With the advancement of Internet of Things (IoT) applications, it is essential to utilize an IoT platform to facilitate data exchange and application deployment. Existing platforms are typically either data-cloud-based or data-centralized, relying on servers as repeaters to exchange data. However, these architectures often face limitations related to triangle routing, network bottlenecks, and data scalability challenges, particularly in AIoT (Artificial Intelligence of Things) applications that require the fusion of numerous high-volume data streams. These challenges can be significantly mitigated through data-decentralized direct sender-to-receiver exchanges with a remote 'Agent', which is responsible for connectivity management. This work presents a prototype data-decentralized AIoT platform (DAIoTtalk) featuring peer-to-peer communications empowered by customized gRPC remote procedure calls based on the publish-subscribe (pub-sub) paradigm. An extension of IoTtalk, DAIoTtalk ensures device management with more adaptable node networking and offers a test bed for low-code development with decentralized communications. We demonstrate through extensive experiments that our design achieves at least 3 times more efficiency than a data-centralized approach. We also develop a case study to showcase the flexibility of our platform.

Video prediction has been studied in several earlier works. However, there are two inherent limitations in existing solutions, one being the ambiguity concern and the other being the degradation problem in lengthy predictions. To overcome these limitations, this work studies using wearable inertial sensors to guide video prediction. We create a data set called Pedestrians with IMU (Ped-IMU) that records people walking around with the wearable devices to collect the relation between inertial measurement unit (IMU) and video data and propose a SensePred model to conquer these limitations. The model takes full wearable sensor information and partial video information as inputs and predicts the missing video information. Simulation results validate the effectiveness of our model.

As natural habitats protection has become a global priority, smart sensing-nets are ever-increasingly needed for effective environmental observation. In a practical monitoring network, it is critical to deploy sensors with sufficient automated intelligence and motion flexibility. Recent advances in robotics and sensors technology have enabled automated mobile sensors deployment in a smart sensing-net. Existing deployment algorithms can be employed to calculate adequate destinations (goals) for sensors to perform respective monitoring tasks. However, given the calculated goal positions, the problem of how to actually coordinate a fleet of robots and schedule moving paths from random initials to reach their goals safely, without collisions, remains largely unaddressed in the wireless sensor networking (WSN) literature. In this paper, we investigate this problem and propose polynomial-time collision-free motion algorithms based on batched movements to ensure all the mobile sensors reach their goals successfully without incurring collisions. We observe that the grouping (batching) strategy is similar to the coloring procedure in graph theory. By constructing a conflict graph, we model the collision-free path scheduling as the well-known k-coloring problem, from which we reduce to our k-batching problem (determining the minimum number of required batches for a successful deployment) and prove its NP completeness. Since the k-batching problem is intractable, we develop CFMA (collision-free motion algorithm), a simple yet effective batching (coloring) heuristic mechanism, to approximate the optimal solution. Performance results show that our motion algorithms outperform other existing path-scheduling mechanisms by producing 100% sensors reachability (success probability of goals reaching), time-bounded deployment latency with low computation complexity, and reduced energy consumption. Note to Practitioners— This research was originally motivated by an oceanography project, which studied marine microbes by sending a team of tiny robots (sensors) randomly scattered on the ocean floor. For hard-to-access habitats like deserts or oceans, where manual placement of sensors is costly or impossible, automatically scheduling robots movements to calculated positions from random initials is essential for an effective monitoring. Our contribution is unique in two ways. First, traditional path-planning research focuses more on independent robots navigating the environment, whereas we target on the network automation problem, coordinating a fleet of robots working together to perform an environmental sensing task. Second, we observe that existing movement methods are too complicated and energy-consuming due to the calculation on-the-go nature. To provide a practical solution, we suggest and design our motion algorithms from a new perspective: pre-scheduled batched movements. Our approach requires a central server for grouping (batching) calculations, but can effectively reduce computation complexity and save significant energy expenditure exerted on resource-limited sensors. The proposed collision-free motion algorithm (CFMA) is a suboptimal yet efficient batching solution, which can be easily applied in real-life open-space monitoring networks. Insights from this foundational research will facilitate future opportunities to implement and verify our method in operational monitoring testbeds.

