Vincent K. N. Lau

Ying Cui, Vincent K. N. Lau, Yueping Wu

In this paper, we propose a two-timescale delay-optimal base station Discontinuous Transmission (BS-DTX) control and user scheduling for downlink coordinated MIMO systems with energy harvesting capability. To reduce the complexity and signaling overhead in practical systems, the BS-DTX control is adaptive to both the energy state information (ESI) and the data queue state information (QSI) over a longer timescale. The user scheduling is adaptive to the ESI, the QSI and the channel state information (CSI) over a shorter timescale. We show that the two-timescale delay-optimal control problem can be modeled as an infinite horizon average cost Partially Observed Markov Decision Problem (POMDP), which is well-known to be a difficult problem in general. By using sample-path analysis and exploiting specific problem structure, we first obtain some structural results on the optimal control policy and derive an equivalent Bellman equation with reduced state space. To reduce the complexity and facilitate distributed implementation, we obtain a delay-aware distributed solution with the BS-DTX control at the BS controller (BSC) and the user scheduling at each cluster manager (CM) using approximate dynamic programming and distributed stochastic learning. We show that the proposed distributed two-timescale algorithm converges almost surely. Furthermore, using queueing theory, stochastic geometry and optimization techniques, we derive sufficient conditions for the data queues to be stable in the coordinated MIMO network and discuss various design insights.

Junting Chen, Vincent K. N. Lau

Multi-user multi-input-multi-output (MU-MIMO) systems transmit data to multiple users simultaneously using the spatial degrees of freedom with user feedback channel state information (CSI). Most of the existing literatures on the reduced feedback user scheduling focus on the throughput performance and the user queueing delay is usually ignored. As the delay is very important for real-time applications, a low feedback queue-aware user scheduling algorithm is desired for the MU-MIMO system. This paper proposed a two-stage queue-aware user scheduling algorithm, which consists of a queue-aware mobile-driven feedback filtering stage and a SINR-based user scheduling stage, where the feedback filtering policy is obtained from the solution of an optimization problem. We evaluate the queueing performance of the proposed scheduling algorithm by using the sample path large deviation analysis. We show that the large deviation decay rate for the proposed algorithm is much larger than that of the CSI-only user scheduling algorithm. The numerical results also demonstrate that the proposed algorithm performs much better than the CSI-only algorithm requiring only a small amount of feedback.

Junting Chen, Vincent K. N. Lau

Max weighted queue (MWQ) control policy is a widely used cross-layer control policy that achieves queue stability and a reasonable delay performance. In most of the existing literature, it is assumed that optimal MWQ policy can be obtained instantaneously at every time slot. However, this assumption may be unrealistic in time varying wireless systems, especially when there is no closed-form MWQ solution and iterative algorithms have to be applied to obtain the optimal solution. This paper investigates the convergence behavior and the queue delay performance of the conventional MWQ iterations in which the channel state information (CSI) and queue state information (QSI) are changing in a similar timescale as the algorithm iterations. Our results are established by studying the stochastic stability of an equivalent virtual stochastic dynamic system (VSDS), and an extended Foster-Lyapunov criteria is applied for the stability analysis. We derive a closed form delay bound of the wireless network in terms of the CSI fading rate and the sensitivity of MWQ policy over CSI and QSI. Based on the equivalent VSDS, we propose a novel MWQ iterative algorithm with compensation to improve the tracking performance. We demonstrate that under some mild conditions, the proposed modified MWQ algorithm converges to the optimal MWQ control despite the time-varying CSI and QSI.

Songfu Cai, Vincent K. N. Lau

In this paper, we consider a MIMO networked control system with an energy harvesting sensor, where an unstable MIMO dynamic system is connected to a controller via a MIMO fading channel. We focus on the energy harvesting and MIMO precoding design at the sensor so as to stabilize the unstable MIMO dynamic plant subject to the energy availability constraint at the sensor. Using the Lyapunov optimization approach, we propose a closed-form dynamic energy harvesting and dynamic MIMO precoding solution, which has an event-driven control structure. Furthermore, the MIMO precoding solution is shown to have an eigenvalue water-filling structure, where the water level depends on the state estimation covariance, energy queue and the channel state, and the sea bed level depends on the state estimation covariance. The proposed scheme is also compared with various baselines and we show that significant performance gains can be achieved.

Vincent K. N. Lau, Chung Ha Koh

In this paper, we analyze the fundamental power-delay tradeoff in point-to-point OFDM systems under imperfect channel state information quality and non-ideal circuit power. We consider the dynamic back- pressure (DBP) algorithm, where the transmitter determines the rate and power control actions based on the instantaneous channel state information (CSIT) and the queue state information (QSI). We exploit a general fluid queue dynamics using a continuous time dynamic equation. Using the sample-path approach and renewal theory, we decompose the average delay in terms of multiple unfinished works along a sample path, and derive an upper bound on the average delay under the DBP power control, which is asymptotically accurate at small delay regime. We show that despite imperfect CSIT quality and non-ideal circuit power, the average power (P) of the DBP policy scales with delay (D) as P = O(Dexp(1/D)) at small delay regime. While the impacts of CSIT quality and circuit power appears as the coefficients of the scaling law, they may be significant in some operating regimes.

