TY - CONF
T1 - A Framework for Analyzing Spectrum Characteristics in Large Spatio-temporal Scales
AU - Zeng, Yijing
AU - Chandrasekaran, Varun
AU - Banerjee, Suman
AU - Giustiniano, Domenico
N1 - Funding Information:
Spectrum data compression: Airpress [65] also noted the scalability issue of spectrum inventory. Thus, it mainly focuses on how to minimize the size of data with maximal compression ratio 64. We take a step further and consider data compression as a preprocessing step to transfer the data into a less complex space with signal features retained, so that we can enable different apps efficiently. 10 CONCLUSIONS We have presented BigSpec, a general-purpose framework that can enable different spectrum related apps efficiently on large volume of spectrum data. Although we only evaluate the performance of BigSpec using three example apps in this paper, we believe that the key idea of BigSpec enables us to gain a deeper understanding of spectrum utilization in large spatio-temporal scales with little prior knowledge. We envision BigSpec to be extended with other building blocks to enable more interesting apps by the community in the future. We foresee that the new insights generated using BigSpec are of considerable value in assisting users, service providers, and regulation authorities to better measure and utilize spectrum. ACKNOWLEDGEMENT We thank our shepherd Ashutosh Sabharwal and the anonymous reviewers for their detailed feedback. We are grateful to Madison Metro bus for letting us collecting data, and Steve Bauder at Wisconsin Public Broadcasting for answering our questions about TV spectrum. We appreciate our lab mates for setting up the data collection platform, Jerry Zhu for useful discussions at the early stage of this project, and Robin Corcos for proofreading early versions of this paper. Yijing Zeng, Varun Chandrasekaran, and Suman Banerjee were supported in part by US National Science Foundation grant CNS-1838733, CNS-1719336, CNS-1647152, and CNS-1629833. Domenico Giustiniano was sponsored in part by the NATO Science for Peace and Security Programme under grant G5461, and Madrid Regional Government through TAPIR-CM project S2018/TCS-4496.
Publisher Copyright:
© 2019 Association for Computing Machinery.
PY - 2019
Y1 - 2019
N2 - Understanding spectrum characteristics with little prior knowledge requires fine-grained spectrum data in the frequency, spatial, and temporal domains; gathering such a diverse set of measurements results in a large data volume. Analysis of the resulting dataset poses unique challenges; methods in the status quo are tailored for specific spectrum-related applications (apps), and are ill equipped to process data of this magnitude. In this paper, we design BigSpec, a generalpurpose framework that allows for fast processing of apps. The key idea is to reduce computation costs by performing computation extensively on compressed data that preserves signal features. Adhering to this guideline, we build solutions for three apps, i.e., energy detection, spatio-temporal spectrum estimation, and anomaly detection. These apps were chosen to highlight BigSpec's efficiency, scalability, and extensibility. To evaluate BigSpec's performance, we collect more than 1 terabyte of spectrum data spanning a year, across 300MHz-4GHz, covering 400 km2. Compared with baselines and prior works, we achieve 17× run time efficiency, sublinear rather than linear run time scalability, and extend the definition of anomaly to different domains (frequency & spatio-temporal). We also obtain high-level insights from the data to provide valuable advice on future spectrum measurement and data analysis.
AB - Understanding spectrum characteristics with little prior knowledge requires fine-grained spectrum data in the frequency, spatial, and temporal domains; gathering such a diverse set of measurements results in a large data volume. Analysis of the resulting dataset poses unique challenges; methods in the status quo are tailored for specific spectrum-related applications (apps), and are ill equipped to process data of this magnitude. In this paper, we design BigSpec, a generalpurpose framework that allows for fast processing of apps. The key idea is to reduce computation costs by performing computation extensively on compressed data that preserves signal features. Adhering to this guideline, we build solutions for three apps, i.e., energy detection, spatio-temporal spectrum estimation, and anomaly detection. These apps were chosen to highlight BigSpec's efficiency, scalability, and extensibility. To evaluate BigSpec's performance, we collect more than 1 terabyte of spectrum data spanning a year, across 300MHz-4GHz, covering 400 km2. Compared with baselines and prior works, we achieve 17× run time efficiency, sublinear rather than linear run time scalability, and extend the definition of anomaly to different domains (frequency & spatio-temporal). We also obtain high-level insights from the data to provide valuable advice on future spectrum measurement and data analysis.
KW - Spatio-temporal data analysis
KW - Spectrum measurement
UR - http://www.scopus.com/inward/record.url?scp=85088102089&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85088102089&partnerID=8YFLogxK
U2 - 10.1145/3300061.3345450
DO - 10.1145/3300061.3345450
M3 - Paper
AN - SCOPUS:85088102089
T2 - 25th Annual International Conference on Mobile Computing and Networking, MobiCom 2019
Y2 - 21 October 2019 through 25 October 2019
ER -