Cybersickness remains a major barrier to the widespread adoption of virtual reality (VR) technologies, motivating researchers to investigate its causes and mitigation strategies through comparative human-subjects studies. These experiments may employ either within- or between-subjects designs. Although within-subjects designs require fewer participants and potentially offer higher statistical power, researchers need to wash out carryover effects of cybersickness symptoms before each session to avoid confounding the results. Prior studies have employed washout periods of varying lengths; some required testing conditions on different days, while others allowed only short breaks of less than 15 minutes. Although shorter washout periods are more convenient for experimenters, their impact on study outcomes has not been systematically investigated. In this work, we conducted a meta-analysis to evaluate the effects of study design and washout period length on cybersickness self-reports measured with the Simulator Sickness Questionnaire (SSQ). We found that short washout periods reduce statistical power compared to long washout periods or between-subjects designs. Based on these findings, we provide guidelines to help researchers design more reliable cybersickness studies and improve the generalizability of their results.

This paper presents Direction-Based Authentication (DBA), a novel authentication method for virtual reality that combines symbolic input and contextual information to balance efficiency with memorability. In DBA, the password consists of four view directions selected by the user across four virtual environments. The user can either physically turn their head or use buttons to select directions, and remember the directions by either symbol or visual context. We conducted a within-subjects study (N=32) to evaluate the efficiency, memorability, and security compared to methods based on symbols or context only. The results demonstrated that password entry with DBA was more efficient than the contextual-only approach. While recall rate was not significantly higher than the symbolic-only method, participants’ subjective ratings indicated that DBA better supported memorability. All three methods were highly resistant to observational attack, although the user-defined passwords appeared homogeneous in certain cases. Overall, the study shows that combining symbolic and contextual information is promising to balance efficiency with memorability for VR authentication, but potential usability and security issues can arise without careful consideration. Based on these findings, we discussed future directions for optimizing the usability and security of DBA and insights regarding participants’ reliance on symbolic vs. visual information.

Most existing works on resource allocation in cooperative OFDMA systems have focused on homogeneous users with same service and demand. In this paper, we focus on resource allocation in a service differentiated cooperative OFDMA system where each user has individual QoS requirement. We investigate joint power allocation, relay selection and subcarrier assignment to maximize overall system rates with considerations of QoS guarantees and service support. By introducing QoS price, this combinatorial problem with exponential complexity is converted into a convex one, and a two-level dual-primal decomposition based QoS-aware resource allocation (QARA) algorithm is proposed to tackle the problem. Simulation results reveal that our proposed algorithm significant outperforms previous works in terms of both services support and QoS satisfaction.

This paper considers joint schedule and resource allocation in downlink cooperative OFDMA systems with multiple sources, multiple relays and a single destination. Each user has individual QoS requirements and the relays operate in amplify-and-forward and half-duplex mode. We try to find optimal power allocation, relay selection and subcarrier assignment to maximize overall system rates. The problem is formulated as a Mixed Binary Integer Programming (MBIP). Although the original problem is comprehensive in nature, a novel two-level dual-primal decomposition based algorithm is proposed to tackle the problem. Optimal solutions are given in closed-form and the algorithm has a polynomial complexity in general. The efficiency of the algorithm is finally illustrated by numerical results.

Traditional congestion control protocols assume that each link provides a fixed capacity, while it is not always the case in wireless networks which have shared and variable medium. In this paper, we incorporate variable link capacity as a function of resource allocated, and random-access interferen