Recent advancements in wireless technology have spurred the growth of indoor localization applications based on wireless signals. Most existing methods model the relationship between input data and the target's location using the$l_{2}$loss function, which assumes that the residual between the predicted location and the ground truth follows a Gaussian distribution with a fixed standard deviation. Nevertheless, this approach may not conform to the actual distribution and lacks confidence measures for individual predictions, potentially compromising accuracy. Furthermore, in the realm of regression, Semi-Supervised Learning (SSL) remains relatively unexplored due to the absence of a reliable method to quantify prediction uncertainty, especially when labeled data is scarce. To address these challenges, we developed a novel indoor localization algorithm that employs probabilistic distribution modeling. Our approach focuses on indoor localization with Channel Impulse Response (CIR) as the input. Crucially, it leverages Maximum Likelihood Estimation (MLE) to model the localization error as a Gaussian distribution, and utilizes Residual Log-likelihood Estimation (RLE) to capture arbitrary error distributions. This enables us to extract the confidence of each prediction and utilize the most reliable ones as pseudo-labels for the unlabeled data. By employing probabilistic distribution modeling, we observed a significant improvement in localization accuracy over the$l_{2}$loss function. Additionally, by integrating pseudo-labeled data for model retraining, our algorithm achieves superior performance compared to existing state-of-the-art machine learning-based and SSL-based methods. This is demonstrated by our evaluation on two public datasets, showcasing the efficiency of the proposed method.

With the continuous enhancement of the computing power of Internet-of-Things (IoT) hardware and the development of lightweight AI algorithms, a new trend of IoT development has emerged, featuring the decentralization of intelligence and the differentiated user needs. Traditional centralized data processing and unified models are inapplicable to intelligent IoT devices considering privacy protection and personalized service requirements. In this paper, we propose a Decentralized Opportunistic Learning Architecture (DOLA) that enables device-to-device opportunistic collaboration based on semantic consensus. This framework effectively enhances the flexibility of device collaboration and meets the personalized needs of minority groups. Experiments conducted on three public datasets demonstrate that our framework can improve the average recognition accuracy by more than 1.46%.

Driven by new business, data and technology, the Internet-of-Things(IoT) has gradually evolved from a local Internet-of-Things and a wide-area Internet-of-Things to a prospective global Intelligent Internet-of-Things(IIoT). This article analyzes the typical business evolution requirements such as AR/VR, smart factories and intelligent transportation and then extracts the 6S characteristics of the Intelligent Internet-of-Things. As the development unfolds in such a trend, the traditional fixed-mode architecture of the Internet-of-Things and the idea of centralized management in the cloud are no longer suitable for the future development of the Internet-of-Things. Therefore, we first propose the MOBI (Modular Orchestration based on Business Intent) architecture, then analyze the core functional modules of the architecture, finally present the challenges that need to be resolved in the process of realizing the architecture. The purpose of this article is to provide a basis and reference for the industry to study the evolution of the Intelligent Internet-of-Things, so as to stimulate the rapid iterative evolution of the industry.

In Internet of Things (IoT), each object requires a unique identifier to identify itself and index its detailed profile to support mutual recognitions among multiple objects. However, existing IoT identifiers belonging to different identification schemes are heterogeneous from each other, which create a great challenge for the applications that need to resolve the heterogeneous identifiers. To address this challenge, we propose an algorithm to automatically recognize the heterogeneous identification schemes used by various IoT identifiers, based on a sequence-to-sequence (seq2seq) model consisting of an encoder and a decoder. The encoder uses one Long Short-Term Memory (LSTM) to map the identifier sequence to a vector of fixed dimensionality, and the decoder uses another LSTM to unfold the vector into a target sequence representing the identification scheme of this identifier. To evaluate our algorithm, we create a new dataset named ID-20 with 20 categories of IoT identifiers and conduct experiments on it. The results demonstrate the superiority of our algorithm against other state-of-the-art methods, with an identifier recognition accuracy of up to 94.57%.

