Title: Collaborative Research: EARS: Large-Scale Statistical Learning based Spectrum Sensing and Cognitive Networking, funded by NSF under grant CNS-1659962, 10/01/2016 – 12/31/2018.

Project Summary:

As cognitive radio (CR) research advances to multihop and complex systems over large geographic regions, the spectrum utilization enhancement should be generalized to fully exploit the spectrum usage diversity in three dimensions (3D): time, frequency, and space, with new emphasis on the under-explored spatial dimension. Accordingly, this project focuses on the following three research objectives. The first one is to utilize the recent advancements in statistical learning over big data to develop efficient 3D spectrum sensing schemes, where a hierarchical approach is taken in developing novel finite-bit and single-bit learning techniques to efficiently explore the correlation structure across the three dimensions, with an advanced distributed approach also developed. The second one is to develop two key building blocks in large-scale CR networking based on the 3D spectrum sensing: 1) a novel multi-scale routing scheme to enhance the overall spectrum utilization, with a focus on exploiting the layered spectrum usage correlation structure in the spatial dimension; and 2) a reliable hierarchical common control channel identification scheme. The last research objective is to validate some key aspects in the proposed sensing and networking schemes via both intensive simulations and a concept-proving testbed. Throughout the project, an interdisciplinary approach is taken to combine the methods of statistical learning, signal processing, and wireless networking, with the core built upon the hierarchical treatment of both spectrum usage statistics and CR networking methodologies. The project provides both theories and algorithms for large-scale spectrum sensing and cognitive networking. Through a coherent education plan, the research findings will be incorporated into courses, and disseminated to the community via journal papers and conference presentations.

Figure 1. Motivating example: 3D spectrum sensing and routing.

Project Goals:

The goals break into the following three thrusts:

· Large-scale cooperative statistical-learning based spectrum sensing model and algorithm development. Goals in this thrust include:

Ø Observation model specification for 3D (spectrum, space, and time) statistical spectrum sensing

Ø Latent model specification for 3D statistical spectrum sensing

Ø Prior model specification for 3D statistical spectrum sensing

Ø Efficient computational methods for big data

Ø Distributed learning algorithms for large-scale spectrum sensing

· Applying the 3D spectrum sensing outcomes obtained in the previous thrust to address key networking issues in large-scale geographic CR networks, including spectrum utilization oriented multi-scale CR routing and hierarchical common control channel (CCC) identification. Particular goals include the following:

Ø 2.1 Consistent network-state aggregation over Markov Random Field (MRF)

Ø 2.2 Finding multi-constrained path (MCP) in random field

Ø 2.3 Exploiting time-domain correlation for NAI maintenance and re-routing

Ø 2.4 Centralized optimization for the CCC hierarchy

Ø 2.5 Distributed optimization, protocol development, and evaluation for the CCC hierarchy

· Performance evaluation through simulation and testbed development

Project Personnel

PIs: Shuguang Cui (TAMU Site), Huiyan Sang (Co-PI, TAMU Site), Liuqing Yang (Colorado State Site), and Tao Shu (Auburn Site)

Graduate Students (Auburn Site)

· Li Sun

· Jing Hou

· Tian Liu

· Jian Chen

· Xueyang Hu

· Rui Zhu (Ph.D. at Oakland University, co-advised with Prof. Huirong Fu)

Past Graduate Students

· Dong Han (Oakland University, graduated in Dec. 2016)

· Ye Yan (Oakland University, graduated in Dec. 2016)

Major Research Activities

The research activities are organized in four important areas: (1) spectrum sensing and analytics, (2) large-scale context-aware networking, (3) security and privacy issues for context-aware networking, and (4) development of spectrum sensing testbeds. Our activities in each area is elaborated as follows.

