Nexcepta Showcased NextG Communications and AI Technologies at AUSA 2024
Oct. 16, 2024
Exciting Day for Nexcepta, Inc. at the Association of the United States Army - AUSA Annual Meeting and Exposition!
We're thrilled to share that our team has exhibited in Washington, DC, showcasing our cutting-edge technologies in NextG secure communications, spectrum awareness, and multi-domain AI. The response has been incredible, and we've generated significant interest in how our innovations can enhance critical network operations, communications, and data services.
Thank you to everyone who visited our booth and engaged with our team. We’re looking forward to the conversations and potential collaborations that lie ahead.
Paper accepted to the ACM Workshop on Machine Learning for NextG Networks at ACM MobiCom
Sep. 13, 2024
A paper from Nexcepta authors has been accepted to the ACM Workshop on Machine Learning for NextG Networks in conjunction with ACM Conference on Mobile Computing and Networking (MobiCom): “Low-Latency Task-Oriented Communications with Multi-Round, Multi-Task Deep Learning”. Please see below for details.
Title Low-Latency Task-Oriented Communications with Multi-Round, Multi-Task Deep Learning
Abstract: In this paper, we address task-oriented (or goal-oriented) communications where an encoder at the transmitter learns compressed latent representations of data, which are then transmitted over a channel. At the receiver, a decoder performs a machine learning task, specifically for classifying the received signals. The deep neural networks corresponding to the encoder-decoder pair are jointly trained, taking both channel and data characteristics into account. Our objective is to achieve high accuracy in the underlying task while minimizing the number of channel uses determined by the encoder’s output size. To this end, we propose a dynamic update of channel uses in multi-round transmissions. The transmitter incrementally sends an increasing number of encoded samples over the channel based on the feedback from the receiver, and the receiver utilizes the signals from a previous round to enhance task performance, rather than only considering the latest transmission. This approach employs multi-task learning to jointly optimize accuracy across varying number of channel uses, treating each configuration as a distinct task. Our method leverages the difference in distributions observed at the final layer output of the decoder between correct and incorrect decisions. The decision to allocate additional channel uses is based on comparing the maximum of decoder's final layer output entries with a threshold, which can be adjusted to meet specific accuracy and latency requirements. We characterize both the accuracy and the delay (total number of channel uses) of our method, demonstrating that it achieves the accuracy of conventional approaches requiring large numbers of channel uses, but with reduced delay by incorporating signals from a prior round. We consider the CIFAR-10 dataset, convolutional neural network architectures, and AWGN and Rayleigh channel models for performance evaluation. Our results show that this multi-round, multi-task learning (MRMTL) approach significantly improves the efficiency of task-oriented communications, balancing accuracy and latency effectively.
Papers accepted to IEEE MILCOM
June 15, 2024
Two papers from Nexcepta authors have been accepted to IEEE Military Communications (MILCOM) Conference: “Integrated Sensing and Communications to Uncover Moving Targets in NextG Networks” and “Continual Deep Reinforcement Learning to Prevent Catastrophic Forgetting in Jamming Mitigation”. See below for details.
Title: I-SCOUT: Integrated Sensing and Communications to Uncover Moving Targets in NextG Networks
Abstract: Integrated Sensing and Communication (ISAC) represents a transformative approach within 5G and beyond, aiming to merge wireless communication and sensing functionalities into a unified network infrastructure. This integration offers enhanced spectrum efficiency, real-time situational awareness, cost and energy reductions, and improved operational performance. ISAC provides simultaneous communication and sensing capabilities, enhancing the ability to detect, track, and respond to spectrum dynamics and potential threats in complex environments. In this paper, we introduce I-SCOUT, an innovative ISAC solution designed to uncover moving targets in NextG networks. We specifically repurpose the Positioning Reference Signal (PRS) of the 5G waveform, exploiting its distinctive autocorrelation characteristics for environment sensing. The reflected signals from moving targets are processed to estimate both the range and velocity of these targets using the cross ambiguity function (CAF). We conduct an in-depth analysis of the tradeoff between sensing and communication functionalities, focusing on the allocation of PRSs for ISAC purposes. Our study reveals that the number of PRSs dedicated to ISAC has a significant impact on the system's performance, necessitating a careful balance to optimize both sensing accuracy and communication efficiency. Our results demonstrate that I-SCOUT effectively leverages ISAC to accurately determine the range and velocity of moving targets. Moreover, I-SCOUT is capable of distinguishing between multiple targets within a group, showcasing its potential for complex scenarios. These findings underscore the viability of ISAC in enhancing the capabilities of NextG networks, for both commercial and tactical applications where precision and reliability are critical.
