Pub Date : 2021-11-29DOI: 10.1109/MILCOM52596.2021.9652972
R. Axford
Margin and availability are routinely used in transponded satellite communications (SATCOM) to specify required reliability. The objective of this paper is to provide a thorough treatment of these two metrics. Drawing from both literature and experience, it collects a set of validated techniques into a concise package. It focusses on methods that are easily applied and widely used. The paper also offers new insights.
{"title":"Margin and Availability in Transponded SATCOM","authors":"R. Axford","doi":"10.1109/MILCOM52596.2021.9652972","DOIUrl":"https://doi.org/10.1109/MILCOM52596.2021.9652972","url":null,"abstract":"Margin and availability are routinely used in transponded satellite communications (SATCOM) to specify required reliability. The objective of this paper is to provide a thorough treatment of these two metrics. Drawing from both literature and experience, it collects a set of validated techniques into a concise package. It focusses on methods that are easily applied and widely used. The paper also offers new insights.","PeriodicalId":187645,"journal":{"name":"MILCOM 2021 - 2021 IEEE Military Communications Conference (MILCOM)","volume":"48 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121750419","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-11-29DOI: 10.1109/MILCOM52596.2021.9653113
Ru Xie, Liming Wang, Xiaojie Tao
Cloud computing is gaining popularity due to prominent advantages of dynamic provisioning, economies of scale and low expenditures. However, co-resident attacks pose great threats to security and reliability of cloud infrastructure. Previous work has shown the effectiveness of secure virtual machine (VM) allocation strategies to defend against attacks and improve security. Unfortunately, existing approaches cannot distinguish potential malicious tenants before running VMs, so they adopt a reckless strategy of stacking VMs, which indirectly mitigates threats but fails to provide adequate security or balance workload. This paper presents an approach to reduce attack risk and balance workload by recognizing potential attackers before VM allocation and applying a secure allocation strategy to prevent malicious tenants from accessing normal ones. We analyze tenant behavior and VM usage data to identify potential attackers, assisted by machine learning methods. A new metric is proposed to measure co-resident attack risk and a novel risk-control VM allocation strategy is designed to minimize it. Implementation and evaluation on a dataset consisting of real-world VM workload demonstrate that our approach significantly outperforms existing approaches in minimizing the risk of co-resident attacks and balancing workload of datacenter as well as individual tenants.
{"title":"A Secure VM Allocation Strategy Based on Tenant Behavior Analysis and Anomaly Identification","authors":"Ru Xie, Liming Wang, Xiaojie Tao","doi":"10.1109/MILCOM52596.2021.9653113","DOIUrl":"https://doi.org/10.1109/MILCOM52596.2021.9653113","url":null,"abstract":"Cloud computing is gaining popularity due to prominent advantages of dynamic provisioning, economies of scale and low expenditures. However, co-resident attacks pose great threats to security and reliability of cloud infrastructure. Previous work has shown the effectiveness of secure virtual machine (VM) allocation strategies to defend against attacks and improve security. Unfortunately, existing approaches cannot distinguish potential malicious tenants before running VMs, so they adopt a reckless strategy of stacking VMs, which indirectly mitigates threats but fails to provide adequate security or balance workload. This paper presents an approach to reduce attack risk and balance workload by recognizing potential attackers before VM allocation and applying a secure allocation strategy to prevent malicious tenants from accessing normal ones. We analyze tenant behavior and VM usage data to identify potential attackers, assisted by machine learning methods. A new metric is proposed to measure co-resident attack risk and a novel risk-control VM allocation strategy is designed to minimize it. Implementation and evaluation on a dataset consisting of real-world VM workload demonstrate that our approach significantly outperforms existing approaches in minimizing the risk of co-resident attacks and balancing workload of datacenter as well as individual tenants.","PeriodicalId":187645,"journal":{"name":"MILCOM 2021 - 2021 IEEE Military Communications Conference (MILCOM)","volume":"72 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126442729","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-11-29DOI: 10.1109/MILCOM52596.2021.9653058
C. Kam, S. Kompella, A. Ephremides
We study the problem of a two-user, single-channel cognitive radio network, in which the objective is to maximize the secondary user (SU) throughput subject to a constraint on the probability of collision experienced by the primary user (PU). The transmit/idle dynamics of the primary user is modeled as a binary Markov chain, and the secondary senses the channel and decides on its transmission and sensing strategy based on the estimated evolution of the primary user transmission state. Because of the Markov model of the primary user dynamics, the age of the information sensed by the secondary has an impact on its belief when it is not sensing. We apply a Lyapunov optimization algorithm to solve the constrained throughput optimization problem, which utilizes the AoI-dependent PU idle/transmit probability to make the SU's sense/transmit decision in each slot. We then apply the Lyapunov framework to identify the tradeoff between three fundamental information qualities: AoI, accuracy, and completeness. Characterizing these types of tradeoffs can be a useful intermediate step towards optimizing a variety of objectives.
