Ubiquitous Intelligence and Computing

Due to the research and technological advances, ubiquitous or pervasive computing is emerging rapidly as an exciting new discipline to provide computing and communication services all the time and everywhere. While with many ongoing initiatives, it is too far to achieve the vision that Mark Weiser described. After a comprehensive analysis on traditional paradigms, we argue that, not users-friendly, i.e., users can not get services from computer easily, is one of the main reasons. In this paper, a new computing paradigm, i.e., Transparent Computing will be presented to solve this problem partially. Accordingly, we propose and develop a pilot system, which runs in a network environment and operates at the assembler instruction level. This system lets users demand heterogeneous OSes and applications upon them from centered simple servers, similar to choose different TV channels in daily life. We also present some primitive real and experimental results to show that it is a feasible and efficient solution for future computing infrastructure.

This paper presents a novel interaction technique with multiple large displays in smart space. We call it D2LP, short for Drag and Drop by Laser Pointer, where a specially designed laser pointer uPen is a handheld interactive device for users. By D2LP, large displays driven by different computers can be made as one seamless integrated uniform system. With a uPen in hand users can directly point and draw on the surface of displays, drag and drop digital objects among multiple displays continuously and smoothly. Report on evaluation experiments shows that D2LP can effectively support the freely interacting with multiple wall-sized displays and it is preferred over the conventional mouse-driven cursor based system for the application in smart spaces with rich information displays.

Many wall-size display systems are built to provide large-scale visualization. These systems may be quite successful for some limited applications, but are very inflexible, since these systems only have fixed display regions. This paper integrates steerable projectors whose beam can be moved under computer control onto a wall-size display system to strengthen its display ability. With the steerable projectors, the integrated display system, named

Flexible Display

, provide an extendable display region. This consists of a large-scale display region and several movable display regions, such that the integrated display system has great potential in the area of human-computer interaction and information visualization. This paper applies the Flexible Display to a virtual museum application to give the users fluent navigation experience. For the application, the Flexible Display provides the following functions: 1) intensity and resolution enhancement of sub-region of display wall, 2) information augmentation, and 3) “stepping user interfaces” for its viewers interacting with display wall.

Logistics has become a fast growing industry in recent years. In particular, the large hub ports have heavy congestions in getting the containers in and out of the ports. Therefore the ports are regarded as the bottleneck in the logistics. In this paper, we analyze the new requirements of port logistics to figure out heavy congestions and then present the system prototype for identification and positioning or tracking in the port environments. Our solution is composed of three parts: RFID, electronic container seal, and RTLS system. The system design focuses on three parts: 1) the standard compliance with the ISO, 2) the energy saving mechanism to maintain the longevity of tags as long as possible, and 3) the high identification rate in the presence of multiple tags. The performance evaluation of the system is also included in the paper. These systems will provide a complete range of port logistics services.

In this research, a ubiquitous learning support system making use of the schoolbag is presented to assist elementary school pupils in their personal items management. Through some sensors, the micro-computer embedded in the schoolbag can monitor what is getting in or taking out of the schoolbag, and maintain a schoolbag item list. Moreover, the teacher can make up a schedule that specifies required items on a given day. The micro-computer then compares the schedule with the maintained item list and provides the reminding service for the pupils.

In 2004, the Computer Science Assistive Technology Laboratory (CSATL) of Utah State University (USU) started a project whose objective is to develop RoboCart, a robotic shopping assistant for the visually impaired. RoboCart is a continuation of our previous work on RG, a robotic guide for the visually impaired in structured indoor environments. The determinism provided by exteroception of passive RFID-enabled surfaces is desirable when dealing with dynamic and uncertain environments where probabilistic approaches like Monte Carlo Markov localization (MCL) may fail. We present the results of a pilot feasibility study with two visually impaired shoppers in Lee’s MarketPlace, a supermarket in Logan, Utah.

This paper presents a smart identification card system using facial biometric information for identity authentification. For a trade-off between the security and the cost, this system utilizes an architecture containing three security levels of identity authentification, “manual face verification offline”, “manual face verification online” and “automatic (biometric) face verification”, which satisfy the different security requirements of various applications of identification cards. For the function of “manual face verification online”, we bring out an idea based on decomposing the face image into two parts which are stocked into the card and the database of system respectively. And for the function of “automatic face verification”, we proposed a novel face verification scheme based on class-specific face models. The technique Active Appearance Model is applied, as the way of face modelling, to realize the proposed scheme. A prototype application of such system, which contains a fix version for PC and a mobile version for Pocket PC, is also introduced in this paper.