Title: Continual Deep Reinforcement Learning to Prevent Catastrophic Forgetting in Jamming Mitigation
Abstract: Deep Reinforcement Learning (DRL) has been highly effective in learning from and adapting to RF environments and thus detecting and mitigating jamming effects to facilitate reliable wireless communications. However, traditional DRL methods are susceptible to catastrophic forgetting (namely forgetting old tasks when learning new ones), especially in dynamic wireless environments where jammer patterns change over time. This paper considers an anti-jamming system and addresses the challenge of catastrophic forgetting in DRL applied to jammer detection and mitigation. First, we demonstrate the impact of catastrophic forgetting in DRL when applied to jammer detection and mitigation tasks, where the network forgets previously learned jammer patterns while adapting to new ones. This catastrophic interference undermines the effectiveness of the system, particularly in scenarios where the environment is non-stationary. We present a method that enables the network to retain knowledge of old jammer patterns while learning to handle new ones. Our approach substantially reduces catastrophic forgetting, allowing the anti-jamming system to learn new tasks without compromising its ability to perform previously learned tasks effectively. Furthermore, we introduce a systematic methodology for sequentially learning tasks in the anti-jamming framework. By leveraging continual DRL techniques based on PackNet, we achieve superior anti-jamming performance compared to standard DRL methods. Our proposed approach not only addresses catastrophic forgetting but also enhances the adaptability and robustness of the system in dynamic jamming environments. We demonstrate the efficacy of our method in preserving knowledge of past jammer patterns, learning new tasks efficiently, and achieving superior anti-jamming performance compared to traditional DRL approaches.
Serving as the IEEE MILCOM Unclassified Technical Chair and Track Chair for Networking Protocols
June 15, 2024
Dr. Sastry Kompella serves as the IEEE MILCOM (Military Communications Conference) Unclassified Technical Chair and Dr. Yalin Sagduyu serves as the IEEE MILCOM Track Chair for Networking Protocols and Performance.
Organizing the ACM Workshop on Machine Learning for NextG Networks at ACM MobiCom
June 1, 2024
Organizing the ACM Workshop on Machine Learning for NextG Networks in conjunction with ACM Conference on Mobile Computing and Networking (MobiCom), 2024.
Organized the ACM Workshop on Wireless Security and Machine Learning (WiseML)
May. 30, 2024
Dr. Yalin Sagduyu has co-organized the ACM Workshop on Wireless Security and Machine Learning (WiseML) in conjunction with the ACM Conference on Security and Privacy in Wireless and Mobile Networks (WiSec) 2024.
Served on the “Addressing DoD’s Spectrum Needs in the Age of Programmable Spectrum” Panel
May. 20, 2024
Dr. Tugba Erpek has served on the "Addressing DoD’s Spectrum Needs in the Age of Programmable Spectrum" panel at the Potomac Officers Club (POC) - 2024 5G Forum.
Served on the panel “Role of AI in 6G Open and Programmable RAN” at IEEE INFOCOM Workshop
May. 20, 2024
Dr. Yalin Sagduyu has served on the panel “Role of AI in 6G Open and Programmable RAN” at IEEE INFOCOM Workshop on Next Generation Open and Programmable RAN (NG-OPERA) on May 20, 2024.
Congratulations to Dr. Sastry Kompella as a Newly Elected 2024 IEEE Fellow
Dec. 21, 2023
We are delighted to share that Dr. Sastry Kompella has been elected as a 2024 IEEE Fellow for his contributions and leadership in advancing dynamic spectrum access and wireless communications. Congratulations to Sastry!!!
Honored to receive the Myron Hiram Nichols Telemetry Spectrum Award
Oct. 26, 2023
Dr. Sastry Kompella attended the International Telemetry Conference (ITC 2023) and honored to receive the Myron Hiram Nichols Telemetry Spectrum Award on "Towards Building a Common Operating Picture for Risk Informed Spectrum Management" . Congratulations!