Edge computing is a bridge for realizing the convergence between physical space and cyber space in the Internet of Things (IoT) paradigm. Large numbers of physical objects produce a huge amount of data that needs to be efficiently processed in the edge side. This situation urgently requires novel ideas and framework in the design and management of edge computing to improve and enhance its performance. In this article, we propose an approach and principle of software-defined edge computing (SDEC) from the perspective of cyber-physical mapping, where the ultimate goal is to achieve a highly automatic and intelligent edge computing system. The SDEC can also help realize flexible management and intelligent collaboration among various edge hardware resources and services by way of software. To this end, we design an SDEC-based open IoT system architecture which decouples upper level IoT applications from the underlying physical edge resources and builds dynamically reconfigurable smart edge services. The software-definition mechanism of the SDEC platform is proposed to introduce the detailed processes that the underlying physical devices are defined in the form of software. We also describe an illustrative application case about smart factory to present the practical effectiveness of the proposed scheme. Finally, we outline several challenges which are worthy of in-depth study and research. The SDEC paradigm can share, reuse, recombine, and reconfigure edge resources and services so that the overall service capability of the edge side can be improved.

Words in a natural language corpus obey a Zipfian distribution, which means only a minority percentage of the total are frequent but the vast majority are rare. Word embedding techniques heavily rely on the frequencies of words in the corpus and cannot provide reliable representations for words appearing infrequently during training. To address this problem, we propose a novel algorithm to induce embeddings for rare words by leveraging morphological decomposition, stemming and bidirectional translation. Compared to existing approaches, our algorithm maintains a relatively lightweight model but generates qualified representations for a wider range of vocabulary from the same corpus. We have performed evaluations on multiple general domain datasets and specific domain datasets, and the experimental results show that our algorithm achieves better performance than the other state-of-the-art techniques.

Since wearable sensors often have high substitutability, it is imperative that newly added wearable sensors have automatic training capabilities. In this paper, we propose a novel automatic model training method called Synchronous Automatic Training (SAT) for wearable sensors, which adopts a knowledge distillation technique to fully exploit the knowledge of the existing trained sensors and a synchronous mechanism to ensure the knowledge is transferred to the untrained sensors to assist the automatic model construction. The automatically trained model by using our proposed SAT method has higher accuracy than existing methods.

As the Internet of Things (IoT) expands its reach into virtually every domain, high-speed data processing and shorter response time are becoming more necessary than ever. Embedding machine learning algorithms in IoT devices is a possible effective solution. However, the IoT devices are typically deployed in highly dynamic and uncontrolled environments, which makes machine learning algorithms with automatic learning mechanisms imperative. In this paper, we introduce a framework for synchronous automatic learning. By defining the concept of the centroid distance vector we propose synchronous automatic learning framework which automatically accomplish the reconfiguration of machine learning algorithm based on label refinement. Our evaluation shows that the automatic reconfiguration of machine learning algorithms is achievable with higher accuracy compared to the state-of-the-art methods.

Intelligent interactions of IoT mean that devices can perform proper actions and communicate with each other to support specific application scenarios without human intervention. To do this, the biggest challenge is providing IoT objects with common knowledge and abilities to analyze data and to reason. In this paper, we present a semantic mechanism to realize intelligent interactions of IoT devices on account of the semantic knowledge and machine learning technology. Based on this mechanism, a three-layer IoT system named IoT-Book is built, and every action performed by a device is deduced through analyzing the environment data and personalized preferences of the user. Finally, a smart home use case is demonstrated, and our approach is proved to be effective.

In recent years, there have been significant research efforts related to vehicle to vehicle (V2V) communication. V2V communication was originally targeted for safety-related applications but interest in other possible application areas including communication support for traffic management has increased. This paper describes a system architecture that can provide high throughput and rapid data dissemination in vehicular networks. While vehicle mobility results in frequent routing changes, we expect that in a highway scenario vehicles will maintain relative positions and speeds for some time resulting in a quasi-stationary network configuration. The proposed approach system is intended to be robust and work with a very simple protocol specification taking advantage of the quasi-stationary behavior of vehicle networks. It also addresses some of the major concerns facing multi-hop vehicular communication. We previously proposed an architectural framework for achieving a vehicle network that employs dynamically switched directional links [30]. In this article we provide preliminary simulations results that highlight the scalability, reliability and latency benefits provided by the switched directional links for a given baseline vehicle configuration emulating a quasi-stationary network. A significant improvement in metrics such as PDR and latency is observed.

