In this paper, we address the issue of congestion in future Unmanned Aerial Vehicle (UAVs) traffic system in uncertain weather. We treat the traffic of UAVs as fluid queues, and introduce models for traffic dynamics at three basic traffic components: single link, tandem link, and merge link. The impact of weather uncertainty is captured as fluctuation of the saturation rate of fluid queue discharge (capacity). The uncertainty is assumed to follow a continuous-time Markov process. We define the resilience of the UAV traffic system as the long-run stability of the traffic queues and the optimal throughput strategy under uncertainties. We derive the necessary and sufficient conditions for the stabilities of the traffic queues in the three basic traffic components. Both conditions can be easily verified in practice. The optimal throughput can be calculated via the stability conditions. Our results offer strong insight and tool for designing flows in the UAV traffic system that is resilient against weather uncertainty.

Due to the rapid development of technologies for small unmanned aircraft systems (sUAS’s), the supply and demand market for sUAS’s is expanding globally. With the great number of sUAS’s ready to fly in civilian airspace, an sUAS aircraft traffic management system that can guarantee the safe and efficient operation of sUAS’s is still absent. In this paper, we propose a control protocol design and analysis method for sUAS traffic management (UTM) which can safely manage a large number of sUAS’s. The benefits of our approach are two-fold: at management level, the effort for monitoring sUAS traffic (authorities) and control/planning for each sUAS (operator/pilot) are both greatly reduced under our framework; and at operational level, the behavior of individual sUAS is guaranteed to follow the restrictions. Mathematical proofs and numerical simulations are presented to demonstrate the proposed method.

On time information is the basis of coordination and decision making for emergency responses during or after disasters. However, accurate information about disaster scenarios, which usually comes from the monitoring devices or people in disaster areas, is often unavailable since the communication network is congested or even damaged. The information centric network (ICN) paradigm could be employed in this situation to help reduce the download traffic that needs to go through the bottlenecks. However, the regular ICN does not handle frequent information uploading, which posts a new challenge for integrating ICN into the emergency communication network. In this paper, we propose to use edge caches as the foundation of caching systems that support information updating. There are two major design work involved: the local cache placement and across access networks synchronization of the local writing caches in the disaster area. For cache synchronization, we present three different schemes and analyze their system capacity in supporting update traffic from the system's perspective and individual local cache's perspective. For local cache placemen, we propose schemes with flexible scale up and scale down abilities, and present performance analysis on each scheme.

The United States is deploying the broadband wireless network for public safety and emergency response. While this will provide a foundational network when disasters or other emergencies appear, the network capacity could still be insufficient when large scale emergency events, such as earthquakes or flooding, happen. On-demand broadcast is a possible solution to address this situation, but its effectiveness remains unknown. In this paper, we aim to uncover the effectiveness of on-demand broadcast at different situations. Specifically, we study the scaling of average response time on the information requests with regard to the indicators of the content diversity - the number of distinct content files and the content popularity distribution. To achieve this goal, we first derive the lower bound of the average response time. We further investigate broadcast for streaming videos and uncover that parallel broadcast could have the same lower bound as the sequential broadcast under certain conditions. Based on the lower bound we derived, we evaluate the scaling of response time with regard to the number of files under different distributions. We further provide numerical results to demonstrate the accuracy of the approximation and the value of the lower bound on evaluating the optimality of a basic heuristic on-demand broadcast scheme.

Wide deployment of communication devices on vehicles is on the horizon due to the development of intelligent transportation systems. Although these communication devices are originally designed for highway efficiency and safety applications, the mobile communication ability on vehicles also enables a lot of other emerging applications. In this paper we study an application scenario on utilizing vehicles to carry messages for a set of sensor collectors, which are deployed at places without other communication infrastructure. A core problem that we study is on how to schedule messages so that the message delay is minimized. To achieve this goal, we propose three different strategies: the first feasible vehicle strategy, the optimal bound strategy, and the multiple attempt strategy. We further provide mathematical analysis on these strategies that are also verified by simulations, and compare the performance of these three strategies in simulations.We draw the conclusion that the optimal bound strategy is an effective strategy as it achieves low delay of the messages. At the same time it avoids the disadvantages such as high energy consumption and large storage size as in the multiple attempt strategy.

In this paper we investigate the scalability of communication architectures to provide traffic information services in a vehicular network based on the peer-to-peer (P2P) network technology. We study a general scenario in a metropolitan area where there are a huge number of data sources disseminating traffic data accessible to the vehicles through a multitude of roadside units. The large quantities of both data sources and customers post paramount challenges to the network scalability design. The existing work in P2P multicast, including the clustering based architecture, can only solve the scalability problem for applications with a limited number of data sources. We study the scenario that has a large number of data sources and prove that the clustering based architecture does not scale well with the traffic load at each node. Therefore, we propose a proxy based scalable architecture, in which data download traffic on each node is proved to keep constant even when the network size grows. We further verify the accuracy of the analysis using simulations.

