement. It is inadequate in terms of the stringent requirements for real-time information in IIoT networks. In this study, we propose a resource allocation policy to meet the stringent AoI requirements of different applications in a cell-free IIoT network. In particular, we design a priority-aware frame structure to enable differentiated AoI services. By adjusting the number of devices transmitted at the same time, the proposed scheme supports devices of each priority to adjust their own transmission period without affecting other devices. The scheme is flexible and can be well adapted to dynamically varying network environments. In addition, our AoI analysis also takes into account the correlation between device location and performance.
Energy-efficient AoI optimization in industrial Internet of Things --- We consider that a wireless sensor monitors a dynamical system and reports real-time measurements to a processing center through an unreliable wireless channel. We study the problem of optimizing the sensor data freshness in terms of AoI while saving energy by scheduling the sensor to sleep when needed. The problem is formulated as a Markov decision process that takes both AoI and energy consumption into account, to which we theoretically prove that the optimal scheduling policy forms a cyclic sleep-wake pattern. We we theoretically derive the optimal sleep period when the AoI function is linear, and further propose an algorithm to find the optimal sleeping period for the generalized AoI function. Simulation results demonstrate that the introduction of sleep mode and the optimal sleep scheduling policy saves a lot of energy, which is a good way to balance the freshness of data and energy consumption.
Publications
X. Cao, J. Wang, Y. Cheng, and J. Jin, "Optimal sleep scheduling for energy-efficient AoI optimization in industrial Internet of Things," IEEE Internet of Things, vol. 10, no. 11, June 2023, pp. 9662-9674
O.T. Ajayi, X. Cao, H. Shan, and Y. Cheng, "Self-renewal machine learning approach for fast wireless network optimization," in Proc. IEEE MASS 2023,invited paper, September 25-27, 2023, Toronto, Canada.
T. Chang, X. Cao, and Y. Cheng, "Age of local information for fusion freshness in Internet of Things," Proc. 2023 IEEE/CIC International Conference on Communications in China (ICCC), Dalian, China, August 10-12, 2023
S. Wang and Y. Cheng, "A deep learning assisted appraoch for minimizing the age of information in a WiFi network," in Proc. IEEE MASS 2022,invited paper, October 20 - 22, 2022, Denver, Colorado
J.M. Parella, O.T. Ajayi, and Y. Cheng, "Adaptive messaging based on the age of information in VANETs," accepted to IEEE GLOBECOM 2022, 4-8 December 2022, Rio de Janeiro, Brazil (Hybrid)
S. Wang, Y. Cheng, L.X. Cai, and X. Cao, "Minimizing the age of information for monitoring over a WiFi network," accepted to IEEE GLOBECOM 2022, 4-8 December 2022, Rio de Janeiro, Brazil (Hybrid)
S. Zhang, B. Yin, W. Zhang, and Y. Cheng, "Topology aware deep learning for wireless network optimization," IEEE Transactions on Wireless Communications, vol. 21, no. 11, pp. 9791-9805, November 2022.
J. Wang, X. Cao, B. Yin, and Y. Cheng, "Sleep-Wake Sensor Scheduling for Minimizing AoI-Penalty in Industrial Internet of Things," IEEE Internet of Things, vol. 9, no. 9, pp. 6404-6417, May 2022.
S. Zhang, O. T. Ajayi, and Y. Cheng, "A Self-supervised learning approach for accelerating wireless network optimization," IEEE Transactions on Vehicular Technology, vol. 72, no. 6, June 2023, pp. 8074-8087.
Y. Cheng, B. Yin, and S. Zhang, "Deep learning for wireless networking: The next frontier," IEEE Wireless Communications, vol. 28, no. 6, pp. 176-183, December 2021.
S. Zhang, B. Yin, and Y. Cheng, "Deep reinforcement learning for scheduling in multi-hop wireless networks," invited paper, in Proc. The 18th IEEE International Conference on Mobile Ad-Hoc and Smart Systems (MASS 2021), Virtual Conference, October 4-7, 2021.
Broader Impact Activities
Energy efficiency is of critical importance in modern IoT applications. Our contributions of the optimal sleeping strategy for minimizing the age of information are expected to benefit many freshness-sensitive IoT applications incorporating the sleep-wake sensors.
The novel metric of age of information with penalty (or AoI-penalty) is an important contribution to the research community on age of information. The AoI-penalty can not only capture the heterogeneous eagerness of the active sensors to provide fresh information (as the contents may be of different levels of importance) but also incorporate energy-efficiency based prioritization. The AoI-penalty based optimization offers a general framework to study the scheduling problem over multiple sleep-wake sensors. The age-penalty function for a trajectory prediction application is directly linked with driving safety. It is expected that this penalty function can find other applications in VANETs beyond our method reported in this project.
The machine learning assisted approach for accurate 802.11 MAC performance analysis, given generic heterogeneous traffic models, is of independent importance to the wireless networking research community. This ML assisted approach can be applicable to other random access based MAC protocols, such as, the 802.15.4 Zigbee protocol.