1. Spectrum Sensing and Analytics

Spectrum sensing and analytics is essential for a cognitive radio network to achieve spectrum awareness and adaptability. Our work in this area is mainly focused on analyzing the spatial and temporal correlation structures of the spectrum and utilizing these correlation to develop efficient and robust spectrum sensing mechanisms. In particular, the following achievements have been accomplished:

(a) 3D spectrum measurements: Two testbeds were developed to measure the radio spectrum condition in the space, time, and frequency domains (3D). The first testbed is based on two sets of USRP (Universal Software Radio Peripheral) systems and the GNU Radio. The measurements on this testbed were mainly focused on the VHF bands, with a particular emphasis on the FM radio broadcast band (75-105 MHz). We placed this testbed at 13 different locations on the first floor of DHE and at each location the two USRPs were separated by several preset distances. At each location, each USRP uses GNU Radio to plot a real-time waterfall graph that represent the strength of received signals in a two-dimensional time-frequency window. Pooling the two USRPs’ waterfall graphs together will allow us to evaluate the correlation of the radio signal in the 3D (space, time, and frequency) domains. Our second testbed focuses on the 2.4 GHz ISM band and consists of a modified TP-Link 802.11 WiFi router and an Android App developed by our team. In particular, the TP-Link WiFi router was re-programmed and are able to perform automatic frequency sweeping among the 11 802.11g WiFi channels according to a prescribed pattern (denoting the sequence of channels and the sojourn time at each channel). An Android App was developed to measure the received signal strength (RSS) of the beacon broadcast by the WiFi router. The App is able to synchronize with the frequency sweeping of the WiFi router, and is able to record the received RSS and calculate their statistics such as max, min, mean, etc. The App also logs all measured data for future analysis (such as calculating the correlation). In the measurement, we fixed the location of the WiFi router and placed a Google Nexus Android Pad where our APP is installed at 27 locations on the first and second floors of DHE, including both indoor and outdoor. A large amount of spectrum data in the frequency, time, and space domains has been collected during the measurements.

(b) Spectrum correlation analysis from empirical measurement data: Extensive analysis on the spatial and temporal correlation structure of the spectrum has been performed based on the CRAWDAD dataset provided by the University of Utah and our own measurement data. The CRAWDAD database records the empirically measured link signature (channel response) between any pair of sender and receiver in a 44-node (USRP devices) wireless network in a typical indoor office-type environment. The total number of links (transmitter-receiver pair) is 44*43= 1892. For each link, 5 distinct measurements were made at different times. As a result, the dataset has recorded over 9300 traces for the link signature.

Our goal is to decide how well a link signature can be estimated by the signature of nearby links (i.e., the spatial correlation). We are also interested in deciding how well a link signature measured at one time can be estimated by the signatures of the same link measured at other times (the temporal correlation). Different from existing works that pre-assume certain analytical models for the spectrum correlation, our study attempts to construct a spectrum correlation model out of the empirical dataset. To this end, we have utilized a comprehensive set of representative machine learning algorithms, including various kernel methods for support vector machine (SVM), decision tree, ensemble method, multivariate linear regression, and neural network. Our results show that the spectrum data presents strong spatial and temporal correlation. Utilizing these correlation allows one to utilize the link signatures measured at different locations and times to significantly approach the signature of a target link. Our findings have extensive implication to a number of applications. For example, our work provides a concise and empirical-data-based spectrum representation for large-scale cognitive radio networks, as desired in our proposal. Our finding may be also useful for wireless security, such as attacking PHY-layer secret key extraction.

(c) QoS-compliant sequential channel sensing: We studied QoS-compliant sequential channel sensing techniques for cognitive radios when spectrum availability and quality is not known a priori. A resource-constrained CR relies on sequential channel sensing and probing to resolve spectrum uncertainty and search for good transmission opportunities in real time. We are interested in maximizing the effective throughput the CR can achieve with a desired confidence (success probability) under a spectrum access delay constraint. The optimization is formulated as a finite-horizon optimal stopping problem under the objective of maximizing a given percentile of the rate of return at the stopping time. This formulation cannot be directly solved by classical optimal stopping theory, because the latter only supports a mean-reward objective function. A novel transformation is developed to convert the problem into solving a series of sub problems, each of which optimizes a transformed mean-reward of the original problem and therefore can be solved using classical optimal stopping method. We prove the monotonicity of the sub problems, based on which we develop a fast algorithm to efficiently find the unique solution to the original problem. To account for different MAC mechanisms used in practice, our analysis considers both non-recall and recall channel decision strategies. Extensive simulations are performed to verify the effectiveness and significance of the optimization. We show that significant gains (e.g., over 30%) on the QoS-compliant capacity can be achieved by the proposed algorithm when compared with the counterparts.

