Abstract:Large Language Models (LLMs) have achieved significant success in various natural language processing tasks, but the role of wireless networks in supporting LLMs has not been thoroughly explored. In this paper, we propose a wireless distributed Mixture of Experts (WDMoE) architecture to enable collaborative deployment of LLMs across edge servers at the base station (BS) and mobile devices in wireless networks. Specifically, we decompose the MoE layer in LLMs by placing the gating network and the preceding neural network layer at BS, while distributing the expert networks among the devices. This deployment leverages the parallel inference capabilities of expert networks on mobile devices, effectively utilizing the limited computing and caching resources of these devices. Accordingly, we develop a performance metric for WDMoE-based LLMs, which accounts for both model capability and latency. To minimize the latency while maintaining accuracy, we jointly optimize expert selection and bandwidth allocation based on the performance metric. Moreover, we build a hardware testbed using NVIDIA Jetson kits to validate the effectiveness of WDMoE. Both theoretical simulations and practical hardware experiments demonstrate that the proposed method can significantly reduce the latency without compromising LLM performance.
Abstract:In this paper, we propose a novel multi-task, multi-link relay semantic communications (MTML-RSC) scheme that enables the destination node to simultaneously perform image reconstruction and classification with one transmission from the source node. In the MTML-RSC scheme, the source node broadcasts a signal using semantic communications, and the relay node forwards the signal to the destination. We analyze the coupling relationship between the two tasks and the two links (source-to-relay and source-to-destination) and design a semantic-focused forward method for the relay node, where it selectively forwards only the semantics of the relevant class while ignoring others. At the destination, the node combines signals from both the source node and the relay node to perform classification, and then uses the classification result to assist in decoding the signal from the relay node for image reconstructing. Experimental results demonstrate that the proposed MTML-RSC scheme achieves significant performance gains, e.g., $1.73$ dB improvement in peak-signal-to-noise ratio (PSNR) for image reconstruction and increasing the accuracy from $64.89\%$ to $70.31\%$ for classification.
Abstract:Lightweight and efficient neural network models for deep joint source-channel coding (JSCC) are crucial for semantic communications. In this paper, we propose a novel JSCC architecture, named MambaJSCC, that achieves state-of-the-art performance with low computational and parameter overhead. MambaJSCC utilizes the visual state space model with channel adaptation (VSSM-CA) blocks as its backbone for transmitting images over wireless channels, where the VSSM-CA primarily consists of the generalized state space models (GSSM) and the zero-parameter, zero-computational channel adaptation method (CSI-ReST). We design the GSSM module, leveraging reversible matrix transformations to express generalized scan expanding operations, and theoretically prove that two GSSM modules can effectively capture global information. We discover that GSSM inherently possesses the ability to adapt to channels, a form of endogenous intelligence. Based on this, we design the CSI-ReST method, which injects channel state information (CSI) into the initial state of GSSM to utilize its native response, and into the residual state to mitigate CSI forgetting, enabling effective channel adaptation without introducing additional computational and parameter overhead. Experimental results show that MambaJSCC not only outperforms existing JSCC methods (e.g., SwinJSCC) across various scenarios but also significantly reduces parameter size, computational overhead, and inference delay.
Abstract:Intelligent reflecting surface (IRS) is expected to play a pivotal role in future wireless sensing networks owing to its potential for high-resolution and high-accuracy sensing. In this work, we investigate a multi-target direction-of-arrival (DoA) estimation problem in a semi-passive IRS-assisted sensing system, where IRS reflecting elements (REs) reflect signals from the base station to targets, and IRS sensing elements (SEs) estimate DoA based on echo signals reflected by the targets. {First of all, instead of solely relying on IRS SEs for DoA estimation as done in the existing literature, this work fully exploits the DoA information embedded in both IRS REs and SEs matrices via the atomic norm minimization (ANM) scheme. Subsequently, the Cram\'er-Rao bound for DoA estimation is derived, revealing an inverse proportionality to $MN^3+NM^3$ under the case of identity covariance matrix of the IRS measurement matrix and a single target, where $M$ and $N$ are the number of IRS SEs and REs, respectively. Finally, extensive numerical results substantiate the superior accuracy and resolution performance of the proposed ANM-based DoA estimation method over representative baselines.
Abstract:Since decades ago, multi-antenna has become a key enabling technology in the evolution of wireless communication systems. In contrast to conventional multi-antenna systems that contain antennas at fixed positions, position-flexible antenna systems have been proposed to fully utilize the spatial variation of wireless channels. In this paper, movable antenna (MA) systems are analyzed from channel measurement, modeling, position optimization to performance evaluation. First, a broadband channel measurement system with physical MAs is developed, for which the extremely high movable resolution reaches 0.02 mm. A practical two-ray model is constructed based on the channel measurement for a two-dimensional movable antenna system across 32$\times$32 planar port positions at 300 GHz. In light of the measurement results, spatial-correlated channel models for the two-dimensional MA system are proposed, which are statistically parameterized by the covariance matrix of measured channels. Finally, the signal-to-interference-and-noise ratio (SINR)-maximized position selection algorithm is proposed, which achieves 99% of the optimal performance. The performance of different MA systems in terms of spectral efficiency are evaluated and compared for both planar and linear MA systems. Extensive results demonstrate the advantage of MAs over fixed-position antennas in coping with the multi-path fading and improving the spectral efficiency by 10% in a 300 GHz measured channel.
