CREWS: Collaborative Robust Edge WiFi Sensing with Asynchronous and Incomplete Observations
This work provides a more robust and practical collaborative WiFi sensing system for edge deployments, which is important for applications requiring reliable sensing in volatile network environments.
This paper addresses the challenge of collaborative WiFi sensing in real-world edge deployments, where asynchronous and incomplete observations are common due to heterogeneous computing and network dropouts. The proposed CREWS framework restricts accuracy degradation to merely 2.2 percentage points under severe conditions (50% transient dropout rate or out-of-distribution jitter), outperforming state-of-the-art baselines.
Existing collaborative WiFi sensing systems rely on perfect node synchronization and complete data availability. However, real-world edge deployments suffer from heterogeneous computing and network dropouts, leading to asynchronous and incomplete features. We propose CREWS, a robust collaborative sensing framework that inherently resists these network volatility. First, CREWS employs a topology-agnostic aggregator invariant to the arrival order and subset size of incoming features. Second, rather than discarding delayed observations, it utilizes a staleness-aware adaptive replay mechanism. By treating stale features from lagging nodes as system-induced hard samples, CREWS transforms synchronization delays into beneficial training regularization. We theoretically prove the joint convergence of this architecture and demonstrate how replay bounds the bias-variance trade-off. Extensive evaluations and an 8-node heterogeneous hardware testbed demonstrate its superior resilience. Under severe conditions i.e., 50\% transient dropout rate or out-of-distribution jitter, CREWS restricts accuracy degradation to merely 2.2 percentage points, substantially outperforming state-of-the-art baselines.