{"title":"The Role of AoI in a Cognitive Radio Network: Lyapunov Optimization and Tradeoffs","authors":"C. Kam, S. Kompella, A. Ephremides","doi":"10.1109/MILCOM52596.2021.9653058","DOIUrl":"https://doi.org/10.1109/MILCOM52596.2021.9653058","url":null,"abstract":"We study the problem of a two-user, single-channel cognitive radio network, in which the objective is to maximize the secondary user (SU) throughput subject to a constraint on the probability of collision experienced by the primary user (PU). The transmit/idle dynamics of the primary user is modeled as a binary Markov chain, and the secondary senses the channel and decides on its transmission and sensing strategy based on the estimated evolution of the primary user transmission state. Because of the Markov model of the primary user dynamics, the age of the information sensed by the secondary has an impact on its belief when it is not sensing. We apply a Lyapunov optimization algorithm to solve the constrained throughput optimization problem, which utilizes the AoI-dependent PU idle/transmit probability to make the SU's sense/transmit decision in each slot. We then apply the Lyapunov framework to identify the tradeoff between three fundamental information qualities: AoI, accuracy, and completeness. Characterizing these types of tradeoffs can be a useful intermediate step towards optimizing a variety of objectives.","PeriodicalId":187645,"journal":{"name":"MILCOM 2021 - 2021 IEEE Military Communications Conference (MILCOM)","volume":"71 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115064896","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-11-29DOI: 10.1109/MILCOM52596.2021.9652907
Jinyang Li, Runyu Ma, Vikram Sharma Mailthody, Colin Samplawski, Benjamin M. Marlin, Songqing Chen, Shuochao Yao, T. Abdelzaher
Convolutional Neural Networks (CNN) have shown great success in many sensing and recognition applications. However, the excessive resource demand remains a major barrier against their deployment on low-end devices. Optimizations, such as model compression, are thus a need for practical deployment. To fully exploit existing system resources, platform-aware optimizations emerged in recent years, where an execution-time model becomes a necessity. However, non-monotonicity over the network configuration space makes execution time modeling a challenging task. Data-driven approaches have the advantage of being portable over different platforms by treating the hardware and software stack as a black box but at the cost of extremely long profiling time. On the other hand, analytical models can be found in the architecture and system literature that do not need heavy profiling but require laborious analysis by domain experts. In this paper, we focus on building a general latency model for convolutional layers that account for the majority of the total execution time in CNN models. We identify two major non-linear modes in the relationship between latency and convolution parameters, and analyze the mechanism behind them. The resulting model has better interpretability and can reduce profiling workload. The evaluation results show that our model outperforms baselines on different platforms and CNN models.
