With the demand of high data rate and low latency in fifth generation (5G), deep neural network decoder (NND) has become a promising candidate due to its capability of one-shot decoding and parallel computing. In this paper, three types of NND, i.e., multi-layer perceptron (MLP), convolution neural network (CNN) and recurrent neural network (RNN), are proposed with the same parameter magnitude. The performance of these deep neural networks are evaluated through extensive simulation. Numerical results show that RNN has the best decoding performance, yet at the price of the highest computational overhead. Moreover, we find there exists a saturation length for each type of neural network, which is caused by their restricted learning abilities.

Wireless sensor networks (WSN) acts as the backbone of Internet of Things (IoT) technology. In WSN, field sensing and fusion are the most commonly seen problems, which involve collecting and processing of a huge volume of spatial samples in an unknown field to reconstruct the field or extract its features. One of the major concerns is how to reduce the communication overhead and data redundancy with prescribed fusion accuracy. In this paper, an integrated communication and computation framework based on meta-learning is proposed to enable adaptive field sensing and reconstruction. It consists of a stochasticgradient-descent (SGD) based base-learner used for the field model prediction aiming to minimize the average prediction error, and a reinforcement meta-learner aiming to optimize the sensing decision by simultaneously rewarding the error reduction with samples obtained so far and penalizing the corresponding communication cost. An adaptive sensing algorithm based on the above two-layer meta-learning framework is presented. It actively determines the next most informative sensing location, and thus considerably reduces the spatial samples and yields superior performance and robustness compared with conventional schemes. The convergence behavior of the proposed algorithm is also comprehensively analyzed and simulated. The results reveal that the proposed field sensing algorithm significantly improves the convergence rate.Index Terms-Learn to sense, meta-learning, wireless sensor networks, field sensing and reconstruction, stochastic gradient descent (SGD), reinforcement learning.

No abstract

In existing communication systems, the channel state information of each UE (user equipment) should be repeatedly estimated when it moves to a new position or another UE takes its place. The underlying ambient information, including the specific layout of potential reflectors, which provides more detailed information about all UEs' channel structures, has not been fully explored and exploited.In this paper, we rethink the mmWave channel estimation problem in a new and indirect way, i.e., instead of estimating the resultant composite channel response at each time and for any specific location, we first conduct the ambient perception exploiting the fascinating radar capability of a mmWave antenna array and then accomplish the location-based sparse channel reconstruction. In this way, the sparse channel for a quasi-static UE arriving at a specific location can be rapidly synthesized based on the perceived ambient information, thus greatly reducing the signalling overhead and online computational complexity. Based on the reconstructed mmWave channel, single-beam mmWave communication is designed and evaluated which shows excellent performance. Such an approach in fact integrates radar with communication, which may possibly open a new paradigm for future communication system design.

No abstract