Video telephony is becoming popular over smart-phones and tablets. Unlike the Desktop era, smartphone users are often `mobile' and this impacts how the video is processed and transmitted over the network. The significant increase in the motion content in such videos change the composition of video frames. Coupled with wireless packet losses, it often leads to poor quality of video received by the end user. In this work, we propose RVD, a framework for Reliable Video Delivery in mobile telephony by accounting for video object motion comprising foreground end-user motion and background scene changes in the network transmission of video. Multilayer perceptron (MLP) based non-linear regression model is used to analyze the impact of redundancy on received video quality under network variations and different degrees of video motion. RVD achieves 17-25% bandwidth savings for a target video quality, and 50-56% quality improvement over video-oblivious approaches.

We characterize the performance of both video and network layer properties of Skype, the most popular video telephony application. The performance in both mobile and stationary scenarios is investigated; considering network characteristics such as packet loss, propagation delay, available bandwidth and their effects on the perceptual video quality, measured using spatial and temporal no-reference video metrics. Based on 200+ live traces, we study the performance of this mobile video telephony application. We model video quality as a function of input network parameters and derive a feed-forward Artificial-Neural-Network that accurately predicts video quality given network conditions (0.0206 ≤ MSE ≤ 0.570). The accuracy of this model improves significantly by incorporating end-user mobility as an input to the model.

Collaborative spectrum sensing is a key technology in cognitive radio networks (CRNs). Although mobility is an inherent property of wireless networks, there has been no prior work studying the performance of collaborative spectrum sensing under attacks in mobile CRNs. Existing solutions based on user trust for secure collaborative spectrum sensing cannot be applied to mobile scenarios, since they do not consider the location diversity of the network, thus over penalize honest users who are at bad locations with severe path-loss. In this paper, we propose to use two trust parameters, location reliability and malicious intention (LRMI), to improve both malicious user detection and primary user detection in mobile CRNs under attack. Location reliability reflects path-loss characteristics of the wireless channel and malicious intention captures the true intention of secondary users, respectively. We propose a primary user detection method based on location reliability (LR) and a malicious user detection method based on LR and Dempster-Shafer (D-S) theory. Simulations show that mobility helps train location reliability and detect malicious users based on our methods. Our proposed detection mechanisms based on LRMI significantly outperforms existing solutions. In comparison to the existing solutions, we show an improvement of malicious user detection rate by 3 times and primary user detection rate by 20% at false alarm rate of 5%, respectively.

Collaborative spectrum sensing is a key technology in cognitive radio networks (CRNs). It is inaccurate if spectrum sensing nodes are malicious. Although mobility is an inherent property of wireless networks, there has been no prior work studying the detection of malicious users for collaborative spectrum sensing in mobile CRNs. Existing solutions based on user trust for secure collaborative spectrum sensing cannot be applied to mobile scenarios, since they do not consider the location diversity of the network, thus over penalize honest users who are at locations with severe pathloss. In this paper, we propose to use two trust parameters, Location Reliability and Malicious Intention (LRMI), to improve malicious and primary user detection in mobile CRNs under attacks. Location Reliability reflects pathloss characteristics of the wireless channel and Malicious Intention captures the true intention of secondary users, respectively. Simulations of our proposed detection mechanisms, LRMI, show that mobility helps train location reliability and detect malicious users. We show an improvement of malicious user detection rate by 3 times and primary user detection rate by 20% at false alarm rate of 5%, respectively.