The TMO (Time-triggered Message-triggered Object) model is a real-time object model for real-time distributed computing. A TMO instance is a kind of autonomous object having two special types of member threads. The first type is a group of time-triggered threads activated by pre-defined timing constraints such as periods and deadlines. And the second type is a group of message-triggered threads that are activated by distributed IPC messages from remote TMO’s and finish their computing within pre-given deadlines. With these properties of the TMO, ubiquitous and real-time distributed applications can be easily organized as a logical network of TMO’s. Since the TMO model has been proposed, there have been many successful progresses in developing TMO engines based on open-source kernel platforms. The kernels are TMO-Linux for general embedded systems and TMO-eCos for small embedded/ubiquitous systems. In this paper, after introducing the general architectures and functions of the developed TMO kernels for embedded systems, some comparisons of their characteristics are also given to specify their suitable usage domains.

Previous researches show that speech interactive systems provide users not only a user-friendly interface but a feedback scheme in speech. However, these systems are built on PCs but not available anytime or anywhere. Therefore, an embedded system design is proposed for ubiquitous speech interactive applications. In addition to provide multiform interfaces, this paper also proposes multiple speech interactions by four scenarios: remote control, dictionary query, programmable dialogue, and lexicon-in sentence-out. For cost down purpose, two issues are presented to allow real-time process on a low cost and resource-limited SPCE061A (49.152 MHz) micro controller unit (MCU). One is the short-time start/end point detection of speech streams and the other is the timing arrangement of feature extraction/saving. By utilizing 18 speech streams (average 2.56 seconds) in our experiments, the average processing time for one recognition operation is 0.92 second and the recognition rate is 89.39 %.

This paper proposes an object-based ubiquitous multimedia contents storage architecture for mobile devices that employs iSCSI protocol for ubiquitous storage access over IP network and the object-based storage device model for low power. It also provides a multimedia content player that operates directly with the proposed storage architecture. We prototype the proposed storage architecture and the multimedia content player upon Linux-based desktop environments. Performance evaluation by playing MP3 multimedia contents reveals that the proposed storage architecture reduces the total power consumption by 9%, compared with existing network storage. This enhancement is mainly contributed to the fact that a large portion of the file system is moved from the mobile device into the object-based multimedia contents storage.

The problem of flash memory is that it cannot be overwritten unless erased in advance. In order to avoid having to erase during every update, non-in-place-update schemes have been widely used. In case of non-in-place update mechanism, garbage collection is needed to reclaim the obsolete space. In this paper, we study a new garbage collection scheme to reduce its cost such as the number of erase operations and the number of data copies. The proposed scheme determines the victim blocks by exploiting usage information of data blocks such as age, utilization and erase count. In addition, the proposed scheme predicts the future I/O workload and controls the number of victims to avoid disturbing the normal I/O operations. Experimental results show that the proposed scheme can perform well especially when the degree of locality is high.

A dangerous workplace like the iron production company needs a durable monitoring of workers to protect them from an critical accident. This paper concerns about a robust and accurate location tracking method using ubiquitous RFID wireless network. The sensed RSSI signals obtained from the RFID readers are very unstable in the complicated and propagation-hazard workplace like the iron production company. So, the existing particle filter can not provide a satisfactory location tracking performance. To overcome this limitation, we propose a double layered particle filter, where the lower layer classifies the block in which the tag is contained by the SVM classifier and the upper layer estimates the accurate location of tag owner by the particle filter within the classified block. This layered structure improves the location estimation and tracking performance because the evidence about the location from the lower layer makes a effective restrict on the range of possible locations of the upper layer. We implement the proposed location estimation and tracking system using the ubiquitous RFID wireless network in a noisy and complicated workplace (100m × 50m) where which 49 RFID readers and 9 gateways are located in the fixed locations and the maximally 100 workers owning active RFID tags are moving around the workplace. Many extensive experiments show that the proposed location estimation and tracking system is working well in a real-time and the position error is about 2m at maximum.

Neural networks, Bayesian networks, Markov models, and state predictors are different methods to predict the next location. For all methods a lot of parameters must be set up which differ for each user. Therefore a complex configuration must be made before such a method can be used. A hybrid predictor can reduce the configuration overhead utilizing different prediction methods or configurations in parallel to yield different prediction results. A selector chooses the most appropriate prediction result from the result set of the base predictors. We propose and evaluate three principal hybrid predictor approaches – the warm-up predictor, the majority predictor, and the confidence predictor – with several variants. The hybrid predictors reached a higher prediction accuracy than the average of the prediction accuracies of the separately used predictors.

