Search for: All records

This work formulates a collaborative decision making framework that exploits over-the-air computation to efficiently aggregate soft information from distributed sensors. This new AirCompFDM protocol approximates the sufficient statistic (SS) of optimum binary hypothesis testing at a server node in this distributed sensing environment under different operation constraints. Leveraging pre/post-processing functions on over-the-air aggregation of sensor log-likelihood ratios, AirCompFDM significantly improves bandwidth efficiency with little detection loss, even from modest numbers of participating sensors and imperfect phase pre-compensation. Without phase pre-compensation, the benefit of over-the-air sensor aggregation diminishes but still can mitigate the effect of channel noise. Importantly, AirCompFDM outperforms the traditional bandwidth hungry polling scheme, even under low SNR. Furthermore, we analyze the Chernoff information and obtain the approximate effect of sensor aggregation on the probability of detection error that can help develop advanced detection strategies. 

We propose an over-the-air learning framework for collaborative decision making in wireless sensor networks. The low complexity framework leverages low-latency sensor transmission for a decision server to coordinate measurement sensors for hypothesis testing through over-the-air aggregation of sensor data over a multiple-access channel. We formulate several collaborative over-the-air hypothesis testing problems under different practical protocols for collaborative learning and decision making. We develop hypothesis tests for these network protocols and deployment scenarios including channel fading. We provide performance benchmark for both basic likelihood ratio test and generalized likelihood ratio test under different deployment conditions. Our results clearly demonstrate gain provided by increasing number of collaborative sensors. 

The prospect of massive deployment of devices for Internet-of-Things (IoT) motivates grant-free access for simultaneously uplink transmission by multiple nodes. Blind demixing represents a promising technique for recovering multiple such source signals over unknown channels. Recent studies show Wirtinger Flow (WF) algorithm can be effective in blind demixing. However, existing theoretical results on WF step size selection tend to be conservative and slow down convergence rates. To overcome this limitation, we propose an improved WF (WF-OPT) by optimizing its step size in each iteration and expediting the convergence. We provide a theoretical guarantee on the strict contraction of WF-OPT and present the upper bounds of the contraction ratio. Simulation results demonstrate the expected convergence gains. 

As applications of Internet-of-things (IoT) rapidly expand, unscheduled multiple user access with low latency and low cost communication is attracting growing more interests. To recover the multiple uplink signals without strict access control under dynamic co-channel interference environment, the problem of blind demixing emerges as an important obstacle for us to overcome. Without channel state information, successful blind demixing can recover multiple user signals more effectively by leveraging prior information on signal characteristics such as constellations and distribution. This work studies how forward error correction (FEC) codes in Galois Field can generate more effective blind demixing algorithms. We propose a constrained Wirtinger flow algorithm by defining a valid signal set based on FEC codewords. Specifically, targeting the popular polar codes for FEC of short IoT packets, we introduce signal projections within iterations of Wirtinger Flow based on FEC code information. Simulation results demonstrate stronger robustness of the proposed algorithm against noise and practical obstacles and also faster convergence rate compared to regular Wirtinger flow algorithm. 

We propose an over-the-air learning framework for collaborative decision making in wireless sensor networks. The low complexity framework leverages low-latency sensor transmission for a decision server to coordinate measurement sensors for hypothesis testing through over-the-air aggregation of sensor data over a multiple-access channel. We formulate several collaborative over-the-air hypothesis testing problems under different practical protocols for collaborative learning and decision making. We develop hypothesis tests for these network protocols and deployment scenarios including channel fading. We provide performance benchmark for both basic likelihood ratio test and generalized likelihood ratio test under different deployment conditions. Our results clearly demonstrate gain provided by increasing number of collaborative sensors. 

We propose a new nonconvex framework for blind multiple signal demixing and recovery. The proposed Riemann geometric approach extends the well known constant modulus algorithm to facilitate grant-free wireless access. For multiple signal demixing and recovery, we formulate the problem as non-convex problem optimization problem with signal orthogonality constraint in the form of Riemannian Orthogonal CMA(ROCMA). Unlike traditional stochastic gradient solutions that require large data samples, parameter tuning, and careful initialization, we leverage Riemannian geometry and transform the orthogonality requirement of recovered signals into a Riemannian manifold optimization. Our solution demonstrates full recovery of multiple access signals without large data sample size or special initialization with high probability of success.