Multisensor data fusion combines various information sources to produce a more accurate or complete description of the environment. This article studies an object identification (OID) system using multiple distributed cameras and Internet-of-Things (IoT) devices for better visualizability and reconfigurability. We first propose a data processing and fusing method to merge the detection results of different IoT devices and video cameras, in order to locate, identify, and track target objects in the monitored area. Then, we develop the FusionTalk system by integrating the data fusion techniques with IoTtalk, an IoT device management platform. FusionTalk is designed with flexibility, modularity, and expansibility, where cameras, IoT devices, and network applications are modularized and can be conveniently plugged in/out, reconfigured, and reused through graphical user interfaces. In FusionTalk, the scope and the target of surveillance can be flexibly configured and associated, and administrators can be warned and easily visualize the movement and behavior of specific objects. Our experimental evaluation of the data fusion algorithm in various scenarios shows an identification accuracy above 95%. Finally, theoretical and numerical analyses on the failure probability of pairing IoT devices with video objects by FusionTalk are presented. Extensive experiments are performed to demonstrate the pairing effectiveness in real-world scenarios with failure probability less than 0.01%.

Both cameras and IoT devices have their particular capabilities in tracking moving objects. Their correlations are, however, unclear. In this work, we consider using a drone to track ground objects. We demonstrate how to retrieve IoT data from devices, which are attached on human objects, and correctly tag them on the human objects captured by a drone view. This is the first work correlating IoT data and computer vision from a drone camera. Potential applications of this work include aerial surveillance, people tracking, and intelligent human-drone interaction.

This paper presents a robot for people identification and tracking developed on robot operating system (ROS). It achieves modulized, light-weight, low-cost, and high-performance design goals even with the existence of distributed sensors. The key idea is to utilize wearable devices to enhance the people tracking capability of a robot through instant wireless communications and multi-sensory data fusion. Experimental results in a realistic environment demonstrate that our robot can keep tracking a specific person at a safe distance even without seeing the biological features of the person, who walks in a crowd with complex trajectories.

The 3GPP Narrowband Internet of Thing (NB-IoT) is the promising technology that can provide multiple types of resource unit (RU) with a special repetition mechanism to improve the scheduling flexibility and transmission reliability. Since the IoT devices need to operate for a very long time, the energy consumption becomes a critical issue. In this paper, we study how to guarantee the quality of service (QoS) while minimizing the energy consumption for IoT devices. We first model the problem and then propose an energy-efficient scheme, which consists of two stages. The first stage tries to incur the lowest energy consumption of devices and satisfy their QoS requirement. The second stage determines the scheduling order to ensure the delay constraint while maintaining energy efficiency. Simulation results show that our scheme can serve more devices while saving their energy.

For next generation mobile communications, Cloud-RAN (C-RAN) is an emerging network architecture to provide broadband services. C-RAN separates computation entities, i.e., Baseband Units (BBUs), from base stations (BSs) and puts BBUs in a cloud located in a centralized network. With C-RAN, UEs can receive data from multiple collaborative cells and thus can leverage the dynamic point selection (DPS) technology to improve network efficiency. When user equipments (UEs) enter a hotspot and can not be served due to congestion, data offloading from the hotspot to its neighboring cells may take place to balance heterogeneous cells' loads. This work shows how to integrate such DPS offloading with the Discontinuous Reception (DRX) mechanism, which allows UEs to turn off their radio interfaces in a periodical manner. We address the resource allocation problem in heterogeneously-loaded C-RAN by optimizing UEs' energy consumption based on DRX while reserving sufficient bandwidths for UEs considering their quality-of-service (QoS) through DPS. We propose an offloading-based DPS scheduling scheme by exploiting not only maximal instantaneous throughputs but also minimal energy cost. Simulation results show that our scheme can improve throughput, resource utilization, and energy consumption as compared to existing schemes.