Abstract In this paper, we survey recent results in vehicular ad hoc networks (VANETs) data dissemination. We describe methods proposed to enforce dissemination scope such as geocast/broadcast and multicast. A growing category consisting of methods designed to achieve disruption tolerance in vehicular networks is presented. We describe the key ideas of representative technologies in each category. In addition, we consider location service and security issues that are crucial for data dissemination in VANET. We conclude by sharing our thoughts on further challenges. Copyright © 2009 John Wiley & Sons, Ltd.

Welcome to this special issue of Wiley Journal of Wireless Communications and Mobile Computing (WCMC).To date, wireless vehicular communications have attached much attention for improving road safety, intelligent management and data exchange services, and providing Internet access on the move to ensure wireless pervasive and ubiquitous connectivity.The field of wireless vehicular communications can be typically identified as vehicle-to-person communications, vehicle-to-infrastructure communications, vehicle-to-vehicle communications, and vehicular communication networks.The combination of unique features of wireless vehicular communications and networking issues opens new opportunities for many interesting research areas, for example, real time safety applications, and intelligent diver information services.The goal of this special issue is to present good-performance, highly scalable, robust, and secure vehicular technologies for universal realization of wireless vehicular communications.To guarantee the quality, in this special issue, we have selectively collected 15 papers out of over 60 submissions, and clustered them in four groups: three are survey papers, four papers deal with channel modeling and physical layer issues, six papers investigating MAC and network layer issues, and two papers focusing on applications.Detailed overview of the selected works is given below.The first group includes three papers, which survey recent advances in wireless vehicular communications.The first paper, by Chen et al., provides recent results for data dissemination in vehicular ad hoc networks (VANETs).This paper also discusses security challenges in this field and the need of supporting technologies to enable efficient data dissemination for automotive applications.The paper by Leng et al. gives a global review of media access control techniques for VANETs.This paper summarizes recent work in this domain including MAC standardization activities, wireless channel resource management, QoS capability enhancement in MAC layer, and reliable MAC broadcasting strategies.The third paper, by Martinez et al., makes a survey of several publicly available mobility generators, network simulators, and VANET simulators.This paper contrasts their software characteristics, graphical user interface, popularity, ease of use, input requirements, output visualization capability, and accuracy of simulation.

Vehicle-to-Vehicle (V2V) communications can be a promising means for providing active safety in driving. Broadcasting can be an essential technology for the dissemination of safety information from a vehicle to vehicles and a roadside unit (RSU) to vehicles. In this paper we propose a reliably suppressed broadcasting protocol for V2V communications. The proposed protocol effectively selects forwarding nodes in consideration of reliability in data delivery, efficiency in forwarding overhead, and delay, such that dissemination for safety information can be achieved with a high delivery ratio and small number of forwarding nodes. The proposed protocol does not require position information and neighboring information. With our suppression technique used in the proposed protocol, we enhance an existing position-based broadcasting protocol. We evaluate the performances of our proposed protocol and the enhanced position-based protocol by comparing them with other protocols through simulations. We also implement interoperability between the proposed protocol and the enhanced position-based protocol to support broadcasting on a vehicle ad-hoc network comprised of both vehicles with GPS and vehicles without GPS.

In this paper, we outline a framework for achieving high data rate connectivity in a multi-hop vehicular network. The data streams are sent to the receiver on the highest data rate path from the source vehicles. The method does not rely upon a location service for obtaining position information. Moreover, as the source and the receivers move from one position to the other, the paths are recomputed dynamically without any explicit position information notification. The forwarding method adapts rapidly, and in a distributed manner, as the conditions on the roads change due to vehicle mobility, radio propagation etc. Based on performance evaluations, we observe that the proposed protocol converges rapidly towards the forwarding method that ensures highest achieved data rates.

This paper addresses the design of dynamic local peer group (LPG) organizations and protocols for ad hoc peer-to-peer networking among neighboring vehicles to support vehicle communications. We first outline the dynamic LPG framework to organize vehicles that have frequently-changing neighbors and have no inherent relationships with one another. We then describe two dynamic LPG organization schemes and protocols to form and maintain the dynamic LPG of vehicles

Inter-vehicle communications is a topic of growing interest, with a number of target applications under consideration. One significant application for inter-vehicle communications is dissemination of vehicle safety information such as road and traffic-related events and conditions. To support vehicle safety information, a reliable and efficient inter-vehicle communications system which can meet stringent safety application performance requirements is needed. This paper reviews developments in routing schemes targeted for ad hoc vehicle networks and considers them in the context of safety communications