(d) Spectrum sensing and modeling in mmWave bands and its application for LOS blockage prediction: Rather than targeting the conventional radio spectrum under 6 GHz, this research focuses on a much higher frequency – the mmWave bands (30 to 300 GHz). The sensing and modeling part of this research fulfills Goal 1 of the project, while its application to LOS blockage prediction is to fulfill Goal 2. In particular, while a cellular mmWave channel can offer extremely high (Gbps-level) capacity by taking advantage of its wide bandwidth, it suffers unique challenges. For example, a mmWave channel is vulnerable to the line-of-sight (LOS) blockage. Due to the short wavelength of mmWave, the signal cannot circumvent or penetrate through an obstacle once it blocks the signal propagation LOS path. An LOS blockage could easily translate into a 20 to 30 dB loss on the signal strength in an outdoor communication environment, leading to a sudden outage of the physical channel. LOS blockage happens frequently in a mobile scenario, whereby a user’s movement could easily encounter obstacles such as treetops, pedestrians, and buildings. The resultant intermittent connection at the physical layer will generate a substantial degradation on the performance of higher layers. In this research, we propose an LOS-blockage prediction mechanism that can accurately predict when the LOS component in an mmWave channel will be blocked, as well as how long the blockage will last. These predictions enable forehand handshaking with the handover destination base station, and thus can significantly reduce the handover delay. The proposed prediction is based on the realtime sensing of the geometry of the mmWave multipaths channel, which has a sparse structure and consists of one dominant LOS component and a handful of relatively weak NLOS components. The prediction is made by sensing and detecting the blockage of peripheral NLOS components. These NLOS components may not be strong enough to be used for high-speed communication, but they are typically strong enough to be detectable. By collaborating with Prof. Husheng Li at the University of Tennessee at Knoxville (UTK), the feasibility and effectiveness of the proposed mechanisms are validated based on experiments over a real mmWave communication testbed located in Prof. Li’s lab at UTK.

2. Large-Scale Context-Aware Networking

In this research, we have investigated how communication context, including spectrum, location, user behavior, and transmission can be exploited to improve the efficiency and security of wireless networking. In particular, the following accomplishments have been obtained.

(a) Malicious-packet-dropping detection in multi-hop ad hoc wireless networking: We studied how the time-domain correlation in packet transmsision in multi-hop wireless ad hoc routing can be utilized to accurately detect packet dropping attacks launched by insider attackers. Specifically, link error and malicious packet dropping are two sources for packet losses in multi-hop wireless ad hoc network. While observing a sequence of packet losses in the network, we are interested in determining whether the losses are caused by link errors only, or by the combined effect of link errors and malicious drop. We are especially interested in the insider-attack case, whereby malicious nodes that are part of the route exploit their knowledge of the communication context to selectively drop a small amount of packets critical to the network performance. Because the packet dropping rate in this case is comparable to the channel error rate, conventional algorithms that are based on detecting the packet loss rate cannot achieve satisfactory detection accuracy. To improve the detection accuracy, we propose to exploit the correlations between lost packets, as calculated from the auto-correlation function (ACF) of the packet-loss bitmap–a bitmap describing the lost/received status of each packet in a sequence of consecutive packet transmissions. The basic idea behind this method is that even though malicious dropping may result in a packet loss rate that is comparable to normal channel losses, the stochastic processes that characterize the two phenomena exhibit different correlation structures (equivalently, different patterns of packet losses). Therefore, by detecting the correlations between lost packets, one can decide whether the packet loss is purely due to regular link errors, or is a combined effect of link error and malicious drop. Furthermore, to ensure truthful calculation of these correlations, we develop a homomorphic linear authenticator (HLA) based public auditing architecture that allows the detector to verify the truthfulness of the packet loss information reported by nodes. This construction is privacy preserving, collusion proof, and incurs low communication and storage overheads. To reduce the computation overhead of the baseline scheme, a packet-blockbased mechanism is also proposed, which allows one to trade detection accuracy for lower computation complexity. Through extensive simulations, we verify that the proposed mechanisms achieve significantly better detection accuracy than conventional methods such as a maximum-likelihood based detection.