{"title":"Towards an Accurate Latency Model for Convolutional Neural Network Layers on GPUs","authors":"Jinyang Li, Runyu Ma, Vikram Sharma Mailthody, Colin Samplawski, Benjamin M. Marlin, Songqing Chen, Shuochao Yao, T. Abdelzaher","doi":"10.1109/MILCOM52596.2021.9652907","DOIUrl":"https://doi.org/10.1109/MILCOM52596.2021.9652907","url":null,"abstract":"Convolutional Neural Networks (CNN) have shown great success in many sensing and recognition applications. However, the excessive resource demand remains a major barrier against their deployment on low-end devices. Optimizations, such as model compression, are thus a need for practical deployment. To fully exploit existing system resources, platform-aware optimizations emerged in recent years, where an execution-time model becomes a necessity. However, non-monotonicity over the network configuration space makes execution time modeling a challenging task. Data-driven approaches have the advantage of being portable over different platforms by treating the hardware and software stack as a black box but at the cost of extremely long profiling time. On the other hand, analytical models can be found in the architecture and system literature that do not need heavy profiling but require laborious analysis by domain experts. In this paper, we focus on building a general latency model for convolutional layers that account for the majority of the total execution time in CNN models. We identify two major non-linear modes in the relationship between latency and convolution parameters, and analyze the mechanism behind them. The resulting model has better interpretability and can reduce profiling workload. The evaluation results show that our model outperforms baselines on different platforms and CNN models.","PeriodicalId":187645,"journal":{"name":"MILCOM 2021 - 2021 IEEE Military Communications Conference (MILCOM)","volume":"214 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134178298","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-11-29DOI: 10.1109/MILCOM52596.2021.9653073
Leila Mounsif, Damien Roque, C. Poulliat
In this paper, we study the joint phase and timing estimation problem in Faster-than-Nyquist (FTN) systems. We use the Cramér-Rao lower bound (CRB) as a cost function to design optimal pilot sequences subject to an energy constraint. We rely on harmonic approximations of the bound to establish closed-form relations between the ultimate synchronization performance and the transmitted waveform characteristics (i.e., pulse shaping filter and signaling density). We show that increasing the symbol rate at fixed bandwidth is beneficial to phase estimation and detrimental to timing synchronization. Therefore, we propose a joint timing/phase pilot design to accommodate this trade-off. Lastly, we illustrate the strengths of the proposed pilots with respect to traditional Zadoff–Chu sequences in presence of a residual carrier frequency offset.
{"title":"Optimal Pilot Sequences for Phase and Timing Synchronization in FTN Systems","authors":"Leila Mounsif, Damien Roque, C. Poulliat","doi":"10.1109/MILCOM52596.2021.9653073","DOIUrl":"https://doi.org/10.1109/MILCOM52596.2021.9653073","url":null,"abstract":"In this paper, we study the joint phase and timing estimation problem in Faster-than-Nyquist (FTN) systems. We use the Cramér-Rao lower bound (CRB) as a cost function to design optimal pilot sequences subject to an energy constraint. We rely on harmonic approximations of the bound to establish closed-form relations between the ultimate synchronization performance and the transmitted waveform characteristics (i.e., pulse shaping filter and signaling density). We show that increasing the symbol rate at fixed bandwidth is beneficial to phase estimation and detrimental to timing synchronization. Therefore, we propose a joint timing/phase pilot design to accommodate this trade-off. Lastly, we illustrate the strengths of the proposed pilots with respect to traditional Zadoff–Chu sequences in presence of a residual carrier frequency offset.","PeriodicalId":187645,"journal":{"name":"MILCOM 2021 - 2021 IEEE Military Communications Conference (MILCOM)","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133612292","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-11-29DOI: 10.1109/MILCOM52596.2021.9653045
Autumn Twitchell, M. Rice
Alamouti-encoded 16-APSK can solve the two-antenna problem in aeronautical mobile telemetry, but detection and decoding must be performed when the propagation delays from the two transmit antennas are different. The ML sequence estimator is derived for Alamouti-encoded 16-APSK in the presence of a differential delay. The ML estimator requires too many states to be practical. A reduced-complexity sequence estimator is shown to have a BER performance within 1 dB of the single-input/single-output 16-APSK in the AWGN environment.