Software defined network (SDN) provides flexible and scalable routing by separating control plane and data plane. With centralized control, SDN has been widely used in traffic engineering, link failure recovery, and load balancing. This work considers the flow update problem, where a set of flows need to be migrated or rearranged due to change of network status. During flow update, efficiency and consistency are two main challenges. Efficiency refers to how fast these updates are completed, while consistency refers to prevention of blackholes, loops, and network congestions during updates. This paper proposes a scheme that maintains all these properties. It works in four phases. The first phase partitions flows into shorter routing segments to increase update parallelism. The second phase generates a global dependency graph of these segments to be updated. The third phase conducts actual updates and then adjusts dependency graphs accordingly. The last phase deals with deadlocks, if any, and then loops back to phase three if necessary. Through simulations, we validate that our scheme not only ensures freedom of blackholes, loops, congestions, and deadlocks during flow updates, but is also faster than existing schemes.

For 5G wireless communications, the 3GPP Narrowband Internet of Things (NB-IoT) is one of the most promising technologies, which provides multiple types of resource unit (RU) with a special repetition mechanism to improve the scheduling flexibility and enhance the coverage and transmission reliability. Besides, NB‐IoT supports different operation modes to reuse the spectrum of LTE and GSM, which can make use of bandwidth more efficiently. The IoT application grows rapidly; however, those massive IoT devices need to operate for a very long time. Thus, the energy consumption becomes a critical issue. Therefore, NB‐IoT provides discontinuous reception operation to save devices’ energy. But, how to further reduce the transmission energy while ensuring the required ultra‐reliability is still an open issue. In this paper, we study how to guarantee the reliable communication and satisfy the quality of service (QoS) while minimizing the energy consumption for IoT devices. We first model the problem as an optimization problem and prove it to be NP‐complete. Then, we propose an energy‐efficient, ultra‐reliable, and low‐complexity scheme, which consists of two phases. The first phase tries to optimize the default transmit configurations of devices which incur the lowest energy consumption and satisfy the QoS requirement. The second phase leverages a weighting strategy to balance the emergency and inflexibility for determining the scheduling order to ensure the delay constraint while maintaining energy efficiency. Extensive simulation results show that our scheme can serve more devices with guaranteed QoS while saving their energy effectively.

Machine-to-machine (M2M) communication is one of the key technologies to realize Internet of Things (IoT). Since IoT applications are mainly for smart sensing, such as metering, home surveillance, disaster detection, and e-health, their special sensing/uploading behaviors will result in periodic and/or event-driven small data transmissions, which may potentially decrease the radio resource efficiency. On the other hand, the widespread deployment of IoT raises the concurrent massive connectivity of IoT devices. How to solve these two problems is a critical issue. In this paper, we investigate an uplink resource allocation problem which considers the periodic, event-driven, and query-based IoT traffic behaviors over LTE-M. The proposed approach takes advantage of data aggregation and both spatial and temporal reuse. Our solution exploits long-term static scheduling for periodic data to ensure the latency and data rate, and employs short-term dynamic scheduling for event-driven, query-based data to improve transmission efficiency. Therefore, both small data and massive connectivity problems are relieved. Extensive simulation results show that the proposed scheme can improve resource efficiency and enlarge network capacity effectively.

For next generation wireless communications, Cloud-RAN (C-RAN) has become an emerging network architecture of mobile communications, which separates computation entities-Baseband Units (BBUs) from original base stations (BSs), and puts BBUs in the cloud then forms a centralized network architecture. With C-RAN architecture, user equipments (UEs) can receive data from multiple collaborative cells and thus can leverage dynamic point selection (DPS) technology to improve network efficiency. Note that since the UEs are powered by batteries, energy saving is always a critical issue under C-RAN architecture. In current standard of 3GPP LTE-A, it has defined Discontinuous Reception (DRX) mechanism to allow UEs to turn off their radio interfaces and go to sleep to save energy. However, how to save UEs' energy under DPS in C-RAN is still an open issue. Therefore, this paper addresses the resource allocation problem by asking how to optimize the energy conservation of UEs based on DRX while serving UEs as more as possible under the consideration of UEs' quality of service (QoS) in C-RAN with DPS. To solve this problem, we propose an energy-efficient DPS (EE-DPS) scheduling scheme. The key idea of our scheme is to serve the UEs in the intersection of cells continuously and allocate resource tightly to avoid additional wake-up intervals. Extensive simulation results show that our scheme can serve most number of UEs while achieving high throughput as well as lower energy consumption, compared to the existing schemes.