Abstract:This paper investigates robust semantic communications over multiple-input multiple-output (MIMO) fading channels. Current semantic communications over MIMO channels mainly focus on channel adaptive encoding and decoding, which lacks exploration of signal distribution. To leverage the potential of signal distribution in signal space denoising, we develop a diffusion model over MIMO channels (DM-MIMO), a plugin module at the receiver side in conjunction with singular value decomposition (SVD) based precoding and equalization. Specifically, due to the significant variations in effective noise power over distinct sub-channels, we determine the effective sampling steps accordingly and devise a joint sampling algorithm. Utilizing a three-stage training algorithm, DM-MIMO learns the distribution of the encoded signal, which enables noise elimination over all sub-channels. Experimental results demonstrate that the DM-MIMO effectively reduces the mean square errors (MSE) of the equalized signal and the DM-MIMO semantic communication system (DM-MIMO-JSCC) outperforms the JSCC-based semantic communication system in image reconstruction.
Abstract:Degraded broadcast channels (DBC) are a typical multiuser communication scenario, Semantic communications over DBC still lack in-depth research. In this paper, we design a semantic communications approach based on multi-user semantic fusion for wireless image transmission over DBC. In the proposed method, the transmitter extracts semantic features for two users separately. It then effectively fuses these semantic features for broadcasting by leveraging semantic similarity. Unlike traditional allocation of time, power, or bandwidth, the semantic fusion scheme can dynamically control the weight of the semantic features of the two users to balance the performance between the two users. Considering the different channel state information (CSI) of both users over DBC, a DBC-Aware method is developed that embeds the CSI of both users into the joint source-channel coding encoder and fusion module to adapt to the channel. Experimental results show that the proposed system outperforms the traditional broadcasting schemes.
Abstract:Federated Multi-Modal Learning (FMML) is an emerging field that integrates information from different modalities in federated learning to improve the learning performance. In this letter, we develop a parameter scheduling scheme to improve personalized performance and communication efficiency in personalized FMML, considering the non-independent and nonidentically distributed (non-IID) data along with the modality heterogeneity. Specifically, a learning-based approach is utilized to obtain the aggregation coefficients for parameters of different modalities on distinct devices. Based on the aggregation coefficients and channel state, a subset of parameters is scheduled to be uploaded to a server for each modality. Experimental results show that the proposed algorithm can effectively improve the personalized performance of FMML.
Abstract:This paper studies an integrated sensing and communication (ISAC) system, where a multi-antenna base station transmits beamformed signals for joint downlink multi-user communication and radar sensing of an extended target (ET). By considering echo signals as reflections from valid elements on the ET contour, a set of novel Cram\'er-Rao bounds (CRBs) is derived for parameter estimation of the ET, including central range, direction, and orientation. The ISAC transmit beamforming design is then formulated as an optimization problem, aiming to minimize the CRB associated with radar sensing, while satisfying a minimum signal-to-interference-pulse-noise ratio requirement for each communication user, along with a 3-dB beam coverage constraint tailored for the ET. To solve this non-convex problem, we utilize semidefinite relaxation (SDR) and propose a rank-one solution extraction scheme for non-tight relaxation circumstances. To reduce the computation complexity, we further employ an efficient zero-forcing (ZF) based beamforming design, where the sensing task is performed in the null space of communication channels. Numerical results validate the effectiveness of the obtained CRB, revealing the diverse features of CRB for differently shaped ETs. The proposed SDR beamforming design outperforms benchmark designs with lower estimation error and CRB, while the ZF beamforming design greatly improves computation efficiency with minor sensing performance loss.
Abstract:Large Language Models (LLMs) have achieved significant success in various natural language processing tasks, but how wireless communications can support LLMs has not been extensively studied. In this paper, we propose a wireless distributed LLMs paradigm based on Mixture of Experts (MoE), named WDMoE, deploying LLMs collaboratively across edge servers of base station (BS) and mobile devices in the wireless communications system. Specifically, we decompose the MoE layer in LLMs by deploying the gating network and the preceding neural network layer at BS, while distributing the expert networks across the devices. This arrangement leverages the parallel capabilities of expert networks on distributed devices. Moreover, to overcome the instability of wireless communications, we design an expert selection policy by taking into account both the performance of the model and the end-to-end latency, which includes both transmission delay and inference delay. Evaluations conducted across various LLMs and multiple datasets demonstrate that WDMoE not only outperforms existing models, such as Llama 2 with 70 billion parameters, but also significantly reduces end-to-end latency.