(b) Cognitive Context-aware optimization for distributed storage in mobile cloud computing: This research activity fulfills Goal 2 of the project. In addition to spectrum sensing outcomes, this research also attempts to incorporate additional networking context information, including users’ mobility and file usage pattern, into the optimization of distributed data storage and retrieval in a large-scale mobile cloud computing environment. In particular, mobile cloud storage (MCS) is being extensively used nowadays to provide data access services to various mobile platforms such as smart phones and tablets. For cross-platform mobile apps, MCS is a foundation for sharing and accessing user data as well as supporting seamless user experience in a mobile cloud computing environment. However, the mobile usage of smart phones or tablets is quite different from legacy desktop computers, in the sense that each user has his/her own mobile usage pattern. Therefore, it is challenging to design an efficient MCS that is optimized for individual users. In this research, we investigate a distributed MCS system whose performance is optimized by exploiting the fine-grained context information of every mobile user. In this distributed system, lightweight storage servers are deployed pervasively, such that data can be stored closer to its user. We systematically optimize the data access efficiency of such a distributed MCS by exploiting three types of user context information: mobility pattern, network condition (spectrum sensing outcome), and data access pattern. We propose two optimization formulations: a centralized one based on mixed-integer linear programming (MILP), and a distributed one based on stable matching. We then develop solutions to both formulations. We showed that the exact solution to the MILP problem is NP-hard, thus we develop a polynomial-time approximate solution. This centralized optimization has the advantage of being able to find a high-performance, near-optimal storage configuration, but has the disadvantage of requiring the global information of the network and context conditions. We then show that the optimization can also be formulated as a distributed stable matching problem (SMP), where a user tries to decide the best storage nodes to use based on its local knowledge/observation of the network and context conditions. Compared with the centralized MILP solution, the proposed stable matching method is distributed, only requires local information, and is more scalable. Comprehensive simulations are performed to evaluate the effectiveness of the proposed solutions by comparing them against their counterparts under various spectrum, network, and mobility conditions. 

3. Security and Privacy Issues in Context-Aware Networking

(a) Privacy-preserving localization: Location is a commonly used context information in networking. Existing location-privacy studies are mainly focused on preventing the leakage of user’s location in utilizing the location information (e.g., in location-based services), however, the possible privacy leakage in the calculation of the user’s location, i.e., the localization, has been largely ignored. Such a privacy leakage stems from the fact that a localization algorithm typically takes the location of anchors (reference points for localization) as input, and generates the target’s location as output. As such, the location of anchors and target could be leaked to others. An adversary could further utilize the leakage of anchor’s locations to attack the localization infrastructure and undermine the accurate estimation of the target’s location. To address this issue, in this research, we study the multi-lateral privacy preserving localization problem, whereby the location of a target is calculated without the need of revealing anchors’ location, and the knowledge of the localization outcome, i.e., the target’s location, is strictly limited to the target itself. To fully protect user’s privacy, our study protects not only the user’s exact location information (the geo-coordinates), but also any side information that may lead to a coarse estimate of the location. We developed three privacy-preserving solutions by leveraging combinations of information hiding and homomorphic encryption. These solutions provide different levels of protection for location side information and resilience to node collusion, and have the advantage of being able to trade user’s privacy requirements for better computation and communication efficiency.