The ability to recognize, interpret and express emotions plays a key role in human communication. Current computer interfaces have become able to ”see”, thanks to advanced video sensors and video processing algorithms; however, until recently they could not plausibly ”guess” user intentions, because available feature extraction techniques could not provide the adequate level of service needed to support sophisticated interpretation capabilities. Our approach relies on a set of novel face and posture recognition techniques efficient and robust enough to be at the basis of a fully video-enabled intelligent pervasive workplace, capable of providing value added services based on the real time of facial and postural data. We propose to build on our current work in this area to create an infrastructure for lightweight facial and posture analysis allowing a variety of extended interactions between users and their work, market and entertainment environments.

In this paper, a novel Multi-agent control system with fuzzy inference learning and its physical testbed are presented. In the Multi-agent system, distributed controlling, monitoring and cooperative learning are achieved through ubiquitous computing paradigm. The physical testbed named Distributed Embedded Intelligence Room (DEIR) is equipped with a fair amount of embedded devices interconnected in three types of physical networks, namely LonWorks network, RS-485 network and IP network. The changes of environment states and user actions are recorded by software agents and are processed by fuzzy inference learning algorithm to form fuzzy rules that capture user behaviour. With these rules, fuzzy logic controllers can perform user preferred control actions. Comparative analysis shows our control system has achieved noticeable improvement in control accuracy compared to the other offline control system.

This paper presents IPD (intelligent pervasive middleware) that provides automatic home services (consumer electronics: TV, DVD, audio, light, and air-conditioner) for human through analysis of the biometrics and environment contexts. The IPD receives the biometrics context (pulse, facial expression and body temperature, human location in smart home and human motion) from sensor devices. We handled the context’s pattern analysis in two steps. The first step selects consumer electronics (TV, DVD, audio, air-conditioner, light, project) from IPD’s rules. In the second step, IPD predicts detailed home service (for example, a detailed home service of the TV includes news, sports, and drama), using the supervised algorithm-based pattern analyzer. We used the SVM (support vector machine) for detailed service pattern analysis. We experimented on the intelligent pervasive middleware in two directions, and it was shown to have an effective performance in practical application. We are currently studying the association technique of home service (by using data mining) that can happen when IPD predicts home service by the home service predictor.

In real-time moving object tracking, each time an object moves, a new position should be sent from the moving object to the server. In this case, numerous moving objects cause a very large volume of location updating. It could bring communication, computation and update overhead between moving object and server. This paper proposes a new location tracking method that is applicable to continuous updating of current position of moving object. The techniques are based on movement anticipation with pre-evaluated arrival time information; the moving object compares actual position with anticipated position at predefined time interval and checks deviation between two positions. Only if predefined threshold exceed, the location updating, from moving objects to the server, is occurred. Using predefined route information, this approach is effective with update reduction for real-time tracking of moving object.

In order for the hidden computer to interact with human beings, a variety of sensors are used to collect data of human behavior. Using video cameras is quite straightforward on this matter. However, it is still a difficult task to do human motion analysis through only video data. In this paper, we discuss the necessity and the way to combine information from video sensors and other types of sensors, especially, the RFID reader. Data acquisition of RFID and video data is discussed first. Combinatorial information fusion techniques for multiple sources of information are then introduced. On the other hand, rule based methods provide an alternative solution for combining the two sources of information. An example in monitoring real-life learning behavior is also given.

With new channels and devices based on ubiquitous computing, UCommerce provides an environment where buyers and sellers are literally able to commerce anytime, anywhere, and anyway they like. In U-Commerce, a lot of opportunities are expectable in various industrial areas, and especially in insurance industry. In this paper, we propose new types of insurance services in U-Commerce. And as a new channel to practice it, we propose Well-Bing Store with three concepts: ‘Corporeal Insurance Product’, ‘U-Cart’, and ‘Health Examination’. And supportive processes in Well-Being Store are depicted in detail as well. By introducing Well-Being Store, we transform a physical place into an intelligent space which cares insured customers previously before they get sick. This is actualized by managing a list of foodstuffs in a U-Cart accordingly, and by keeping a health examination regularly with help of an insurance company and a hospital.

