Over-the-Air Fusion of Sparse Spatial Features for Integrated Sensing and Edge AI over Broadband Channels

TL;DR

This paper introduces Spatial AirFusion, a novel over-the-air fusion framework for integrated sensing and edge AI in 6G networks, leveraging spatial sparsity and multi-carrier aggregation to enhance sensing accuracy over wireless channels.

Contribution

It proposes a new over-the-air fusion method exploiting spatial sparsity and multi-carrier techniques, along with algorithms for optimal voxel-carrier pairing and power allocation.

Findings

Significantly reduces computation errors in sensing tasks.

Improves sensing accuracy compared to conventional methods.

Efficient algorithms for voxel-carrier pairing and power allocation.

Abstract

The 6G mobile networks feature two new usage scenarios -- distributed sensing and edge artificial intelligence (AI). Their natural integration, termed integrated sensing and edge AI (ISEA), promises to create a platform that enables intelligent environment perception for wide-ranging applications. A basic operation in ISEA is for a fusion center to acquire and fuse features of spatial sensing data distributed at many edge devices (known as agents), which is confronted by a communication bottleneck due to multiple access over hostile wireless channels. To address this issue, we propose a novel framework, called Spatial Over-the-Air Fusion (Spatial AirFusion), which exploits radio waveform superposition to aggregate spatially sparse features over the air and thereby enables simultaneous access. The framework supports simultaneous aggregation over multiple voxels, which partition the 3D sensing region, and across multiple subcarriers. It exploits both spatial feature sparsity with channel diversity to pair voxel-level aggregation tasks and subcarriers to maximize the minimum receive signal-to-noise ratio among voxels. Optimally solving the resultant mixed-integer problem of Voxel-Carrier Pairing and Power Allocation (VoCa-PPA) is a focus of this work. The proposed approach hinges on derivations of optimal power allocation as a closed-form function of voxel-carrier pairing and a useful property of VoCa-PPA that allows dramatic solution space reduction. Both a low-complexity greedy algorithm and an optimal tree-search algorithm are then designed for VoCa-PPA. The latter is accelerated with a customised compact search tree, node pruning and agent ordering. Extensive simulations using real datasets demonstrate that Spatial AirFusion significantly reduces computation errors and improves sensing accuracy compared with conventional over-the-air computation without awareness of spatial sparsity.