(b) Renovating location-based routing (LBR) to achieve integrated efficiency and privacy: We developed HC-LBR routing, the first location-based routing (LBR) mechanism that simultaneously retains high communication efficiency and privacy for Internet of Things (IoT). Existing LBR schemes were originally designed for wireless sensor networks (WSNs). Although they offer attractive efficiency and scalability, their privacy performance presents a vulnerability when being used in IoT. This is because, unlike a conventional WSN where all nodes are owned by the same user, an IoT network has an open architecture in which nodes owned by different users are mixed and work together. Therefore, when LBR is directly used in an IoT network, the sharing of location information among alien nodes allows one user to peek into the communication privacy of other users. HC-LBR overcomes this privacy weakness by computing efficient geographic routes directly based on Hilbert-Curve-encrypted location information. HC-LBR consists of two components: A Kademlia-tree-based routing algorithm that supports efficient geographic routing in HC encrypted space, and Rand-Mix, a lightweight traffic mixer that uses multiple traffic flows to enhance the privacy of HC routing. Our initial evaluation verifies the high efficiency and privacy of the proposed methods.

(c) An in-network AES equivalent (IAE) encryption mechanism: Advances in wearable devices and pervasive computing provide unprecedented opportunity for ubiquitous realtime e-Healthcare and patient monitoring by placing wirelessly connected sensors in, on, and around the body of patients. Due to the privacy-sensitive and mission-critical nature of these wireless body area sensor networks (WBASNs), as well as the desire to use them for long-time uninterrupted monitoring of patients’ vital physiological signals, the privacy/security and energy efficiency of WBASNs are of primary concerns. In this work, we proposed a novel In-network AES Equivalent (IAE) mechanism to protect the security/privacy and maintain good energy efficiency for WBASNs at the same time. IAE achieves this goal by outsourcing part of the energy-consuming cryptographic operation to other deliberately-selected peer sensor nodes so as to balance the energy consumption of the entire network. An analytical model is proposed to characterize the computation and communication energy consumption of IAE, based on which we optimize the outsourcing under given security constraints. Through extensive simulations, we verify the effectiveness and efficiency of the proposed mechanism in prolonging the network lifetime under given security requirements.

4. Development for Spectrum Sensing and Measurement Testbeds

The PI’s team has developed two testbeds for 3D spectrum sensing. The first testbed is based on several USRP (Universal Software Radio Peripheral) devices and GNU Radio, operating on the VHF bands, with a particular emphasis on the FM radio broadcast band (75-105 MHz). The second testbed consists of a modified TP-Link 802.11 WiFi router and several Google Nexus tablets running our self-developed channel sensing and measurement app. This testbed operates over the 2.4 GHz ISM band. The TP-Link router has been re-programmed, so that it can perform automatic frequency sweeping over the 11 802.11g WiFi channels according to a pre-defined pattern. Our android App can synchronize with the frequency sweeping of the WiFi router, measure the received signal strength (RSS) of the beacon broadcast by the WiFi router, and calculate their statistics such as max, min, mean, etc. The App also logs all measured data for future off-line analysis.

Figure 2. Programmable channel-sweeping AP

Figure 3. Android App for automatic channel measurement, recording, and analysis

Figure 4. Field tests

Project Outcomes (from all three collaborating sites, underlined authors are/were students under my advice)

1. Jingchao Bao, Tao Shu, and Husheng Li, “Geometry-based Blockage Prediction in mmWave Communications — A Sensing Approach,” submitted to IEEE ICC 2018, Nov. 2017. 

2. Rui Zhu, Tao Shu, and Huirong Fu, “Empirical statistical inference attack against PHY-layer key extraction in real environments,” in IEEE Milcom 2017, Oct. 2017.

3. Dong Han, Ye Yan, Tao Shu, Liuqing Yang, and Shuguang Cui, “Cognitive context-aware distributed storage optimization in mobile cloud computing: A stable matching based approach,” in IEEE ICDCS 2017, June 2017.