Served on the "Ohio Innovation Tour" Panel
July 11, 2023
Dr. Sastry Kompella has been invited to participate in the "Ohio Innovation Tour: Artificial Intelligence Partnership with Industry, National Labs, & Talent and Workforce Development Strategies for Business" panel, on July 12th, 2023. He will describe the collaboration of Nexcepta with the NSF-funded AI-EDGE Institute at the Ohio State University as an industry partner of the institute.
Organized the IEEE INFOCOM Workshop on Age of Information
May 20, 2023
Dr. Sastry Kompella, Chief Scientist of Nexcepta, organized the 6th IEEE International Conference on Computer Communications (INFOCOM) Workshop on Age of Information. Dr. Kompella is a leader in the field of age of information that studies freshness of information and timeliness of communications with critical applications in cyber-physical systems, Internet of Things, and next-generation communication networks.
Selected as Army SBIR Awardee to develop AI/ML solutions
May 8, 2023
Nexcepta is proud to announce its recent selection through the Army Applied SBIR Program in its first AI/ML-focused open-topic solicitation. The U.S. Army seeks cutting-edge technologies to develop and deliver critical AI/ML and wearables solutions to Soldiers.
Awarded AF SBIR Contract on SDN-based Satellite Networking
May 1, 2023
Nexcepta was recently awarded an Air Force SBIR project for building a Software-Defined Networking (SDN)-enabled dynamic resource allocation for satellite networks.
Co-Authored Spectrum Best Practice Document
March 9, 2023
A new spectrum best practices document has been published by the National Spectrum Consortium (NSC) MIL-CIV Task Group and the U.S. Department of Defense (DoD). These recommendations represent a significant step in establishing metrics, criteria, procedures, and controls for sharing the spectrum in the 2025-2110 MHz band between the DoD and civilian entities. Our Chief Scientist Dr. Sastry Kompella contributed and co-authored the document as part of the MIL-CIV Task Group.
Watch the video at https://lnkd.in/ePgd3Y8j
NSC Membership
Dec. 8, 2022
We are excited to announce that Nexcepta has recently joined the National Spectrum Consortium (NSC). As a consortium member, we are working with the Government and other industry partners to develop and deliver enabling technologies related to the use of the electromagnetic spectrum and the information that rides on it.
Keynote Talk and Panel Moderator at MILCOM 2022
Dec. 2, 2022
Dr. Sastry Kompella, our Chief Scientist at Nexcepta, gave a keynote talk on “Spectrum meets AI at the Tactical Edge: Challenges and Architectural Considerations” at the Workshop on "The Internet of Things for Adversarial Environments". The talk discussed the complexity and pitfalls associated with developing AI/ML solutions for tactical wireless networks and communication systems including unmanned systems.
Dr. Sastry Kompella also served as the moderator for the technical panel on "Modernizing Enterprise Spectrum Management Within the DoD". The panel members included the Spectrum Access Research and Development Program (SAR&DP) Manager at OUSD, along with the SAR&DP Tranche 2 projects (OSCAR, RISA and MICCA) working towards creating a system of systems to demonstrate an automated spectrum management. The panel also included a technical director from DISA DSO who discussed emerging doctrine related to joint Electromagnetic Spectrum operations, planning and management capabilities and improved interoperability with related service and intelligence tools and systems.
Chairing a Workshop at MILCOM 2022
Dec. 2, 2022
Nexcepta team attended IEEE MILCOM 2022 this week in Rockville, MD. Our Chief Scientist Dr. Sastry Kompella successfully ran the workshop on Fundamental advances in Information Latency” as co-chair. The goal of this workshop was to showcase recent contributions from the MURI of the same name, on the impact of information freshness and Age of Information (AoI) on military IoT systems.
Partnered for Open6G Cooperative Research Center
Sept. 27, 2022
Nexcepta is pleased to join a select group of companies in partnering for the Institute of the Wireless Internet of Things (WIoT) at Northeastern University and participates in their newly created Open6G cooperative research center. Nexcepta will collaborate with center researchers in various 6G R&D activities.