Sponser: National Science Foundation – NeTS Core Program (NSF-NeTS)

PI(s): Dr. Wenye Wang, Dr. Do Young Eun, and Dr. Huaiyu Dai

Duration: October 2018 – September 2021

Scope of the project

In this project, we envision that the richness and complexity of the multi-radio access technology (RAT), multi-band environments can be exploited to our advantage, in a multi-layer fashion, so as to achieve dramatic improvement in resource utilization and system performance. Our main challenge lies in the extreme heterogeneity of spectrum band occupancy and QoS requirement of devices operating on diverse frequency bands with different channel characteristics over a wide range of spatio-temporal scale. Specifically, we aim to undertake the following tasks:

1) obtain and analyze real-time spectrum activities to build multi-layer spectrum snapshots;

2) build a fundamental framework to construct distributed and time/energy-efficient spectrum allocation algorithms under heterogeneous environments with multi-RAT in a principled manner;

3) study how to achieve fast rendezvous in multi-layer spectrum environments;

4) explore abundant opportunities in multi-layer networks for efficient pairwise and broadcast communications via intra and inter-layer topology design.

In progress...

Objectives
In this project, we aim to address how to effectively monitoring spectrum activities in the multi-RAT spectrum environment, and how to efficiently utilize the spectrum resource for pairwise and broadcast communications.

Updates (in progress)
10/01/2018 – 9/30/2019
During this period, a systematic investigation of the multi-RAT, multi-band spectrum environment has been carried out, from the perspectives of spectrum activity modeling, tenancy data collection, dynamic process characterization, and generic algorithms design. Our results and findings can be summarized as follows:

Modeling and Strategy Design of Spectrum Activity Surveillance (SAS)
Spectrum activity surveillance (SAS) is essential to dynamic spectrum access (DSA)-enabled systems, as it collects spectrum usage data for spectrum efficiency improvement, and detecting spectrum culprits, that is, unauthorized spectrum occupants. Large-scale SAS is an open and challenging problem, due to the difficulty in field tests, and lack of performance characterization for strategy design. Capturing the locality of spectrum activities, such as user occupancy and misuse detection, we formulate the SAS process in a 3-factor space, composed of time, spectrum, and geolocational space domains, such that the SAS problem is transformed from a globally collective activity into a set of localized actions, and design objectives are converted from qualitative attributes to quantitative measures. As an application of the proposed model, we also present performance-guaranteed deployment strategies for systems with multiple spectrum monitors. We expect this work to contribute toward a comprehensive spectrum monitoring-utilization model for multi-layer wireless networks, in which combinations of monitoring techniques and monitor deployment strategies can be analyzed and fairly compared.

Modeling of Spectrum Tenancy and Detection of Tenancy Pattern Changes in LTE Systems
Existing spectrum tenancy models suffer from several limitations, including low granularities in measurements, strong assumptions on the underlying statistical characteristics (time varying or stationary), and the lack of multi-channel models. To address these issues, we measure the spectrum tenancy of a LTE cell in time and frequency granularities of 1 millisecond by 180 KHz, i.e., the smallest resource assignment unit in LTE scheduling. We apply the Martingale Test (MT) based change detection algorithm to investigate the stable period, which measures how often the statistical characteristics of the LTE spectrum tenancy change. The average stable period discovered by the change detection algorithm coincides with that found by the brute-force way of fitting the widely used on/off model to the measurement results. We find that fitting vector autoregression (VAR) and on/off models to measurement results divided according to stable periods, discovered by the change detection algorithm, improves fitting accuracy, because the underlying statistical characteristics are stable in each time period. We also investigate the performance of VAR and on/off models in terms of fitting the tenancy of multiple channels, and find that VAR model achieves better performance. Our measurement data collected from a commercial LTE base station also serves as a benchmark for various spectrum tenancy models. We expect this line of work to have a broad impact on various studies of spectrum utilization, such as spectrum occupancy prediction for individual users.

Non-Markovian Monte Carlo (NMMC) Techniques for Directed Graphs
To address the current limitations and challenges around the Markov Chain Monte Carlo (MCMC) techniques that have become widely popular for many purposes, especially their inapplicability to directed graphs with non-reciprocal edges, we develop a framework for directed graphs called Non-Markovian Monte Carlo (NMMC) by establishing a mapping to convert into the quasi-stationary distribution of a carefully constructed transient Markov chain on an extended state space. As applications, we demonstrate how to achieve any given distribution on a directed graph and estimate the eigenvector centrality using a set of non-Markovian, history-dependent random walks on the same graph in a distributed manner. Our NMMC framework as reported builds upon our careful mapping from the proposed chain to the target distribution on a set of transient states, and entails the machinery of the quasi-stationary distribution of a suitably constructed transient chain and the induced random walks with reinforcement to relocate to positions from their past history, which are amenable to distributed implementation using only locally available information.

