Search for: All records

Channel state information (CSI) plays a vital role in scheduling and capacity-approaching transmission optimization of massive MIMO communication systems. In frequency division duplex (FDD) MIMO systems, forward link CSI reconstruction at transmitter relies on CSI feedback from receiving nodes and must carefully weigh the tradeoff between reconstruction accuracy and feedback bandwidth. Recent application of recurrent neural networks (RNN) has demonstrated promising results of massive MIMO CSI feedback compression. However, the cost of computation and memory associated with RNN deep learning remains high. In this work, we exploit channel temporal coherence to improve learning accuracy and feedback efficiency. Leveraging a Markovian model, we develop a deep convolutional neural network (CNN)-based framework called MarkovNet to efficiently encode CSI feedback to improve accuracy and efficiency. We explore important physical insights including spherical normalization of input data and deep learning network optimizations in feedback compression. We demonstrate that MarkovNet provides a substantial performance improvement and computational complexity reduction over the RNN-based work.We demonstrate MarkovNet’s performance under different MIMO configurations and for a range of feedback intervals and rates. CSI recovery with MarkovNet outperforms RNN-based CSI estimation with only a fraction of computational cost. 

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. 

In this work, we analyze the convergence of constant modulus algorithm (CMA) in blindly recovering multiple signals to facilitate grant-free wireless access. The CMA typically solves a non-convex problem by utilizing stochastic gradient descent. The iterative convergence of CMA can be affected by additive channel noise and finite number of samples, which is a problem not fully investigated previously. We point out the strong similarity between CMA and the Wirtinger Flow (WF) algorithm originally proposed for Phase retrieval. In light of the convergence proof of WF under limited data samples, we adopt the WF algorithm to implement CMA-based blind signal recovery. We generalize the convergence analysis of WF in the context of CMA-based blind signal recovery. Numerical simulation results also corroborate the analysis. 

Edge machine learning can deliver low-latency and private artificial intelligent (AI) services for mobile devices by leveraging computation and storage resources at the network edge. This paper presents an energy-efficient edge processing framework to execute deep learning inference tasks at the edge computing nodes whose wireless connections to mobile devices are prone to channel uncertainties. Aimed at minimizing the sum of computation and transmission power consumption with probabilistic quality-of-service (QoS) constraints, we formulate a joint inference tasking and downlink beamforming problem that is characterized by a group sparse objective function. We provide a statistical learning based robust optimization approach to approximate the highly intractable probabilistic-QoS constraints by nonconvex quadratic constraints, which are further reformulated as matrix inequalities with a rank-one constraint via matrix lifting. We design a reweighted power minimization approach by iteratively reweighted ℓ1 minimization with difference-of-convex-functions (DC) regularization and updating weights, where the reweighted approach is adopted for enhancing group sparsity whereas the DC regularization is designed for inducing rank-one solutions. Numerical results demonstrate that the proposed approach outperforms other state-of-the-art approaches.