Mobile Networks and Applications

There is a growing trend in developing applications for mobile computing systems in which mobile host computers retain their network connections while in transit. This paper proposes an algorithm that enforces a useful property, namely, causal ordering, that delivers messages among mobile hosts. This property ensures that causally related messages directed to the same destination will be delivered in an order consistent with their causality, which is important in applications that involve human interaction such as mobile e‐mail and mobile teleconferencing. Such applications are envisioned by the proponents of Personal Communications Services (PCS). Without this property, users may receive and read original messages and the corresponding replies out of order. Our algorithm, when compared with previous proposals, provides an alternative with a low handoff cost, medium message overhead, and low probability of unnecessary inhibition in delivering messages.

In this paper we discuss how we improved the MChannel group communication middleware for Mobile Ad-hoc Networks (MANETs) in order to let it become both delay- and energy-aware. MChannel makes use of the Optimized Link State Routing (OLSR) protocol, which is natively based on a simple hop-count metric for the route selection process. Based on such metric, OLSR exploits Dijkstra’s algorithm to find optimal paths across the network. We added a new module to MChannel, enabling unicast routing based on two alternative metrics, namely end-to-end delay and overall network lifetime. With such new module, we prove that network lifetime and average end-to-end delay improve, compared to the original OLSR protocol implementation included in the mentioned middleware. Thanks to MChannel’s approach, which implements routing in the user’s space, the improvements achieved in the unicast jOLSR routing protocol are transparently applied to the upstanding MChannel overlay multicast OMOLSR protocol. We also discuss how the proposed new module actually represents a general framework which can be used by programmers to introduce in MChannel novel metrics and path selection algorithms.

A smart meter neighborhood area network is usually regarded as the last mile network, which plays a significant role for communications in smart grid. A neighborhood area network typically consists of smart meters and Data Aggregation Points (DAPs), which collect energy consumption or billing information from smart meters and forward the information to wide area network gateways via wireless communications. The location of DAPs significantly affects the distance and associated transmission routes between DAPs and smart meters. In this paper, we investigate the DAP placement problem and propose solutions to reduce the distance between DAPs and smart meters. Specifically, the DAP placement problem is formulated with two objectives, e.g., the average distance minimization and the maximum distance minimization. The concept of network partition is introduced in this paper and two associated algorithms are developed to address the DAP placement problem. Extensive simulations are conducted based on a real suburban neighborhood topology. The simulation results verify that the proposed solutions are able to remarkably reduce the communication distance between DAPs and their associated smart meters.

Mobile IP is a standard for handling routing for hosts that have moved from their home network. This paper studies the costs of the Mobile IP handoff that occurs when a mobile host moves between networks. Experiments were carried out with Mobile IP and TCP over varying network conditions to observe the effect of handoffs on the transmission. This paper shows that although Mobile IP may be appropriate for current applications, its long handoff periods make it unsuitable for the future.

A MAC protocol for Ultra Wide Band (UWB) radio networks named (UWB)2 is proposed. The algorithm exploits typical features of impulse radio such as large processing gain, and is conceived in conjunction with a synchronization strategy which foresees the presence of a synchronization sequence in each transmitted packet. (UWB)2 adopts a pure Aloha approach; Performance analysis of the synchronization tracking mechanism showed in fact that under the preliminary simplistic hypothesis of an AWGN channel, and for a sufficient number of pulses in the synchronization sequence, a fairly high probability of successful synchronization can be achieved, even in the presence of several users and Multi User Interference (MUI). The multiple access scheme is based on the combination of a common control channel provided by a common Time Hopping (TH) code with dedicated data channels associated to transmitter specific TH codes. Results obtained by simulation indicate that (UWB)2 can be successfully applied when the number of users spans from a few tens to about one hundred, for data rates ranging from a few thousands to a few hundreds of bits per second. Network throughput was above 99.8% in all considered simulation settings. Such achievement confirms that (UWB)2 is a suitable and straightforward solution for large networks of terminals using impulse radio for transmission at low bit rates.

To implement various artificial intelligence and automation applications in smart factories, edge computing and industrial Internet of Things (IIoT) devices must be widely deployed, so as to increase the demand of coping with huge-scale and high-diversity data. Through deployment of fifth-generation (5G) networks (providing wide broadband, low latency, and massive machine type communications), industrial wireless networks, cloud, and fixed/mobile end devices in smart factories are interoperated in a harmony. However, with the huge-scale deployment of 5G networks and the IIoT in smart factories, threats and attacks against various vulnerabilities increase enormously, and cause considerable security and privacy challenges. Consequently, this article investigates crucial security and privacy issues for 5G-IIoT smart factories in three entities (i.e., physical layer, data layer and application layer), and further surveys recent approaches based on deep learning, reinforcement learning, and blockchain. In addition, this article provides future perspectives and challenges along this line of research.

Predicting trustworthiness of mobile services is a fundamental need for mobile service selection. With the popularization of mobile social networks, employing trust propagation to predict trust of a user placed on a mobile service becomes available. However, existing methods based on trust propagation in social networks may suffer from a scalability problem, i.e., their trust computation for two indirectly connected users is likely too time-consuming to be acceptable in very large social networks. To address this issue, this paper proposes a trust propagation method which exploits the peculiar properties of social networks and incorporates a landmark-based method with preprocessing to improve the efficiency of trust prediction. In this method, a small number of landmark users in the social network are firstly selected as referees in trust propagation, and the trust between these landmark users and the other users are then pre-computed. The trust between two indirectly connected users is finally estimated via aggregating the referrals provided by the landmark users. To evaluate the performance of the proposed method, comprehensive experiments are conducted using a real online social network. The experimental results show that our method is quite more efficient than the other four classic trust propagation methods in trust prediction.

VR sickness (Cybersickness) presents an important challenge in virtual reality environments. We present the results of a study on the effects of VR technology and VR video content type on VR sickness and on autonomous nervous system of the user. The participants watched two omnidirectional (360°) videos of different content types (neutral and action) on five distinct video display types (2D TV screen, three generations of Oculus Rift VR HMDs and on the mobile Samsung GearVR HMD). The Simulator Sickness Questionnaire (SSQ) in combination with the measurement of the physiological parameters (skin conductance and skin temperature, respiratory frequency and heart rate) were used to monitor the participants’ physiology. The results show that video content significantly affects the SSQ grading and the skin conductance level. VR sickness effects were significantly reported less often with TV display type than with other VR HMDs.

This study suggests a new product recommendation model to reflect the recent purchasing patterns of customers. There are many methods to measure the similarity between customers or products using one-way collaborative filtering. However, few studies have calculated the similarity of using both customer information and product information. Therefore, in this study, affinity variables that combine customer data with product data are created through a confusion matrix. Various derived variables are also generated to enhance the forecasting performance in enormous analysis data. In this study, various data mining classifiers such as the decision tree, neural network, support vector machine, random forest, and rotation forest are applied, and a sliding-window scheme is considered to construct the recommendation model.