Good deployment of sensors empowers the network with effective monitoring ability. Different from omnidirectional sensors, the coverage region of a directional sensor is determined by not only the sensing radius (distance), but also its sensing orientation and spread angle. Heterogeneous sensing distances and spread angles are likely to exist among directional sensors, to which we refer as heterogeneous directional sensors. In this paper, we target on a bounded monitoring area and deal with heterogeneous directional sensors equipped with locomotion and rotation facilities to enable the sensors self-deployment. Two Enhanced Deployment Algorithms, EDA-I and EDA-II, are proposed to achieve high sensing coverage ratio in the monitored field. EDA-I leverages the concept of virtual forces (for sensors movements) and virtual boundary torques (for sensors rotations), whereas EDA-II combines Voronoi diagram directed movements and boundary torques guided rotations. EDA-I computations can be centralized or distributed that differ in required energy and execution time, whereas EDA-II only allows centralized calculations. Our EDA-II outperforms EDA-I in centralized operations, while EDA-I can be adapted into a distributed deployment algorithm without requiring global information and still achieves comparably good coverage performance to its centralized version. To the best of our knowledge, this is perhaps the first work to employ movements followed by rotations for sensors self-deployment. Performance results demonstrate that our enhanced deployment mechanisms are capable of providing desirable surveillance level, while consuming moderate moving and rotating energy under reasonable execution time.

IoT (Internet of Things) and M2M (machine to machine) have attracted a lot of attention since more and more devices are expected to connect to the Internet for special purposes such as environment monitoring, home automation, industrial surveillance, and e-Health care. Currently, the OM2M (Open source platform for M2M communication) is a promising project which implements oneM2M and SmartM2M standards as an open-source platform for integrating various M2M services, applications, and devices. However, the individual data generated from those IoT/M2M devices is usually quite small, which incurs a lot of control overhead and thus decreases network performance significantly. Therefore, in this work we design and implement an OM2M 'plugin' that can aggregate small IoT data effectively. We will show how the plugin works and verify the effectiveness on the network bandwidth in this demonstration.

Smart video surveillance systems are essentials in modern environments to ensure safety and security for lives and property. In order to capture critical clues of reconnaissance, it is important to automatically and dynamically control the surveillance sensitivity and quality for detecting abnormal/suspicious events to perform high quality monitoring. In this work, we design a smart video surveillance system, which fully utilizes pan-tilt-zoom (PTZ) cameras and well integrates with environment sensors (such as fire/motion/door sensors) while cooperating with wearable devices that can adaptively monitor events and ensure the surveillance quality in terms of pixel-per-foot (PPF), viewing-angles, timeliness, and accuracy. In addition, this system also supports to define the special monitoring tasks and surveillance requirements, such as the surveillance timeliness and duty cycle for particular moving objects to be tracked per time unit. Then, a smart camera dispatch algorithm will calculate and determine the best cameras set with the corresponding configurations to accomplish the surveillance tasks effectively and precisely. We will show how our system realizes smart detecting and monitoring in this demonstration by leveraging PTZ cameras, environmental sensors and wearable devices.

Video surveillance systems are commonly used to monitor environments, such as factories, shopping malls, offices, and schools, for safety and security. However, traditional closed-circuit television cameras can only capture static scenes. When unexpected events happen, such as fire accidents or stranger intrusion, the recorded video data cannot provide immediate and precise information. In this work, we design a camera dispatch system for video surveillance with pan-tilt-zoom (PTZ) cameras, which can monitor event-areas/targets real-timely and flexibly. Users can dynamically define the monitoring requirements, such as pixel-per-foot and viewing-angle of the event-areas/targets. Then, a dispatch algorithm will decide the most necessary cameras and corresponding settings to monitor the event-areas/targets immediately. The control panel is implemented as a web-based service. The demo shows that users can easily monitor potential event-areas/targets anywhere, anytime through a web browser.