Load balancing for distributed servers is a common issue in many applications and has been extensively studied. Several distributed load balancing schemes have been proposed that proactively route individual requests to appropriate servers to best balance the load and shorten request response time. These schemes do not require a centralized load balancer. Instead, each server is responsible for determining, for each request it receives from a client, to which server in the pool the request should be forwarded for processing. We propose a new request routing scheme that is more scalable to increasing number of servers and request load than the existing schemes. The method combines random server selection and next-neighbor load sharing techniques that together prevent the staleness of load information from building up when the number of servers increases. Our simulation shows that it outperforms existing schemes under a piggyback-based load update model.

This paper proposes a restoration mechanism for resource-constrained wireless networks. More specifically, our contribution in this paper is two-fold. First, we propose a dynamic policy-based multi-layer self-healing mechanism that taps the merits of the various protocol layers to provide self-healing (service survivability) with partial network resources in the event of random network failures. In particular, we investigate the use of "higher protocol layers", such as layer 4, to help provide self healing capability. Next, we illustrate the power of modeling and simulation (M&S) to provide insights into the performance of complex aspects of wireless networks, since the complexity of the self-healing mechanisms coupled with the stochastic nature of the underlying network renders the use of analytic techniques infeasible unless otherwise simplifying assumptions are made. The latter, however, distort the system and even lead to misleading conclusions. Via M&S, we both demonstrate the working and analyze the performance of the proposed multi-layer restoration mechanism.

In mobile wireless environments, bandwidth fluctuations are unpredictable, and the throughput experienced by mobile user applications is dynamic. Real-time tracking of bandwidth variations is desirable to effectively manage the throughput available to a user application at any given time. We propose a throughput control mechanism for applicability over wireless networks, transparent to the underlying transport and network technologies. To improve the utilization of the available throughput, we develop a two-component application-layer, real-time control technique for mobile users traversing diverse networks with varying service rates. We first calculate the weighted average throughput experienced by the user's application. Secondly, we appropriately adjust the download rate to the user based on the real-time average throughput information. In this paper, we detail the throughput control algorithm and provide some initial simulation results.

In order to ensure proper quality of service for real-time communication in a mobile wireless Internet environment it is essential to minimize the transient packet loss when the mobile is moving between different cells (subnets) within a domain. Network layer mobility management schemes have been proposed to provide optimized fast-handoff for multimedia streams during a client's frequent movement within a domain. This paper introduces application layer techniques to achieve fast-handoff for real-time RTP/UDP based multimedia traffic in a SIP signaling environment. These techniques are based on standard SIP components such as user agent and proxy which usually participate to set up and tear down the multimedia sessions between the mobiles. Unlike network layer techniques, application layer techniques do not have to depend upon any additional components such as home agent and foreign agent. It thus provides a network access independent solution suitable for application service providers.

The wireless information era has created a phenomenal dependence amongst commercial and military applications on wireless networks. Such an increasing reliance in turn implies increasingly serious consequences caused by outages. Hence (i) alarm models that can help efficiently and quickly perform root cause analysis (RCA) and (ii) self-healing mechanisms that can respond to the dynamics of the underlying network and provide rapid recovery of the affected applications, are critical. We recognize the importance of the above and propose appropriate service survivability mechanisms. More specifically, our contributions are as follows. We derive alarm model specifications that can succinctly and accurately capture the network alarms and symptoms for processing by the inferencing (RCA) engine. Next, we propose a dynamic cost and performance-sensitive self-healing mechanism that can adapt to the networks' dynamics and provide rapid and efficient self-healing. The proposed mechanism has the additional feature of being able to be specified in terms of policies, thereby allowing for an easy integration into the rapidly emerging policy-based management framework. The results of our work will benefit both network designers and managers and help (a) not only in increasing customer retention ratios, but also providing increased revenues in the commercial setting and (b) provide the critically needed quick and seamless restoration of military services.

