Target Sensing with Intelligent Reflecting Surface: Architecture and Performance

Intelligent reflecting surface (IRS) has emerged as a promising technology to reconfigure the radio propagation environment by dynamically controlling wireless signal's amplitude and/or phase via a large number of reflecting elements. In contrast to the vast literature on studying IRS's performance gains in wireless communications, we study in this paper a new application of IRS for sensing/localizing targets in wireless networks. Specifically, we propose a new self-sensing IRS architecture where the IRS controller is capable of transmitting probing signals that are not only directly reflected by the target (referred to as the direct echo link), but also consecutively reflected by the IRS and then the target (referred to as the IRS-reflected echo link). Moreover, dedicated sensors are installed at the IRS for receiving both the direct and IRS-reflected echo signals from the target, such that the IRS can sense the direction of its nearby target by applying a customized multiple signal classification (MUSIC) algorithm. However, since the angle estimation mean square error (MSE) by the MUSIC algorithm is intractable, we propose to optimize the IRS passive reflection for maximizing the average echo signals' total power at the IRS sensors and derive the resultant Cramer-Rao bound (CRB) of the angle estimation MSE. Last, numerical results are presented to show the effectiveness of the proposed new IRS sensing architecture and algorithm, as compared to other benchmark sensing systems/algorithms.

UnifiedSKG: Unifying and Multi-Tasking Structured Knowledge Grounding with Text-to-Text Language Models

Structured knowledge grounding (SKG) leverages structured knowledge to complete user requests, such as semantic parsing over databases and question answering over knowledge bases. Since the inputs and outputs of SKG tasks are heterogeneous, they have been studied separately by different communities, which limits systematic and compatible research on SKG. In this paper, we overcome this limitation by proposing the SKG framework, which unifies 21 SKG tasks into a text-to-text format, aiming to promote systematic SKG research, instead of being exclusive to a single task, domain, or dataset. We use UnifiedSKG to benchmark T5 with different sizes and show that T5, with simple modifications when necessary, achieves state-of-the-art performance on almost all of the 21 tasks. We further demonstrate that multi-task prefix-tuning improves the performance on most tasks, largely improving the overall performance. UnifiedSKG also facilitates the investigation of zero-shot and few-shot learning, and we show that T0, GPT-3, and Codex struggle in zero-shot and few-shot learning for SKG. We also use UnifiedSKG to conduct a series of controlled experiments on structured knowledge encoding variants across SKG tasks. UnifiedSKG is easily extensible to more tasks, and it is open-sourced at https://github.com/hkunlp/unifiedskg Latest collections at https://unifiedskg.com.

Causal Analysis Framework for Recommendation

Recently, recommendation based on causal inference has gained much attention in the industrial community. The introduction of causal techniques into recommender systems (RS) has brought great development to this field and has gradually become a trend. However, a unified causal analysis framework has not been established yet. On one hand, the existing causal methods in RS lack a clear causal and mathematical formalization on the scientific questions of interest. Many confusions need to be clarified: what exactly is being estimated, for what purpose, in which scenario, by which technique, and under what plausible assumptions. On the other hand, technically speaking, the existence of various biases is the main obstacle to drawing causal conclusions from observed data. Yet, formal definitions of the biases in RS are still not clear. Both of the limitations greatly hinder the development of RS. In this paper, we attempt to give a causal analysis framework to accommodate different scenarios in RS, thereby providing a principled and rigorous operational guideline for causal recommendation. We first propose a step-by-step guideline on how to clarify and investigate problems in RS using causal concepts. Then, we provide a new taxonomy and give a formal definition of various biases in RS from the perspective of violating what assumptions are adopted in standard causal analysis. Finally, we find that many problems in RS can be well formalized into a few scenarios using the proposed causal analysis framework.

Resource allocation for semantic-aware networks

Semantic communications have shown its great potential to improve the transmission reliability, especially in low signal-to-noise regime. However, the resource allocation for semantic-aware networks still remains unexplored, which is a critical issue in guaranteeing the transmission reliability of semantic symbols and the communication efficiency of users. To fill this gap, we investigate the spectral efficiency in the semantic domain and rethink the semantic-aware resource allocation issue. Specifically, the semantic spectral efficiency (S-SE) is defined for the first time, and is used to optimize resource allocation in semantic-aware networks in terms of channel assignment and the number of transmitted semantic symbols. Additionally, for fair comparison of semantic and conventional communication systems, a transform method is developed to convert the conventional bit-based spectral efficiency to the S-SE. Simulation results demonstrate the validity and feasibility of the proposed semantic-aware resource allocation model, as well as the superiority of semantic communications in terms of the S-SE.

Intelligent Reflecting Surface-Aided LEO Satellite Communication: Cooperative Passive Beamforming and Distributed Channel Estimation

We consider in this paper a new intelligent reflecting surface (IRS)-aided LEO satellite communication system, by utilizing the controllable phase shifts of massive passive reflecting elements to achieve flexible beamforming, which copes with the time-varying channel between the high-mobility satellite (SAT) and ground node (GN) cost-effectively. In particular, we propose a new architecture for IRS-aided LEO satellite communication where IRSs are deployed at both sides of the SAT and GN, and study their cooperative passive beamforming (CPB) design over line-of-sight (LoS)-dominant single-reflection and double-reflection channels. Specifically, we jointly optimize the active transmit/receive beamforming at the SAT/GN as well as the CPB at two-sided IRSs to maximize the overall channel gain from the SAT to each GN. Interestingly, we show that under LoS channel conditions, the high-dimensional SAT-GN channel can be decomposed into the outer product of two low-dimensional vectors. By exploiting the decomposed SAT-GN channel, we decouple the original beamforming optimization problem into two simpler subproblems corresponding to the SAT and GN sides, respectively, which are both solved in closed-form. Furthermore, we propose an efficient transmission protocol to conduct channel estimation and beam tracking, which only requires independent processing of the SAT and GN in a distributed manner, thus substantially reducing the implementation complexity. Simulation results validate the performance advantages of the proposed IRS-aided LEO satellite communication system with two-sided cooperative IRSs, as compared to various baseline schemes such as the conventional reflect-array and one-sided IRS.

