Search

Objective: The information transfer rate (ITR) is widely accepted as a performance metric for generic brain-computer interface (BCI) spellers, while it is noticeable that the communication speed given by ITR is actually an upper bound which however can never be reached in real systems. A new performance metric is therefore needed. Methods: In this paper, a new metric named average time consumption per character (ATCPC) is proposed. It quantifies how long it takes on average to type one character using a typical synchronous BCI speller. To analytically derive ATCPC, the real typing process is modelled with a random walk on a graph. Misclassification and backspace are carefully characterized. A close-form formula of ATCPC is obtained through computing the hitting time of the random walk. The new metric is validated through simulated typing experiments and compared with ITR. Results: Firstly, the formula and simulation show a good consistency. Secondly, ITR always tends to overestimate the communication speed, while ATCPC is more realistic. Conclusion: The proposed ATCPC metric is valid. Significance: ATCPC is a qualified substitute for ITR. ATCPC also reveals the great potential of keyboard optimization to further enhance the performance of BCI spellers, which was hardly investigated before.

Brain-computer interfaces (BCIs) have been widely focused and extensively studied in recent years for their huge prospect of medical rehabilitation and commercial applications. Transfer learning exploits the information in the source domain and applies in another different but related domain (target domain), and is therefore introduced into the BCIs to figure out the inter-subject variances of electroencephalography (EEG) signals. In this article, a novel transfer learning method is proposed to preserve the Riemannian locality of data structure in both the source and target domains and simultaneously realize the joint distribution adaptation of both domains to enhance the effectiveness of transfer learning. Specifically, a Riemannian graph is first defined and constructed based on the Riemannian distance to represent the Riemannian geometry information. To simultaneously align the marginal and conditional distribution of source and target domains and preserve the Riemannian locality of data structure in both domains, the Riemannian graph is embedded in the joint distribution adaptation (JDA) framework and forms the proposed Riemannian locality preserving-based transfer learning (RLPTL). To validate the effect of the proposed method, it is compared with several existing methods on two open motor imagery datasets, and both multi-source domains (MSD) and single-source domains (SSD) experiments are considered. Experimental results show that the proposed method achieves the highest accuracies in MSD and SSD experiments on three datasets and outperforms eight baseline methods, which demonstrates that the proposed method creates a feasible and efficient way to realize transfer learning.

Non-orthogonal multiple access (NOMA) is a promising technology for the next generation of wireless networks. However, maximizing energy efficiency (EE) for NOMA systems may cause unbearable near-far unfairness. In this letter, a novel fairness-aware energy-efficient power allocation for uplink NOMA systems with imperfect successive interference cancellation (SIC) is proposed. We first introduce a new metric called geometric energy efficiency (GEE), then a dynamic power allocation scheme is developed by maximizing the GEE. Numerical results show that the proposed power allocation scheme achieves at least a 15.1% EE gain over benchmarks while also providing better or comparable fairness performance.

Non-orthogonal multiple access (NOMA) is one of candidate technologies in the next generation radio access networks, in which the joint maximum-likelihood (JML) detector has the optimal estimation performance. In this paper, we propose a rotated-constellation based method to regularize the decision boundaries of received constellation. Through regular decision boundaries, we further derive the approximate closed-form BER expressions of uplink NOMA systems in Rayleigh fading channels. Numerical results verify the effectiveness of rotated-constellation based method and conform to our derivation. It is also shown through simulations that the proposed analysis method outperforms conventional method in terms of accuracy.