A type of disruption/delay tolerant networks, known as pigeon networks, utilizes controllable special purpose vehicles called pigeons to convey messages among segregated areas. This paper aims to provide optimization methods for the challenging problem that how a pigeon should design its route to minimize the average delay of messages. For small scale networks, we construct an exact optimization algorithm that shows significant reduction in delay over the existing heuristic algorithms. For large scale networks, we present both the lower bound and upper bound of the average delay, and devise partitioning-based optimization algorithms that perform close to the lower bound. Both theoretical analysis and numerical results show that the delay could be reduced by as much as 50% compared to the existing algorithm.

Visual reranking has become a widely-accepted method to improve traditional text-based image search results. The main principle is to exploit the visual aggregation property of relevant images among top results so as to boost ranking scores of relevant images, by explicitly or implicitly detecting the confident relevant images, and propagating ranking scores among visually similar images. However, such a visual aggregation property does not always hold, and thus these schemes may fail. In this paper, we instead propose to filter out the most probable irrelevant images using deep contexts, which is the extra information that is not limited in the current search results. The deep contexts for each image consist of sets of images that are returned by searches using the queries formed by the textual context of this image. We compare the popularity of this image in the current search results and the deep contexts to check the irrelevance score. Then the irrelevance scores are propagated to the images whose useful textual context is missed. We formulate the two schemes together to reach a Markov random field, which is effectively solved by graph cuts. The key is that our scheme does not rely on the assumption that relevant images are visually aggregated among top results and is based on the observation that an outlier under the current query is likely to be more popular under some other query. After that, we perform graph reranking over filtered results to reorder them. Experimental results on the INRIA dataset show that our proposed method achieves significant improvements over previous approaches.

In this paper, we investigate a network coding-aware channel allocation and routing scheme for multi-hop cognitive radio networks. We consider network coding and channel availability in cognitive radio networks and maximize the throughput by allocating the channel and link rate appropriately. First, we model the activities of the primary users and the interference among the secondary users in a cognitive radio network and show how to implement network coding in the multi-hop cognitive radio network. Second, we formulate an optimization problem to maximize the throughput of the network. It takes advantage of the network coding opportunities and considers the channel availability constraint. By solving this optimization problem, we can determine how to allocate channels and rates of links in different channel availability scenarios. Furthermore, we compare the performance of our scheme with the coding oblivious routing for different scenarios of channel availability and maximum number of channels in a random wireless network. Our work brings insights on the benefits of network coding-aware routing in multi-hop cognitive radio networks.

In this paper, we propose a backbone construction scheme for multi-hop wireless cognitive radio networks (MWCRN) with a channel bonding technique. In our proposed scheme, the backbone is established purely in licensed bands using cognitive radio (CR) technology. We introduce a channel bonding technique to get higher performance of the network. To get higher reliability of MWCRN, we use backup channels in our proposed scheme. The proposed scheme includes two major steps. The first step is to formulate a backbone network; and the second step is to assign operating channels and backup channels to each backbone link. Simulation results show that the proposed scheme achieves higher network performance of multi-hop wireless cognitive radio networks.

In this paper, we study the communication architecture of smart grid for rural areas that employ cognitive radio technique. Since the communication in a cognitive radio network is normally unreliable, it is a great challenge to support applications which have stringent delay requirements, for instance, the transmission line monitoring application. To this end, we propose to reroute the data traffic with stringent delay requirement through the neighboring cells that work properly. This leads to a challenging scheduling problem, with the goal of maximizing the total throughput of the network while preserving the priority of real-time traffic in the local cell. To solve this problem, we present a scheduling algorithm that outperforms the earliest deadline scheduling and priority queue scheduling. The simulation results verify the effectiveness of our algorithm in achieving the design goals.

In a disruption/delay tolerant network, message carriers are often used to act as relays between segregated nodes and gateways. In this paper, we study a special routing problem for message carriers in delay tolerant networks, the visiting frequency problem. Specifically, we present a detailed analysis on a near-optimal visiting frequency on the nodes in a subregion of the network, to ensure low average message delay, using a partitioning-based optimization approach. Then we propose a multiple visit algorithm and apply the above near-optimal visiting frequency to our multiple visit algorithm, to guarantee the low average message delay for the network. Simulation results demonstrate the superiority of our new algorithm comparing with the existing single visit algorithms.