Characterizing Transient Dynamics of Susceptible-Infected (SI) Epidemics
The susceptible-infected (SI) model has been largely overlooked in the literature, while it is naturally a better fit for modeling the malware propagation in early times when patches/vaccines are not available, or over a wider range of timescales when massive patching is practically infeasible. Nonetheless, its analysis is simply non-trivial, as its important dynamics are all transient and the usual stability/steady-state analysis no longer applies. To this end, we develop a theoretical framework that allows us to obtain an accurate closed-form approximate solution to the original SI dynamics on any arbitrary network, which captures the temporal dynamics over all time and is tighter than the existing approximation, and also to provide a new interpretation via reliability theory. As its applications, we also develop vaccination policies with or without knowledge of already-infected nodes, to mitigate the future epidemic spreading to the extent possible, and demonstrate their effectiveness through numerical simulations.

Multilex Conductance for Gossip-based Information Spreading in Multiplex Networks
We adopt the gossip (random-walk) based model to investigate information spreading in multiplex networks, a special type of multi-layer networks where all layers share the same set of nodes. In practice, the same set of nodes may correspond to individuals who can communicate through multiple networks or platforms, and duplicates of the same node may represent different communication devices or radio access technologies. Two key features of multiplex networks allow potentially much faster information spreading: availability of multiple channels for each user, and more choices on neighbor contacting. We propose a novel metric, multiplex conductance, and use it to effectively quantify the information spreading performance in a general multiplex network. Multiplex conductance is then evaluated for some interesting multiplex networks to facilitate understanding in this new area and motivate effective multiplex network designs. Furthermore, the tradeoff between the cost of additional layers and the improvement of information spreading efficiency is discussed from both the user’s and the network designer’s aspect. Our study sheds light on the impact of structural properties of multiplex networks on information spreading performance, and provides insights for effective multiplex network design.

Interlink Optimization in Multi-layer Interdependent Networks under Cost Constraints
Component (node) failures can be detrimental to multi-layer interdependent networks, to combat which we maximize the robustness of the network by optimizing interlinks under cost constraints. Diverting from the existing approaches which either are intractable for analysis or fail to capture the network dynamics, in this work, we present a surrogate metric based framework for constructing interlinks to maximize the network robustness. In particular, we focus on three representative mechanisms of failure propagation, and propose metrics to track the network robustness for each of these mechanisms. Owing to their mathematical tractability, these metrics allow us to optimize the interlink structure to enhance robustness. Furthermore, we are able to introduce the cost of construction into the interlink design problem, a practical feature largely ignored in relevant literature. We simulate the failure cascades on real world networks to compare the performance of our interlinking strategies with the state of the art heuristics and demonstrate their effectiveness.

Thesis/Dissertations

Seyyedali Hosseinalipour, “Efficient Network Planning and Design for Cloud, Interdependent, and UAV-assisted Networks,” Ph.D. dissertation, Dept. Elect.& Comp. Eng., NC State University, Raleigh, NC, 2020.

Srinjoy Chattopadhyay, “Optimizing and Controlling Spreading Processes in Multilayer Networks,” Ph.D. dissertation, Dept. Elect.& Comp. Eng., NC State University, Raleigh, NC, 2020.

Yufan Huang, “Spreading Processes in Complex Networks: Speed, Competition, and Privacy,” Ph.D. dissertation, Dept. Elect.& Comp. Eng., NC State University, Raleigh, NC, 2019.

Jie Wang, “Modeling and Analysis of Mobile Data Dynamics in Heterogeneous Wireless Networks.”Ph.D. dissertation, Dept. Elect.& Comp. Eng., NC State University, Raleigh, NC, 2019.

Conference Publications

R. Zou, W. Wang, and H. Dai, “Temporal and Spectral Analysis of Spectrum Hole Distributions in an LTE Cell,” 2021 IEEE Global Communications Conference (GLOBECOM), Madrid, Spain, Dec. 2021.

Y. Huang, R. Jin, and H. Dai, “Differential Privacy and Prediction Uncertainty of Gossip Protocols in General Networks,” 2020 IEEE Global Communications Conference (GLOBECOM), Taipei, Taiwan, Dec. 2020.