4. Ye Yan, Dong Han, and Tao Shu, “Privacy preserving optimization of participatory sensing in mobile cloud computing,” in IEEE ICDCS 2017, June 2017.

5. Tao Shu and Shuguang Cui, “Renovating location-based routing for integrated communication privacy and efficiency in IoT,” in IEEE ICC 2017, May 2017.

6. Tao Shu, Yingying Chen, and Jie Yang, ``Protecting multi-lateral localization privacy in pervasive environments,’’ accepted by IEEE/ACM Transactions on Networking (ToN), to appear, Aug. 2015.

7. Tao Shu and Marwan Krunz, ``Privacy-preserving and truthful detection of packet dropping attacks in wireless ad hoc networks,’’ IEEE Transactions on Mobile Computing (TMC), vol. 14, no. 4, pp. 813-828, Apr. 2015.

8. Tao Shu and Husheng Li, ``QoS-compliant sequential channel sensing for cognitive radios,’’ IEEE Journal on Selected Areas in Communications (JSAC), vol. 32, no. 11, pp. 2013-2025, Nov. 2014.

9. Dong Han, Ye Yan, and Tao Shu, ``Context-aware distributed storage in mobile cloud computing,’’ Proc. IEEE Globecom 2015 workshop, Dec. 2015.

10. Dong Han, Ye Yan, and Tao Shu, ``Context-aware distributed storage in mobile cloud computing (2-page abstract),’’ Proc. IEEE MASS 2015 poster, Oct. 2015.

11. Dong Han and Tao Shu, ``Thermal-aware energy-efficient task scheduling for DVFS-enabled data centers,’’ Proc. IEEE ICNC 2015, Feb. 2015.

12. Ye Yan and Tao Shu, ``Energy-efficient in-network encryption/decryption for wireless body area sensor networks,’’ Proc. IEEE Globecom 2014, Dec. 2014.

13. Tao Shu and Husheng Li, ``Rate-percentile-optimal sequential channel sensing and probing in cognitive radio networks under spectrum uncertainty,’’ Proc. IEEE SECON 2013, June, 2013.

14. Husheng Li, Tao Shu, Feng He, and Ju Bin Song, ``Futures market for spectrum trade in wireless communications: modeling, pricing and hedging,’’ Proc. IEEE Globecom 2013, Dec. 2013.

15. Bohai Zhang, Huiyan Sang, and Jianhua Huang. "Full-Scale Approximations of Spatio-Temporal Covariance Models for Large Datasets," Statistica Sinica, v.25, 2015, p. 99.

16. Marc Genton, Simone Padoan, and Huiyan Sang. "Multivariate Max-Stable Spatial Processes," Biometrika, v.102, 2015, p. 215.

17. Bohai Zhang, Bledar A. Konomi, Huiyan Sang, Georgios Karagiannis, and Guang Lin. "Full scale multi-output Gaussian process emulator with nonseparable auto-covariance functions," Journal of Computational Physics, v.300, 2015, p. 623.

18. Weijia Han, Chuan Huang, Jiandong Li, Zan Li, and Shuguang Cui. "Correlation based Spectrum Sensing with Over-sampling in Cognitive Radio," IEEE Journal on Selected Areas of Communications, v.33, 2015.

19. Chun-Hung Liu, Beiyu Rong, and Shuguang Cui. "Optimal Discrete Power Control in Poisson-Clustered Ad Hoc Networks," IEEE Transactions on Wireless Communications, v.14, 2015.

20. Armin Banaei, Daren Cline, Costas Georghiades, and Shuguang Cui. "On Asymptotic Statistics for Geometric Routing Schemes in Wireless Ad-Hoc Networks," IEEE/ACM Transactions on Networking, v.23, 2015.

21. Armin Banaei, Daren Cline, Costas Georghiades, and Shuguang Cui. "On Asymptotic Statistics for Geometric Routing Schemes in Wireless Ad-Hoc Networks.," IEEE/ACM Transactions on Networking, v.23, 2015, p. 559.