In modern wireless networks with multiple orthogonal (non-overlapping) channels available, one essential performance topic is how to effectively exploit channel diversity to enable parallel communications. Generally, having a radio interface hop through all available channels produces better spectrum diversity than binding it permanently to one channel, at the cost of channel switching delays and potentially compromised network connectivity. Moreover, multi-hop communications become challenging due to the lack of a common rendezvous for discovering routes and the difficulty of relaying packets from hop to hop. In this paper, we propose a multi-radio channel-hopping scheme (CHS) that preserves network connectivity. We prove that less than three radios are required by CHS in order to achieve good channel overlapping in widespread IEEE 802.11-based wireless systems. Corresponding channel-diverse routing (CDR) protocol is devised to realize efficient multi-hop communications. Simulation results demonstrate that the proposed CDR outperforms other strategies in static IEEE 802.11a multi-hop networking environments.

One critical issue, for a wireless sensor network (WSN) to operate successfully, is to provide sufficient sensing coverage. Define the smart sensing environment as a sensing system with the capability to sense the environment and respond properly in an automated manner. In this paper, we target on smart sensing environments and deal with heterogeneous sensors (here sensor heterogeneity is defined as sensors having different sensing ranges) equipped with actuation facilities to assist in the sensor self-deployment. A coverage-aware sensor automation (CASA) protocol is proposed to realize an automated smart monitoring network. Two centralized algorithms are included in the CASA protocol suite: enhanced virtual forces algorithm with boundary forces (EVFA-B) and sensor self-organizing algorithm (SSOA). Unlike most previous works that tackle the deployment problem only partially, we intend to address the problem from both global deployment (EVFA-B) and local repairing (SSOA) perspectives. The EVFA-B protocol exerts weighted attractive and repulsive forces on each sensor based on predefined distance thresholds. Resultant forces then guide the sensors to their suitable positions with the objective of enhancing the sensing coverage (after a possibly random placement of sensors). Furthermore, in the presence of sensor energy depletions and/or unexpected failures, our SSOA algorithm is activated to perform local repair by repositioning sensors around the sensing void (uncovered area). This capability of local recovery is advantageous in terms of saving the communication and moving energies. Performance of the proposed sensor deployment strategies is evaluated in terms of surveillance coverage, monitoring density, network self-healing competence, and moving energy consumption. We also implement our CASA protocol suite in a real-life monitoring network (MoNet) to demonstrate the protocol feasibility and validate the MoNet detection capability of emergency events.

IEEE 802.11 wireless network supports multiple link rates at the physical layer. Each link rate is associated with a certain required Signal-to-Interference-and-Noise Ratio (SINR) threshold for successfully decoding received packets. On transmission failures, the 802.11 DCF performs a binary exponential backoff mechanism to discourage channel access attempts, hoping to reduce congestion. When traditional link adaptation is applied, both rate reduction and binary backoff represent double penalties for this wireless link, which may cause overly conservative transmission attempts. On the other hand, once transmission succeeds, 802.11 DCF resets the backoff contention window to the minimum value to encourage channel access attempts. At the same time, traditional link adaptation may also decide to increase the data rate, which leads to overly aggressive transmission attempts. We observe this improper interaction of link rate and backoff mechanism that harms the 802.11 system performance, due to separate consideration of those two parameters. In this paper, we propose to jointly adapt the rate and backoff parameters. Specifically, an Enhanced Adaptation of link Rate and Contention window, abbreviated as EARC, is devised. EARC is a closed-loop (receiver-assisted) link rate adaptation protocol that jointly considers the backoff mechanism. With only one extra byte carried by the DATA packet, EARC incurs little controlling overhead despite its receiver-assisted nature. Moreover, since SINR information commonly utilized by receiver-assisted protocols is not precisely supported in real devices, we introduce a rate selection reference (RSR) table empirically derived by constantly monitoring the environmental energy level and reception behavior. The RSR table then guides the receiver to select the best sustainable rate for the transmitter. Simulation results demonstrate the RSR table is a practical option for making the rate decision, and the proposed EARC approach is effective in maintaining high system throughput, compared to other link adaptation algorithms.