Jie Hu, Vishwaraj Doshi, and Do Young Eun, “Opportunistic Spectrum Access: Does Maximizing Throughput Minimize File Transfer Time?”, in WiOpt, Oct. 2021

Vishwaraj Doshi, Shailaja Mallick, and Do Young Eun, “Competing Epidemics on Graphs – Global Convergence and Coexistence”, in IEEE INFOCOM, May 2021

Doshi, Vishwaraj and Eun, Do Young, “Fiedler Vector Approximation via Interacting Random Walks,” in Proceedings of the ACM on Measurement and Analysis of Computing Systems, September 2020. Deposited in NSF-PAR doi:10.1145/3379502.

Rui Zou and Wenye Wang, “U-CIMAN: Uncover Spectrum and User Information in LTE Mobile Access Networks,” in Proc. of IEEE INFOCOM, July 2020. pdf DOI: 10.1109/INFOCOM41043.2020.9155245.

Rui Zou and Wenye Wang, “Change Detection Based Segmentation and Modeling of LTE Spectrum Tenancy,” In Proc. oof IEEE GLOBECOM, December 2019. pdf DOI: 10.1109/GLOBECOM38437.2019.9013749.

Chul-Ho Lee, Srinivas Tenneti, and Do Young Eun, “Transient Dynamics of Epidemic Spreading and its Mitigation on Large Networks”, in ACM MobiHoc, Catania, Italy, July 2019 (Best Paper Award Finalists).

Chul-Ho Lee, Min Kang, and Do Young Eun, “Non-Markovian Monte Carlo on Directed Graphs” in ACM SIGMETRICS, Phoenix, AZ, June 2019.

Jie Wang, Wenye Wang, and Cliff Wang, “SAS: Modeling and Analysis of Spectrum Activity Surveillance in Wireless Overlay Networks,” in Proceedings of IEEE INFOCOM 2019 – The 38th Annual IEEE International Conference on Computer Communications (INFOCOM 2019), Paris, France, Apr. 2019.

Rui Zou and Wenye Wang, “Change Detection Based Segmentation and Modeling of {LTE} Spectrum Tenancy,” in Proceedings of 2019 IEEE Global Communications Conference: Cognitive Radio and AI-Enabled Network Symposium (Globecom2019 CRAEN), Waikoloa, USA, Dec. 2019.

Journal Publications

Rui Zou, Wenye Wang, “Downlink Decoding Based Accurate Measurement of LTE Spectrum Tenancy”, IEEE Transactions on Mobile Computing, 2021 (Accepted).

Teng Fei, Wenye Wang, “The Vulnerability and Enhancement of AKA Protocol for Mobile Authentication in LTE/5G Networks”, Computer Networks, 2021 (under review).

Jie Wang, Wenye Wang, Cliff Wang, Min Song, “Spectrum Activity Surveillance: Modeling and Analysis from Perspectives of Surveillance Coverage and Culprit Detection”, IEEE Transactions on Mobile Computing, October 2020.

S. Chattopadhyay, H. Dai, and D. Y. Eun, “Controlling Metastable Infection Patterns in Multilayer Networks via Interlink Design,” to appear in IEEE Trans. Network Science and Engineering, DOI: 10.1109/TNSE.2021.3108075.

S. Hosseinalipour, J. Mao, D. Y. Eun, and H. Dai, “Prevention and Mitigation of Catastrophic Failures in Demand-Supply Interdependent Networks,” IEEE Trans. Network Science and Engineering, vol. 7, no. 3, pp. 1710-1723, July-September, 2020.

Doshi, Vishwaraj and Young Eun, Do. (2020). “Fiedler Vector Approximation via Interacting Random Walks”. ACM SIGMETRICS Performance Evaluation Review. 48 (1) 101 to 102 pdf doi:10.1145/3410048.3410107.

Srinjoy Chattopadhyay, Huaiyu Dai, and Do Young Eun, “Maximization of Robustness of Interdependent Networks under Budget Constraints,” IEEE Trans. Network Science and Engineering, to appear. 10.1109/TNSE.2019.2935068.

Yufan Huang and Huaiyu Dai, “Multiplex Conductance and Gossip Based Information Spreading in Multiplex Networks,” IEEE Trans. Network Science and Engineering, vol. 6, no. 3, pp. 391-401, July-Sept., 2019.

Chul-Ho Lee, Min Kang, and Do Young Eun, “Non-Markovian Monte Carlo on Directed Graphs” in Proceedings of the ACM on Measurement and Analysis of Computing Systems, June 2019.