22. Bohai Zhang, Bledar A. Konomi, Huiyan Sang, Georgios Karagiannis, and Guang Lin. "Full scale multi-output Gaussian process emulator with nonseparable auto-covariance functions," Journal of Computational Physics, v.300, 2015, p. 623.

23. Bohai Zhang, Huiyan Sang, and Jianhua Huang. "Full-Scale Approximations of Spatio-Temporal Covariance Models for Large Datasets," Statistica Sinica, v.25, 2015, p. 99.

24. Chun-Hung Liu, Beiyu Rong, and Shuguang Cui. "Optimal Discrete Power Control in Poisson-Clustered Ad Hoc Networks.," IEEE Transactions on Wireless Communications., v.14, 2015, p. 138.

25. Gustavo Tapia, Huiyan Sang, and Alaa Elwany. "Prediction of porosity in metal-based additive manufacturing using spatial Gaussian process models," Additive Manufacturing, v.12, 2016, p. 282.

26. Hang Li, Chuan Huang, Fuad Alsaadi, and Shuguang Cui. "Performance Analysis for Energy Harvesting Communication Systems: From Throughput to Energy Diversity.," IEEE Globecom. San Diego, 2015.

27. Hang Li, Jie Xu, Rui Zhang, and Shuguang Cui. "A General Utility Optimization Framework for Energy Harvesting Based Wireless Communications," IEEE Communications Magazine, v.53, 2015, p. 79.

28. Hongjian Sun, Arumugam Nallanathan, Shuguang Cui, and Chengxiang Wang. "Cooperative Wideband Spectrum Sensing over Fading Channels," IEEE Transactions on Vehicular Technology, v.65, 2015, p. 1382.

29. Marc Genton, Simone Padoan, and Huiyan Sang. "Multivariate Max-Stable Spatial Processes," Biometrika, v.102, 2015, p. 215.

30. Suzhi Bi, Rui Zhang, Zhi Ding, and Shuguang Cui. "Wireless Communication in the Era of Big Data," IEEE Communication Magazine, v.53, 2015, p. 190.

31. Weijia Han, Chuan Huang, Jiandong Li, Zan Li, and Shuguang Cui. "Correlation based Spectrum Sensing with Over-sampling in Cognitive Radio," IEEE Journal on Selected Areas of Communications., v.33, 2015, p. 188.

32. Weijia Han, Huiyan Sang, Min Sheng, Jiandong Li, Shuguang Cui. "Efficient Learning of Statistical Primary Patterns via Bayesian Network.," IEEE International Conference on Communications. London, UK., 2015.

33. Yuan Ma, Yue Gao, Ying-Chang Liang, and Shuguang Cui. "Efficient Blind Cooperative Wideband Spectrum Sensing based on Joint Sparsity," IEEE Globecom. Washington DC, 2016.

34. Yunjuan Zang, Feixiang Ni, Zhiyong Feng, Shuguang Cui, and Zhi Ding. "Wavelet Transform Processing Cellular Traffic Prediction in Machine Learning Networks.," IEEE ChinaSIP. Chengdu, China, 2015.

35. D. Duan, L. Yang and S. Cui. "Spectrum Sensing: To Cooperate or Not to Cooperate?," EAI Transactions on Wireless Spectrum, 2014.

36. X. Cheng, L. Yang, and X. Shen. "D2D for Intelligent Transportation Systems: A Feasibility Study," IEEE Transactions on Intelligent Transportation Systems, 2015.

37. R. Zhang, X. Cheng, and L. Yang. "Cooperation via Spectrum Sharing for Physical Layer Security in Device-to-Device Communications Underlaying Cellular Networks," Proceedings of the IEEE Global Communications Conference, 2015.

38. B. Yu, L. Yang, X. Cheng, and R. Cao. "Resource Optimization for Full-Duplex Decode-and-Forward Relaying," IEEE Transactions on Communications, v.63, 2015, p. 4743.

