This paper presents a class of routing protocols called road-based using vehicular traffic (RBVT) routing, which outperforms existing in city-based ad hoc networks (VANETs). RBVT leverage real-time information to create paths consisting successions road intersections that have, with high probability, network connectivity among them. Geographical forwarding is used transfer packets between on the path, reducing path's sensitivity individual node movements. For dense contention, we optimize...

Traffic congestion causes driver frustration and costs billions of dollars annually in lost time fuel consumption. This paper presents five traffic rerouting strategies designed to be incorporated a cost-effective easily deployable vehicular guidance system that reduces travel time. The proposed proactively compute individually tailored pushed vehicles when signs are observed on their route. the dynamic shortest path (DSP), A* with repulsion (AR*), random k (RkSP), entropy-balanced kSP...

Centralized solutions for vehicular traffic re-routing to alleviate congestion suffer from two intrinsic problems: scalability, as the central server has perform intensive computation and communication with vehicles in real-time; privacy, drivers have share their location well origins destinations of trips server. This article proposes DIVERT, a distributed system avoidance. DIVERT offloads large part at vehicles, thus, process becomes practical real-time. To take collaborative decisions,...

Ad hoc networks can be used not only as data carriers for mobile devices but also providers of a new class services specific to ubiquitous computing environments. Building in ad networks, however, is challenging due the rapidly changing operating contexts, which often lead situations where node hosting certain service becomes unsuitable execution any longer. We propose novel model provisioning based on concept context- aware migratory services. Unlike regular that executes always same node,...

Traffic congestion causes driver frustration and costs billions of dollars annually in lost time fuel consumption. This paper presents three traffic re-routing strategies designed to be incorporated a cost-effective easily deployable vehicular guidance system that reduces the effect congestions. collects real-time data from vehicles road-side sensors computes proactive, individually-tailored which is pushed when signs are observed on their route. Extensive simulation results over two urban...

Cellular traffic prediction enables operators to adapt demand in real-time for improving network resource utilization and user experience. To predict cellular traffic, previous studies either applied Recurrent Neural Networks (RNN) at individual base stations or adapted Convolutional (CNN) work grid-cells a geographically defined grid. These solutions do not consider explicitly the effect of handover on spatial characteristics which may lead lower accuracy. Furthermore, RNN are slow train,...

Federated Learning (FL) is a privacy-preserving distributed deep learning paradigm that involves substantial communication and computation effort, which problem for resource-constrained mobile IoT devices. Model pruning/sparsification develops sparse models could solve this problem, but existing sparsification solutions cannot satisfy at the same time requirements low bidirectional overhead between server clients, good model accuracy, under FL assumption does not have access to raw data...

TrafficView is a device that can be embedded in the next generation of vehicles to provide drivers with real-time view road traffic far beyond what they physically see. Vehicles equipped devices disseminate information using short-range wireless communication. The main benefits disseminating vehicle-to-vehicle fashion are scalability and ease deployment. paper describes prototype presents preliminary experimental results for this prototype.

With mobile devices becoming ubiquitous, the time is ripe to bring sensor data out of close-loop networks into center daily urban life. The Internet has become a great success because its applications appeal regular people. This isn't case with networks, which are generally perceived as "something" remote in forest or on battlefield. few exceptions, first-generation address application-specific, static-sensor deployments accurately monitor sensed environment real time.

High-accuracy and low-latency indoor place prediction for mobile users can enable a wide range of applications domains such as assisted living smart homes. In this paper, we propose GoPlaces, practical system that works on devices without requiring any new infrastructure. GoPlaces does not rely servers or specialized localization infrastructure, except single cheap off-the-shelf WiFi access point supports ranging with Round Trip Time (RTT) protocol. enables personalized naming prediction, it...

Spatial programming (SP) is a space-aware model for outdoor distributed embedded systems. Central to SP are the concepts of space and spatial reference, which provide applications with virtual resource naming in networks A network referenced using its expected physical location properties. Together other features, such as reference consistency access timeout, they help programmers cope highly dynamic configurations network-transparent fashion. We present design implementation smart messages,...

EZCab is a proof-of-concept ubiquitous computing application that allows people to book nearby cabs using their cell phones or PDAs equipped with short-range wireless network interfaces. discovers and books free mobile ad hoc networks of vehicles. We have implemented an prototype on top smart messages, middleware architecture based execution migration, which we had developed provide common environment for outdoor applications. The experimental simulation results demonstrated the feasibility

The next generation of computing systems will be embedded, in a virtually unbounded number, and dynamically connected. current software, network architectures, their associated programming models are not suitable for this scenario. This paper presents distributed model, Cooperative Computing, the Smart Messages architecture large networks embedded systems. In applications dynamic collections migratory execution units, called Messages, working to achieve common goal. Virtually any...

People-centric sensing with smart phones can be used for large scale of the physical world by leveraging cameras, microphones, GPSs, accelerometers, and other sensors on phones. Ranging from manual photo tasks to automated activity monitoring, any task crowd sourced sense data different locations at reduced cost. However, sensed submitted participants is not always reliable as they submit false earn money without executing actual task. Therefore, it important validate data. Validating...

Avatar is a system that leverages cloud resources to support fast, scalable, reliable, and energy efficient distributed computing over mobile devices. An avatar per-user software entity in the runs apps on behalf of user's The avatars are instantiated as virtual machines run same operating with In this way, provide resource isolation execute unmodified app components, which simplifies technology adoption. synchronized (mobile device, avatar) pairs achieve global goal. three main challenges...

Smart phones can collect and share Bluetooth co-location traces to identify ad hoc or semi-permanent social groups. This information, known group members but otherwise unavailable, be leveraged in applications protocols, such as recommender systems delay-tolerant forwarding networks, enhance the user experience. Group discovery using is practical because:(i) embedded nearly every phone has low battery consumption, (ii) short wireless transmission range lead good identification accuracy,...

Catastrophic disasters can destroy large regions and, in the process, leave many victims isolated from rest of world. Recovering communication infrastructure is typically slow and expensive, which not suitable for emergency response. Multihop wireless access networks have potential to quickly provide Internet connectivity victims, but so far no simple practical solution has been proposed help people configure these easily. We are pursuing approach utilizing virtualization techniques...

Recently, we started to experience a shift from physical communities virtual communities, which leads missed social opportunities in our daily routine. For instance, are not aware of neighbors with common interests or nearby events. Mobile computing applications (MSCAs) promise improve connectivity by leveraging information about people, relationships, and places. This paper presents MobiSoC, middleware that enables MSCAs development provides platform for capturing, managing, sharing the...

Applying Machine Learning (ML) to business applications for automation usually faces difficulties when integrating diverse ML dependencies and services, mainly because of the lack a common framework. In most cases, models are developed which targeted specific domain use leading duplicated effort, making reuse impossible. This paper presents Acumos, an open platform capable packaging into portable containerized microservices can be easily shared via platform's catalog, integrated various...

Traditional page recommendation models are endangered by stricter privacy regulations, such as the General Data Protection Regulation (GDPR). The performance of these suffer when only a part users share their personal data, cookies, with web servers, while rest choose to opt-out from sharing data. Furthermore, not designed provide recommendations for who do This paper addresses question how good all users, independent attitudes. We propose Fed4Rec, privacy-preserving framework based on...