{"title":"On Alamouti-Encoded 16-APSK in Aeronautical Mobile Telemetry","authors":"Autumn Twitchell, M. Rice","doi":"10.1109/MILCOM52596.2021.9653045","DOIUrl":"https://doi.org/10.1109/MILCOM52596.2021.9653045","url":null,"abstract":"Alamouti-encoded 16-APSK can solve the two-antenna problem in aeronautical mobile telemetry, but detection and decoding must be performed when the propagation delays from the two transmit antennas are different. The ML sequence estimator is derived for Alamouti-encoded 16-APSK in the presence of a differential delay. The ML estimator requires too many states to be practical. A reduced-complexity sequence estimator is shown to have a BER performance within 1 dB of the single-input/single-output 16-APSK in the AWGN environment.","PeriodicalId":187645,"journal":{"name":"MILCOM 2021 - 2021 IEEE Military Communications Conference (MILCOM)","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133173406","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-11-29DOI: 10.1109/MILCOM52596.2021.9653067
H. Khalifé, Raphael Naves, D. Thebault, V. Conan
Named Data Networking (NDN) promotes a data-centric approach for content distribution by uncoupling the data from hosts and its location in the network. Initially targeting wired networks, the NDN extension to constrained wireless and error prone environments raises several concerns. In this paper we propose NDN-FASYM, a forwarding solution that efficiently copes with links asymmetry, a common challenge in wireless networks and in particular in military deployments. With NDN-FASYM, a node filters Interest messages for a particular data and forwards them only if the reverse link (from the receiver to the sender) is capable to ensure the Quality of Service (QoS) requested by the data. Our validation over a real testbed deployment shows that our solution is able to successfully satisfy the data QoS request and reduces the overhead in terms of sent information in the network. Interestingly, NDN-FASYM requires limited additions to legacy implementations and maintains backward compatibility with NDN standard specifications.
{"title":"Optimized NDN Forwarding in Tactical Networks with Asymmetric Radio Links","authors":"H. Khalifé, Raphael Naves, D. Thebault, V. Conan","doi":"10.1109/MILCOM52596.2021.9653067","DOIUrl":"https://doi.org/10.1109/MILCOM52596.2021.9653067","url":null,"abstract":"Named Data Networking (NDN) promotes a data-centric approach for content distribution by uncoupling the data from hosts and its location in the network. Initially targeting wired networks, the NDN extension to constrained wireless and error prone environments raises several concerns. In this paper we propose NDN-FASYM, a forwarding solution that efficiently copes with links asymmetry, a common challenge in wireless networks and in particular in military deployments. With NDN-FASYM, a node filters Interest messages for a particular data and forwards them only if the reverse link (from the receiver to the sender) is capable to ensure the Quality of Service (QoS) requested by the data. Our validation over a real testbed deployment shows that our solution is able to successfully satisfy the data QoS request and reduces the overhead in terms of sent information in the network. Interestingly, NDN-FASYM requires limited additions to legacy implementations and maintains backward compatibility with NDN standard specifications.","PeriodicalId":187645,"journal":{"name":"MILCOM 2021 - 2021 IEEE Military Communications Conference (MILCOM)","volume":"42 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131794705","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-11-29DOI: 10.1109/MILCOM52596.2021.9653044
Robert E. Anderson, A. Berg, James R. Carter, Dana DeFrancesco, Ronald W. Elliott, Jessica L. Gregris, Donald W. Hershberger, William Y. Joo, Dahesh A. Khalil, Jeremy R. Kozee, Philip C. Minor, Pamela A. Phillips, A. Vigil
DOD SATCOM Earth segment architectures are poised to migrate to Digital IF as the medium of choice. This migration is already underway within commercial SATCOM. Motivators in the case of DOD SATCOM include mission demand for resilience, carrier count congestion, RF bandwidth growth, operational demand for virtualization, modem value proposition, and environmental hardening. Migration at DOD SATCOM gateways is anticipated in three phases, commencing now with foundational groundwork, leading soon to baseline infrastructure, then followed by advanced initiatives such as virtualization and management automation. This migration is being supported by updates to necessary MIL standards and WGS Performance Certification processes to enable WGS access. Subsequent enterprise gateway deployment and robust terrestrial mesh networking achieves SATCOM resilience and mission planning flexibility while enabling the aggregation of baseband operations to regional facilities. Digital IF is fast maturing. Digital IF technology investment is being made and the IEEE-ISTO 4900–2021 “DIFI Standard” for interoperability has just been published. These developments will enable growth and reduce overall lifecycle costs for all manner of communication systems. On all counts, a comprehensive migration to Digital IF, well under way commercially, is also now beginning for DOD SATCOM Earth segment architectures.