AoI analysis over CSMA is still an understudied area. we for the first time demonstrate how to construct a SHS model that incorporate both finite buffer queueing and practical CSMA MAC operations for AoI analysis. This modeling method is expected to enhance the theoretical toolset for the AoI research community.
The service differentiation mechanism and associated spatio-temporal analytical framework, developed for the cell-free IIoT networks, can find applications beyond the AoI analysis. The framework captures multiple main features under wireless IIoT networks, including cell-free mMIMO architecture, frame structure, finite-sized geographic areas, densely deployed devices, device priority, retransmission, and interaction among different transmission links.
The PI gave an invited talk titled as "Application-Oriented Scheduling for Optimizing Information Freshness in Wireless Networks," at the International Conference on Computing, Networking and Communications (ICNC) 2023.
The PI gave a keynote talk titled as "To Find the Supremacy of Machine Learning in Wireless Network Optimization" at the Workshop of Edge Intelligence for 6G Networks, co-located with the IEEE/CIC International Conference on Communications in China (ICCC), Dalian, China, August 10-12, 2023. The talk contains some outcomes from this project. The PI gave the same talk as an invited seminar speaker at the Department of Computer Science and Information Engineering, National Taiwan University.
Multiple invited paper were published in the IEEE International Conference on Mobile Ad-Hoc and Smart Systems conferences (MASS 2021, 2022, and 2023), respectively.
The PI has attended the IEEE Communications Society (ComSoc) Excellence Camp co-located with IEEE International Symposium on Personal, Indoor and Mobile Radio Communications (PIMRC), 5 September 2023, Toronto, Canada. The IEEE ComSoc Excellence Camp program (https://www.comsoc.org/education-training/ieee-comsoc-excellence-camps) was recently created to provide students, postgraduate students and young professional members with opportunities to stay up to date on the emerging technologies. The PI's student Suyang Wang's project "A Deep Learning Assisted Approach for Minimizing the Age of Information in a WiFi Network" was presented in this ComSoc educational camp and won the Bronze Medal Award.
The PI leveraged the studies in his NSF projects to enhance the research components of his courses, "ECE541: Computer Network Performance Analysis" and "ECE545: Advanced Computer Networks." In Fall 2020, two week of lectures about applications of probability and queueing theory in AoI analysis, medium access control analysis, and machine Learning were added into the ECE541 course contents. In Spring 2021, ECE545 course contents were enhanced with discussions of AoI related studies in next generation wireless networks. The PI has been developing a new course "Machine learning in Wireless Networking," which is expected to deliver in 2023 for both graduate students and senior undergraduate students.
In Summer 2021, the PI worked with IIT Summer Engineering Undergraduate Research Immersion Program and offered a research course. The course was advertised by IIT and open to both IIT undergraduate students and visiting students from other universities or other countries. The course was successfully delivered. In the course, students used open-source machine learning tools to solve a scheduling problem in a device-to-device wireless networking context. In Fall 2022, he contributed "Optimizing Information Freshness in Internet of Things (IoT) Systems" to the Illinois Tech Armour R&D Research Project Program, which is offered to both undergraduate and graduate students.
In Fall 2021 and Spring 2022, the PI worked with the exchange student program of IIT and supervised one visiting master student (Jordi Marias I Parella) from Polytechnic University of Catalonia, Spain. The student was teamed up with another PhD student (Oluwaseun Ajayi) and fruitfully conducted the research on AoI controlling in VANETs.
The PI conducted effective outreach activities.His PhD student Oluwaseun Ajayi won the Illinois Tech Socially Responsible Modeling, Computation, and Design (SoReMo) Scholarship award.SoReMo provides unique interdisciplinary research and outreach opportunities for students who want to apply computational, modeling, and design skills to solve a social issue. Oluwaseun's SoReMo award is largely credited to his solid background on machine learning built up through his research activities in the NSF projects.
The PI's research has also attracted a grade-11 student, Hannah Ding from Walter Payton College Preparatory High School, to get involved in academic research in his Lab. Hannah is teamed with Oluwaseun to conduct ML related research. The PI also served as an external mentor to another grade-11 student, Matthew Wojcik from Bartlett High School, for his school research project.
In Fall 2022, the PhD students in this project attended the IIT Welcome Week Student Research Poster Event and displayed the research contributions from the PI's NSF projects with research posters. This is a big event with university-wide attendance of undergraduate/graduate students and faculty members, alumni and family members, and invited judges from industry.
Personnel Involved in This Project
-
PI: Yu Cheng
-
RA: Shuai Zhang
-
RA: Oluwaseun Ajayi
-
RA: Suyang Wang
Computation and Simulation Codes
-
Available based on request.
[ Home | Research Projects | Publications | Honors & Awards | Teaching | Students | Professional Activities | Invited Talk ]
|
|
Last Updated: October, 2021 |