Chengyu Xia, Danny H. K. Tsang, Vincent K. N. Lau

With the growth of neural network size, model compression has attracted increasing interest in recent research. As one of the most common techniques, pruning has been studied for a long time. By exploiting the structured sparsity of the neural network, existing methods can prune neurons instead of individual weights. However, in most existing pruning methods, surviving neurons are randomly connected in the neural network without any structure, and the non-zero weights within each neuron are also randomly distributed. Such irregular sparse structure can cause very high control overhead and irregular memory access for the hardware and even increase the neural network computational complexity. In this paper, we propose a three-layer hierarchical prior to promote a more regular sparse structure during pruning. The proposed three-layer hierarchical prior can achieve per-neuron weight-level structured sparsity and neuron-level structured sparsity. We derive an efficient Turbo-variational Bayesian inferencing (Turbo-VBI) algorithm to solve the resulting model compression problem with the proposed prior. The proposed Turbo-VBI algorithm has low complexity and can support more general priors than existing model compression algorithms. Simulation results show that our proposed algorithm can promote a more regular structure in the pruned neural networks while achieving even better performance in terms of compression rate and inferencing accuracy compared with the baselines.

Fan Zhang, Vincent K. N. Lau, Ling Shi

In this paper, we consider a networked control system (NCS) in which an dynamic plant system is connected to a controller via a temporally correlated wireless fading channel. We focus on communication power design at the sensor to minimize a weighted average state estimation error at the remote controller subject to an average transmit power constraint of the sensor. The power control optimization problem is formulated as an infinite horizon average cost Markov decision process (MDP). We propose a novel continuous-time perturbation approach and derive an asymptotically optimal closed-form value function for the MDP. Under this approximation, we propose a low complexity dynamic power control solution which has an event- driven control structure. We also establish technical conditions for asymptotic optimality, and sufficient conditions for NCS stability under the proposed scheme.

Chengyu Xia, Danny H. K. Tsang, Vincent K. N. Lau

In this paper, we investigate Bayesian model compression in federated learning (FL) to construct sparse models that can achieve both communication and computation efficiencies. We propose a decentralized Turbo variational Bayesian inference (D-Turbo-VBI) FL framework where we firstly propose a hierarchical sparse prior to promote a clustered sparse structure in the weight matrix. Then, by carefully integrating message passing and VBI with a decentralized turbo framework, we propose the D-Turbo-VBI algorithm which can (i) reduce both upstream and downstream communication overhead during federated training, and (ii) reduce the computational complexity during local inference. Additionally, we establish the convergence property for thr proposed D-Turbo-VBI algorithm. Simulation results show the significant gain of our proposed algorithm over the baselines in reducing communication overhead during federated training and computational complexity of final model.

Zezhong Zhang, Guangxu Zhu, Rui Wang et al.

Recently years, the attempts on distilling mobile data into useful knowledge has been led to the deployment of machine learning algorithms at the network edge. Principal component analysis (PCA) is a classic technique for extracting the linear structure of a dataset, which is useful for feature extraction and data compression. In this work, we propose the deployment of distributed PCA over a multi-access channel based on the algorithm of stochastic gradient descent to learn the dominant feature space of a distributed dataset at multiple devices. Over-the-air aggregation is adopted to reduce the multi-access latency, giving the name over-the-air PCA. The novelty of this design lies in exploiting channel noise to accelerate the descent in the region around each saddle point encountered by gradient descent, thereby increasing the convergence speed of over-the-air PCA. The idea is materialized by proposing a power-control scheme which detects the type of descent region and controlling the level of channel noise accordingly. The scheme is proved to achieve a faster convergence rate than in the case without power control.

Fan Zhang, Vincent K. N. Lau

In this paper, we consider dynamic precoder/decorrelator optimization for multimedia streaming in MIMO interference networks. We propose a truly cross-layer framework in the sense that the optimization objective is the application level performance metrics for multimedia streaming, namely the playback interruption and buffer overflow probabilities. The optimization variables are the MIMO precoders/decorrelators at the transmitters and the receivers, which are adaptive to both the instantaneous channel condition and the playback queue length. The problem is a challenging multi-dimensional stochastic optimization problem and brute-force solution has exponential complexity. By exploiting the underlying timescale separation and special structure in the problem, we derive a closed-form approximation of the value function based on continuous time perturbation. Using this approximation, we propose a low complexity dynamic MIMO precoder/decorrelator control algorithm by solving an equivalent weighted MMSE problem. We also establish the technical conditions for asymptotic optimality of the low complexity control algorithm. Finally, the proposed scheme is compared with various baselines through simulations and it is shown that significant performance gain can be achieved.