{"title":"Migration of DOD SATCOM Earth Segment Architectures to Digital IF","authors":"Robert E. Anderson, A. Berg, James R. Carter, Dana DeFrancesco, Ronald W. Elliott, Jessica L. Gregris, Donald W. Hershberger, William Y. Joo, Dahesh A. Khalil, Jeremy R. Kozee, Philip C. Minor, Pamela A. Phillips, A. Vigil","doi":"10.1109/MILCOM52596.2021.9653044","DOIUrl":"https://doi.org/10.1109/MILCOM52596.2021.9653044","url":null,"abstract":"DOD SATCOM Earth segment architectures are poised to migrate to Digital IF as the medium of choice. This migration is already underway within commercial SATCOM. Motivators in the case of DOD SATCOM include mission demand for resilience, carrier count congestion, RF bandwidth growth, operational demand for virtualization, modem value proposition, and environmental hardening. Migration at DOD SATCOM gateways is anticipated in three phases, commencing now with foundational groundwork, leading soon to baseline infrastructure, then followed by advanced initiatives such as virtualization and management automation. This migration is being supported by updates to necessary MIL standards and WGS Performance Certification processes to enable WGS access. Subsequent enterprise gateway deployment and robust terrestrial mesh networking achieves SATCOM resilience and mission planning flexibility while enabling the aggregation of baseband operations to regional facilities. Digital IF is fast maturing. Digital IF technology investment is being made and the IEEE-ISTO 4900–2021 “DIFI Standard” for interoperability has just been published. These developments will enable growth and reduce overall lifecycle costs for all manner of communication systems. On all counts, a comprehensive migration to Digital IF, well under way commercially, is also now beginning for DOD SATCOM Earth segment architectures.","PeriodicalId":187645,"journal":{"name":"MILCOM 2021 - 2021 IEEE Military Communications Conference (MILCOM)","volume":"273 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132182979","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-11-29DOI: 10.1109/MILCOM52596.2021.9652883
P. Draxler, Anjali Sharma, Dane Malangone, Donald E. Cowles, Eric Brown, Jonmei J. Yan
An efficient, model based, digital predistortion (DPD) is presented for high efficiency wide bandwidth envelope tracking (ET) power amplifiers (ETPA) with multi-carrier results. This ET configuration is specifically designed for wideband carrier aggregation (CA) signals that have >200MHz bandwidth and high peak to average ratio. The DPD is based on a new formulation of the inverse RFPA surface which enables improved sparse signal processing techniques. This algorithm has been implemented for open loop and a batch mode closed loop operation. This model and these techniques reduce the number of inverse model terms significantly, using 40 terms of 724, while achieving reduced spectral emissions, improved computational efficiency, linearity and power efficiency. These improvements are achieved by calculating the covariance matrix within the Doubly Orthogonal Matching Pursuit (DOMP) algorithm to assist with parametric estimation through Reduced Computation-DOMP (RC-DOMP) to replace the Moore-Penrose inverse. With all these techniques, we can perform parameter estimation updates within 1 second on a Xilinx ZCU208 RFSOC achieving specification compliant performance, enabling future efficient real time closed loop operation. Measurements of a 4 CA signal with over 200MHz signal bandwidth, around 3.7 GHz is presented, with peak to average ratio of 6dB after crest factor reduction, Po>38.9dBm, EMV <4%, and <-35dBc spectral regrowth with DPD.