A special type of disruption/delay tolerant networks, known as pigeon networks, utilizes controllable vehicles called pigeons to convey messages among segregated hosts. This paper investigates the optimal assignment of multiple pigeons to serve hosts in a pigeon network such that the average message delay is minimized. To achieve this goal, first we provide an asymptotic analysis on the effectiveness of geographical partitioning-based method, which establishes the theoretical foundation for partitioning-based pigeon assignment. Second, we derive a near-optimal assignment of pigeons for a given partitioning through analysis. Finally, we design a pigeon assignment algorithm based on the theoretical analysis we have provided. Simulation results show that our algorithm can achieve lower delay compared to the existing schemes in delay tolerant networks.

In this paper, we investigate the scalability of three communication architectures for advanced metering infrastructure (AMI) in smart grid. AMI in smart grid is a typical cyber-physical system (CPS) example, in which large amount of data from hundreds of thousands of smart meters are collected and processed through an AMI communication infrastructure. Scalability is one of the most important issues for the AMI deployment in smart grid. In this study, we introduce a new performance metric, accumulated bandwidth-distance product (ABDP), to represent the total communication resource usages. For each distributed communication architecture, we formulate an optimization problem and obtain the solutions for minimizing the total cost of the system that considers both the ABDP and the deployment cost of the meter data management system (MDMS). The simulation results indicate the significant benefits of the distributed communication architectures over the traditional centralized one. More importantly, we analyze the scalability of the total cost of the communication system (including MDMS) with regard to the traffic load on the smart meters for both the centralized and the distributed communication architectures. Through the closed form expressions obtained in our analysis, we demonstrate that the total cost for the centralized architecture scales linearly as O(\lambda N), with N being the number of smart meters, and \lambda being the average traffic rate on a smart meter. In contrast, the total cost for the fully distributed communication architecture is O(\lambda^{2\over 3} N^{2\over 3} ), which is significantly lower.

When an emergency happens, public communication networks, including cellular networks, public switched telecommunication networks, and the Internet can be utilized to support emergency communications. The admission control strategies currently employed by the Wireless Priority Service (WPS), however, have drawbacks of being not able to address dynamic traffic demands for different scenarios and locations. They assume a standard profile for emergency and public demand at all locations. In contrast, this paper presents an analytical modeling framework that creates flexible, adaptive algorithms for operators to easily use based on the specifics of each scenario. Numerical results show that this provides an obvious improvement over the existing approaches. The analysis approach provided here, that incorporates the reneging behavior of customers waiting for service, also presents another main contribution compared with past work for computing the average waiting time.

Currently, wireless priority services (WPS) are provided using call queueing with a weighted round-robin schedule based on a fixed parameter. In this paper, we introduce a flexible, analytically-based scheme for dealing with admission control in a cellular network that supports emergency service. In the scheduling scheme, a priority queue is used to ensure the wise utilization of system resources while guaranteeing the priority of public handoff traffic. Also, probabilistic scheduling is considered for adaptively ensuring the high admission rate of emergency traffic and protecting public calls for different traffic demands. The admission of emergency traffic and public originating traffic can be flexibly adjusted. Numerical results about the evaluation of this strategy are presented and show an obvious improvement over related strategies.

In this paper we introduce the preemption threshold based strategy to cellular networks for supporting emergency traffic. With the new strategy, in addition to providing immediate access for emergency traffic, a certain amount of resources can be protected for public traffic by tuning the preemption threshold. Also, with the assumption that preempted users will restart the session after resumption, we find that a conservation law holds for the average duration of calls that succeed or fail to finish. Numerical results about the evaluation of this strategy are presented and show an obvious improvement over related strategies.

In this paper we present a single node decomposition based model for the analysis of wireless mesh networks. With knowledge of the network topology and routing strategy, the performance of each node, including throughput, delay, etc., can be analyzed using a simple M/G/1 model that takes into account the interference due to neighbor transmissions. With this basic analytic framework, closed form expressions for delay in terms of multipath routing variables are presented. A flow deviation optimization algorithm is used to derive the optimal flow over a given set of routes for both single and multiple classes of traffic. Numerical results are presented for different network topologies and compared with simulation studies.

Preemption is a possible way to ensure emergency traffic's priority in wireless network. But a pure preemption policy can be especially harsh to low priority calls. A way to improve this is introducing a queuing mechanism, which means putting preempted calls into a queue so that they can resume when channels become available, as long as they do not give up due to waiting too long in the queue. Also, we bring out a single preemption scheme, to avoid calls being preempted multiple times. The analytical results for performance analysis of these methods are given, and based on them the performance of different schemes are evaluated and compared