Empowering Base Stations with Co-Site Intelligent Reflecting Surfaces: User Association, Channel Estimation and Reflection Optimization

Intelligent reflecting surface (IRS) has emerged as a promising technique to enhance wireless communication performance cost effectively. The existing literature has mainly considered IRS being deployed near user terminals to improve their performance. However, this approach may incur a high cost if IRSs need to be densely deployed in the network to cater to random user locations. To avoid such high deployment cost, in this paper we consider a new IRS aided wireless network architecture, where IRSs are deployed in the vicinity of each base station (BS) to assist in its communications with distributed users regardless of their locations. Besides significantly enhancing IRSs' signal coverage, this scheme helps reduce the IRS associated channel estimation overhead as compared to conventional user-side IRSs, by exploiting the nearly static BS-IRS channels over short distance. For this scheme, we propose a new two stage transmission protocol to achieve IRS channel estimation and reflection optimization for uplink data transmission efficiently. In addition, we propose effective methods for solving the user IRS association problem based on long term channel knowledge and the selected user IRS BS cascaded channel estimation problem. Finally, all IRSs' passive reflections are jointly optimized with the BS's multi-antenna receive combining to maximize the minimum achievable rate among all users for data transmission. Numerical results show that the proposed co site IRS empowered BS scheme can achieve significant performance gains over the conventional BS without co site IRS and existing schemes for IRS channel estimation and reflection optimization, thus enabling an appealing low cost and high performance BS design for future wireless networks.

Achievable Rate Maximization for Underlay Spectrum Sharing MIMO System with Intelligent Reflecting Surface

We consider the achievable rate maximization problem for intelligent reflecting surface (IRS) assisted multiple-input multiple-output systems in an underlay spectrum sharing scenario, subject to interference power constraints at primary users. The formulated non-convex optimization problem is challenging to solve due to its non-convexity as well as coupling design variables in the constraints. Different from existing works that are mostly based on alternating optimization (AO), we propose a penalty dual decomposition based gradient projection (PDDGP) algorithm to solve this problem. We also provide a convergence proof and a complexity analysis for the proposed algorithm. We benchmark the proposed algorithm against two known solutions, namely a minimum mean-square error based AO algorithm and an inner approximation method with block coordinate descent. Specifically, the complexity of the proposed algorithm is $O(N_I^2)$ while that of the two benchmark methods is $O(N_I^3)$, where $N_I$ is the number of IRS elements. Moreover, numerical results show that the proposed PDDGP algorithm yields considerably higher achievable rate than the benchmark solutions.

Rethinking Influence Functions of Neural Networks in the Over-parameterized Regime

Understanding the black-box prediction for neural networks is challenging. To achieve this, early studies have designed influence function (IF) to measure the effect of removing a single training point on neural networks. However, the classic implicit Hessian-vector product (IHVP) method for calculating IF is fragile, and theoretical analysis of IF in the context of neural networks is still lacking. To this end, we utilize the neural tangent kernel (NTK) theory to calculate IF for the neural network trained with regularized mean-square loss, and prove that the approximation error can be arbitrarily small when the width is sufficiently large for two-layer ReLU networks. We analyze the error bound for the classic IHVP method in the over-parameterized regime to understand when and why it fails or not. In detail, our theoretical analysis reveals that (1) the accuracy of IHVP depends on the regularization term, and is pretty low under weak regularization; (2) the accuracy of IHVP has a significant correlation with the probability density of corresponding training points. We further borrow the theory from NTK to understand the IFs better, including quantifying the complexity for influential samples and depicting the variation of IFs during the training dynamics. Numerical experiments on real-world data confirm our theoretical results and demonstrate our findings.

An Overview of Signal Processing Techniques for RIS/IRS-aided Wireless Systems

In the past as well as present wireless communication systems, the wireless propagation environment is regarded as an uncontrollable black box that impairs the received signal quality, and its negative impacts are compensated for by relying on the design of various sophisticated transmission/reception schemes. However, the improvements through applying such schemes operating at two endpoints (i.e., transmitter and receiver) only are limited even after five generations of wireless systems. Reconfigurable intelligent surface (RIS) or intelligent reflecting surface (IRS) have emerged as a new and revolutionary technology that can configure the wireless environment in a favorable manner by properly tuning the phase shifts of a large number of quasi passive and low-cost reflecting elements, thus standing out as a promising candidate technology for the next-/sixth-generation (6G) wireless system. However, to reap the performance benefits promised by RIS/IRS, efficient signal processing techniques are crucial, for a variety of purposes such as channel estimation, transmission design, radio localization, and so on. In this paper, we provide a comprehensive overview of recent advances on RIS/IRS-aided wireless systems from the signal processing perspective. We also highlight promising research directions that are worthy of investigation in the future.