{"title":"Computationally Efficient 2-D Predistortion for Multi-Carrier Envelope Tracking Power Amplifiers","authors":"P. Draxler, Anjali Sharma, Dane Malangone, Donald E. Cowles, Eric Brown, Jonmei J. Yan","doi":"10.1109/MILCOM52596.2021.9652883","DOIUrl":"https://doi.org/10.1109/MILCOM52596.2021.9652883","url":null,"abstract":"An efficient, model based, digital predistortion (DPD) is presented for high efficiency wide bandwidth envelope tracking (ET) power amplifiers (ETPA) with multi-carrier results. This ET configuration is specifically designed for wideband carrier aggregation (CA) signals that have >200MHz bandwidth and high peak to average ratio. The DPD is based on a new formulation of the inverse RFPA surface which enables improved sparse signal processing techniques. This algorithm has been implemented for open loop and a batch mode closed loop operation. This model and these techniques reduce the number of inverse model terms significantly, using 40 terms of 724, while achieving reduced spectral emissions, improved computational efficiency, linearity and power efficiency. These improvements are achieved by calculating the covariance matrix within the Doubly Orthogonal Matching Pursuit (DOMP) algorithm to assist with parametric estimation through Reduced Computation-DOMP (RC-DOMP) to replace the Moore-Penrose inverse. With all these techniques, we can perform parameter estimation updates within 1 second on a Xilinx ZCU208 RFSOC achieving specification compliant performance, enabling future efficient real time closed loop operation. Measurements of a 4 CA signal with over 200MHz signal bandwidth, around 3.7 GHz is presented, with peak to average ratio of 6dB after crest factor reduction, Po>38.9dBm, EMV <4%, and <-35dBc spectral regrowth with DPD.","PeriodicalId":187645,"journal":{"name":"MILCOM 2021 - 2021 IEEE Military Communications Conference (MILCOM)","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133277320","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-11-29DOI: 10.1109/MILCOM52596.2021.9652915
Peter Martin, Jian Fan, Taejin Kim, W. Vesey, Lloyd G. Greenwald
Deep learning (DL) models have been shown to be vulnerable to adversarial attacks. DL model security against adversarial attacks is critical to using DL-trained models in forward deployed systems, e.g. facial recognition, document characterization, or object detection. We provide results and lessons learned applying a moving target defense (MTD) strategy against iterative, gradient-based adversarial attacks. Our strategy involves (1) training a diverse ensemble of DL models, (2) applying randomized affine input transformations to inputs, and (3) randomizing output decisions. We report a primary lesson that this strategy is ineffective against a white-box adversary, which could completely circumvent output randomization using a deterministic surrogate. We reveal how our ensemble models lacked the diversity necessary for effective MTD. We also evaluate our MTD strategy against a black-box adversary employing an ensemble surrogate model. We conclude that an MTD strategy against black-box adversarial attacks crucially depends on lack of transferability between models.
{"title":"Toward Effective Moving Target Defense Against Adversarial AI","authors":"Peter Martin, Jian Fan, Taejin Kim, W. Vesey, Lloyd G. Greenwald","doi":"10.1109/MILCOM52596.2021.9652915","DOIUrl":"https://doi.org/10.1109/MILCOM52596.2021.9652915","url":null,"abstract":"Deep learning (DL) models have been shown to be vulnerable to adversarial attacks. DL model security against adversarial attacks is critical to using DL-trained models in forward deployed systems, e.g. facial recognition, document characterization, or object detection. We provide results and lessons learned applying a moving target defense (MTD) strategy against iterative, gradient-based adversarial attacks. Our strategy involves (1) training a diverse ensemble of DL models, (2) applying randomized affine input transformations to inputs, and (3) randomizing output decisions. We report a primary lesson that this strategy is ineffective against a white-box adversary, which could completely circumvent output randomization using a deterministic surrogate. We reveal how our ensemble models lacked the diversity necessary for effective MTD. We also evaluate our MTD strategy against a black-box adversary employing an ensemble surrogate model. We conclude that an MTD strategy against black-box adversarial attacks crucially depends on lack of transferability between models.","PeriodicalId":187645,"journal":{"name":"MILCOM 2021 - 2021 IEEE Military Communications Conference (MILCOM)","volume":"15 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134074517","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：