As multicast services become prevalent, it is important to find viable solutions for multicasting to mobile nodes. This problem is complicated by the necessity to support multicast services over existing backbone and access networks which may have varying network and/or link layer multicasting capabilities. While most work on supporting multicast services focuses on the IP layer solution, we propose an application-layer approach for providing multicast services to mobile users traversing networks with diverse multicast capabilities. We propose placing multicast proxies in the backbone and access networks to support several multicast-related functions at the application layer including the creation of virtual networks for dynamically tunneling through non-multicast-capable networks. We describe our proposed application-layer multicast architecture and its advantages to third-party service providers for multicasting to mobile users in diverse networks.

A mechanism for the positioning of mobile user location is presented in this paper. Based on the three-point problem from land surveying, this positioning mechanism makes use of network delay to estimate user locations by trigonometry and analytical geometry. Location-sensitive wireless services and applications can be provided without the need of obtaining access to the physical layer information such as signal strength, time of arrival (TOA), angle of arrival (AOA) or adding GPS (Global Positioning System) hardware. It also provides better protection for user privacy by a user initiation mechanism. The novel concept of virtual location based on network latency is introduced to investigate the relative topologies among mobile users and position known servers.

We investigate radio control unit (RCU) and access switch resource allocation in a wireless ATM (asynchronous transfer mode) access network. Since user mobility is an important factor in wireless ATM networks, we propose a framework that allows the examination of user mobility impact on cell-level quality of service (QoS) and resource allocation in conjunction with mobility management functions including handoffs and call-level functions. The proposed analysis may be employed in a network design tool for studying the admission control policy and selecting system parameters to satisfy desired cell-level QoS.

We investigate the connection admission control problem in mobile personal communications networks. In contrast to other studies on this aspect which either focused on only single-class call connections or ignored user mobility, our study directly addresses the connection admission control for multiple-class calls with user mobility. A generic class of coordinate-convex admission control policies is considered. An efficient numerical method is proposed to derive the connection-level quality of service (QoS) and is verified with computer simulation results. The proposed analysis may be employed in a network design tool for studying the admission control policy and selecting system parameters to satisfy the required connection-level QoS.

In this paper, we investigate the connection admission control problem in mobile personal communications networks. Since user mobility and multimedia traffic are important factors in personal communications networks, in contrast to other studies on this aspect which either focused on only single-class call connections or ignored user mobility, our study directly addresses the connection admission control for multiple-class calls with user mobility. A generic class of coordinate-convex admission control policies is considered. An efficient numerical method is proposed to derive the connection-level quality of service (QoS), and is verified with computer simulation results. The results obtained show that, besides offered load, user mobility can have a great impact on the connection-level QoS. The proposed analysis may be employed in a network design tool for studying the admission control policy and for selecting system parameters to satisfy the required connection-level QoS.

In a personal communication services (PCS) network, a set of channels is assigned to every cell. When a phone call arrives, it consumes a channel until the end of the conversation. If no channel is available, the call is dropped. In many cases, channels may return shortly after a call is dropped. Thus, if some buffering mechanism is introduced to the channel allocation algorithm, a cell may accommodate more phone calls. This is referred to as call request buffering. The paper proposes both analytical and simulation models to study the impact of two call request buffering schemes. The results indicate the potential to greatly improve network performance (i.e., reducing the call blocking probability) by increasing the call-setup period by a reasonably small amount. For the offered load engineered at 1% blocking probability, call request buffering carries 5%-50% more load compared with the system without buffering.>

Scheduling of voice traffic on broadcast-bus local area networks is considered. A voice scheduling scheme is proposed and evaluated. In this scheme, active voice calls are organized using a distributed global queue. Access scheduling overhead is reduced by voice packet arrival anticipation. To guarantee fairness and to reduce variance of the waiting time, the voice packet build-up during the overhead period is fairly shared among all calls by a round-robin service discipline. Three different voice packet service cases, namely nonexhaustive, exhaustive, and no-queueing, are evaluated. Simple closed-form approximation formulas have been proposed which show good agreement with simulation. The proposed scheme is shown to have performance approaching that of the centralized system.>

Random access scheduling schemes for broadcast-bus-type local area networks are considered. It is found that a good access scheduling scheme not only has information about the number of packets to be scheduled, but which, more importantly controls the average of this number, which can be achieved by properly choosing the scheduling interval. Furthermore, this scheduling interval is updated in such a way that adjacent intervals overlap, which by correlation gives a better estimate for the number of packets to be scheduled. The scheduling schemes developed using such concepts provide significant performance improvement over schemes using other scheduling approaches previously reported in the literature.>