Differences

This shows you the differences between two versions of the page.

--- mlwins [2023/08/03 00:15]
lingjialiu created
+++ mlwins [2023/08/03 07:48] (current)
lingjialiu
@@ Line 16: / Line 16: @@
   * Principal Investigator in the Leading Institution: Lingjia Liu (ECE at VT)
   * Principal Investigator in the Collaborative Institution: [[http://lizhongzheng.mit.edu/|Lizhong Zheng]] (EECS at MIT)
+  * Student Investigator at VT: Jiarui Xu
 ===== Abstract =====
 ###
-With the growing importance of wireless connectivity for social and economic interaction, there have been rising demands for greater spectrum use by both primary and secondary active radios, amidst the critical requirement of quiet spectrum for scientific exploration by receiver-only passive systems. To improve the overall spectrum utilization of the wireless ecosystem, this project develops a holistic Intelligent Dynamic spEctrum Access (IDEA) framework that can substantially enhance the spectrum utilization, energy-efficiency, and coexistence capability of spectrum sharing networks. In IDEA, enabling technical innovations across multiple disciplines are synergistically developed, including neuromorphic design of energy-efficient computing hardware at the device and circuit level, and artificial intelligence for spectrum sensing and dynamic access at the network level. The spectrum and interference management in IDEA conscientiously treats the coexistence constraints imposed by passive services, in support of scientific and societal returns from remote sensing investments. The outcomes of this research are expected to broadly impact next-generation wireless networks and Internet of Things applications with high traffic demands, such as autonomous driving, smart cities and remote sensing.\\
+Artificial intelligence (AI) is having a transformational effect in many industries, however, the
-\\
+application of AI and its associated machine learning (ML) tools within wireless networks, while promising,
-The goal of this project is to develop an intelligent dynamic spectrum access framework with unprecedented spectrum utilization efficiency and agility to support spectrum coexistence. The developed IDEA network platform supports heterogeneous devices from both primary and secondary active radios as well as passive radios. Key technical innovations are developed across the network to substantially enhance system-level spectrum utilization and active-passive radio coexistence. Specifically, analog/mixed-signal spiking neural network (SNN)-based neuromorphic computing hardware is designed to provide on-board intelligence at ultra-low power for resource-constrained secondary active radios. Model-free deep reinforcement learning is integrated with wireless domain knowledge and the SNN platform to accelerate learning-based spectrum access and coexistence. Advanced spectrum monitoring techniques are developed to quickly detect and characterize various signal emitters in both active and passive services. Finally, software and hardware testbeds are developed for system level evaluation and tradeoff optimization. The IDEA framework not only empowers efficient spectrum access in highly dynamic wireless environments, but also facilitates holistic system design and optimization across devices and circuits, sensing and communications, and networking.
+is still in its nascent stages. Some fundamental difficulties remain in applying learning-based data-driven
+techniques in communication problems: they generally cannot match the results from the more conventional
+model-based approaches in offering insights to understand entire families of communication problems. In
+general, both the learning-based data-driven approach and the model-based approach have some
+advantages, some success stories and some limitations. The objective of this project is to combine these
+two approaches for communication problems focusing on the physical layer of a wireless network. To be
+specific, we will identify novel methods to utilize the knowledge of the structure inherent in a wireless
+network for the learning process to come up new algorithms that can be tailored to specific yet critical
+applications, and hence have better efficiency and performance guarantees.
 ###
-===== Research Breakdown =====
+===== Research Thrusts =====
-**IDEA** covers the following research components:
+We choose to explore the use of learning-based techniques, especially deep neural networks (DNNs), in the physical layer problems of a wireless network. This is not because we expect to see some orders of magnitude performance improvement like we often see from high layer problems, but rather since the physical layer has simple models and clear performance benchmarks, which can provide good guidance when we make changes to the learning
-  * Analog/mixed-signal neuromorphic computing hardware: SNN-aided device designs including multiplexing neural encoding, computing-in-memory, and efficient training for resource constrained secondary radios to enable on-board intelligence at ultra-low power consumption and compact design areas;
+algorithm to reflect the domain knowledge. Three critical problem domains have been identified to combine
-  * Improving spectrum utilization and coexistence through learning: tailored integration of model-free DRL and domain knowledge of spectrum sharing network with improved sample efficiency to increase spectrum utilization in realistic scenarios, along with judiciously designed DSA actions for coexistence;
+domain knowledge with learning algorithms:
-  * Spectrum sensing through concise statistical modeling and learning: efficient spectrum sensing techniques that exploit the inherent structural information of statistics to accurately extract discriminative higher-order statistical features of various signal sources within a short sensing time;
-The following thrusts are organized to address the above mentioned research components:
+  * Thrust 1: Symbol detection for MIMO fading interference channel;
-  * Thrust 1: Energy-Efficient Spiking Neural Networks Design and Optimization;
+  * Thrust 2: Massive MIMO with low-resolution ADCs;
-  * Thrust 2: Accelerating Learning-based DRL to Improve Spectrum Utilization;
+  * Thrust 3: MIMO-OFDM waveform and non-linear radio frequency effects.
-  * Thrust 3: Spectrum Sensing and Interference Control for Active and Passive Radio Coexistence.

Lingjia Liu's Webpage

User Tools

Site Tools

Differences

Page Tools