39. D. Wang, R. Zhang, X. Cheng, and L. Yang. "Relay Selection in Two-Way Full- Duplex Energy-Harvesting Relay Networks," Proceedings of the IEEE Global Communications Conference (Globecom), 2016.

40. D. Wang, X. Cheng, and L. Yang. "Joint Power Allocation and Splitting (JoPAS) for SWIPT in Time-Variant Wireless Channels," Proceedings of the IEEE Global Communications Conference (Globecom), 2016.

41. R. Zhang, X. Cheng, and L. Yang. "Cooperation via Spectrum Sharing for Physical Layer Security in Device-to-Device Communications Underlaying Cellular Networks," Proceedings of the IEEE Global Communications Conference, 2015.

42. R. Zhang, X. Cheng, and L. Yang. "Cooperation via Spectrum Sharing for Physical Layer Security in Device-to-Device Communications Underlaying Cellular Networks," IEEE Transactions on Wireless Communications, v.15, 2016.

43. R. Zhang, X. Cheng, and L. Yang. "Joint Power and Access Control for Phys- ical Layer Security in D2D Communications Underlaying Cellular Networks," Proceedings of the IEEE International Conference on Communications (ICC), 2016.

44. R. Zhang, X. Cheng, L. Yang, and B. Jiao. "Interference Graph Based Resource Allocation (InGRA) for D2D Communications Underlaying Cellular Networks," IEEE Transactions on Vehicular Technology, v.64, 2015, p. 3844.

45. T. Yang, R. Zhang, X. Cheng, and L. Yang. "A Graph Coloring Resource Sharing Scheme for Full-Duplex Cellular-VANET Heterogeneous Networks," Proceedings of International Conference on Computing, Networking and Communications (ICNC 2016), 2016.

46. Y. Zou, J. Zhu, L. Yang, Y.-C. Liang, and Y.-D. Yao. "Securing Physical-Layer Communications for Cognitive Radio Networks," IEEE Communications Magazine, v.53, 2016, p. 48.

Broader Impacts

This project will be critical to building future large-scale cognitive wireless communication systems, which could find many applications in both civilian and military sectors. Furthermore, the PIs propose a concurrent and integrated education program, which addresses the undergraduate and graduate education issues in multiple STEM-oriented aspects: teaching, training, and mentoring. The proposed program will provide broader educational opportunities by holding regular seminars for students, industry, and government agencies, posting results on the project website, and publishing papers at technical conferences/journals. Moreover, the program emphasizes outreach to women and other under-represented minorities, with direct involvement of two female faculty PIs and several female students.

The following activities have been taken to broaden the impacts of this project so far:

· The PI presented the research outcomes in several invited talks, including at the University of Miami, University of Central Florida, University of South Florida, and Kansas State University.

· The PI presented the research outcomes to international audience in several invited talks, including Tsinghua University, South China University of Technology, and Guangxi University.

· The PI presented the research outcomes to the UnCoRe REU students at the CSE department of Oakland University in the summer of 2016.

· The two REU students under the PI’s supervision presented our research in the 2016 Mid-SURE symposium (Mid-Michigan Symposium for Undergraduate Research Experience) at Michigan State University.

· The PI and his students presented their research outcomes in several reputed conferences, including IEEE INFOCOM 2014, Globecom 2014, Globecom 2015, ICDCS 2017, ICC 2017, and Milcom 2017.

· The PI has integrated part of the research outcomes in the course materials he is teaching at Auburn University, including COMP 4320 (Computer Networks), COMP 5320/6320/6326 (Design and Analysis of Computer Networks), and COMP 7970 (Cryptographic and non-cryptograhic methods for network security). 

· This project was also introduced to over 1000 high-school students and their parents during the 2017 Open House Engineering Day of the Samuel Ginn College of Engineering at Auburn University. This helps to foster the high-school students' interests in taking science and technology as their future career.