This paper proposes an algorithm for real-time license plate detection. In this algorithm, the relatively easy car plate features are adopted including the simple statistical feature and Harr-like feature. The simplicity of the object features used is very helpful to real-time processing. The classifiers based on statistical features decrease the complexity of the system. They are followed by the classifiers based on Haar-like features, which makes the final classifier invariant to the brightness, color, size and position of license plates. The experimental results obtained by the proposed algorithm exhibit the encouraging performance.

Wide deployment of ubiquitous information environment makes it possible to provide various information to many users. “Regional information” should be one of the most valuable information if “context-aware” provision to community residents is made. But “contexts” of community residents are private information and should be treated very carefully. In this paper, we propose a novel information system called RUIS -Regional Ubiquitous Information System. RUIS can obtain and store regional information from network, realize robust security mechanism of personal/private information protection, and provide information that matches to residents’ contexts. Using RUIS features we can realize“smart watch over” for children. We are working on the implementation of this application based on RUIS prototype.

As an emerging computing model, Ubiquitous Computing (UbiComp) has become a source of challenging research. In this paper, we investigate the impact of mobility and resource-limitedness of UbiComp upon service provision, and design an adaptive service delivery model which can dynamically deliver satisfying services to mobile users in tune with the variation of context. In our design, we propose a user-oriented Quality of Service (QoS) model to measure whether services are suitable for being delivered to the mobile users in UbiComp environments. Moreover, we utilize first-order logic inference and fuzzy logic evaluation in the process of service delivery. We also perform several experiments to evaluate the service delivery model in terms of scalability and performance.

An intelligent vehicle safety system can be constructed by exchanging emergency-related information between any unrelated vehicles, such as urgency stop, traffic accident, and obstacles. In the most of vehicle safety communication applications, an emergency message is propagated in the form of broadcasting. However, it causes a lot of problems in terms of efficiency due to the multi-hop propagation and radio collision problems. This paper presents a selective message forwarding method by proposing the stem and branch structure. With benefit of the structure, only one vehicle performs forwarding a received emergency event among the vehicles that are included in the same wireless coverage. Moreover, the proposed scheme improves the efficiency of message transmission with the selective assignment of priority for forwarding message. To analyze its performance, it has been evaluated by network simulator.

One of the central ideas of ubiquitous computing research is to create smart real-world artifacts and environments that can spontaneously cooperate in order to enable novel applications that no single device could provide. A critical factor in such applications is the availability of relevant context knowledge, which may be provided by a variety of different sensors, domain models, and applications. Locating, accessing, and preserving such knowledge is not a trivial task, especially in dynamic environments where new knowledge sources and applications may unexpectedly appear while existing ones may vanish. In this paper we present a pragmatic yet powerful approach to manage shared knowledge by introducing a novel concept called RDF:Stores. We will present the general properties of RDF:Stores and describe how they can be implemented using off-the-shelf software components.

Intelligent vision based surveillance has been an important research issue of distributed computing and communication in the future. Since the smart applications are implemented on cyberspace, how to control smart objects and to increase the degree of self awareness has been a challenging issue. This paper was addressed in a vision-based surveillance smart application. An Extended Petri Net was proposed for modeling the vision-based surveillance smart application. This proposed system not only provided the hyperspace facilities but also gave the certainly smart/intelligent fusion for receiver’s needs. Our testbed were focused on the swing and slide those were very common facilities in the park field or playground. We constructed several surveillance cameras with the different specified position and view angle. To obtain the moving objects which were grabbed video and it could be analyzed the potential dangers in real time. We believe that the more smart applications can be developed the more smart real world will be come true in near future.

Smart spaces pose significant technical challenges in heterogeneous device interaction/integration, user/environment perception, as well as system interoperability. As devices become more powerful and connected, the users are required to understand complex device functionalities in order to carry out simple tasks, and thus experience more and more frustration with the increase of device types and complexity. In this paper, we first examine some key issues and challenges in smart spaces, and then we propose a lightweight middleware which enables the spontaneous device interaction and can hide the complexity of heterogeneous device connection from the end users. Implementation details of the middleware prototypes are presented.

In this paper an analytical model is established to optimize the cost efficiency of broadcast operations while guarantee required RE and delay. To facilitate analysis on cost, a new metric—WCOF (Wireless Channel Occupation Frequency), is introduced to measure the average cost per node incurred by a round of broadcast. For simplicity delay is measured by the least number of broadcasts to achieve required RE. Performance of Counter based and distance based schemes are investigated in detail using the new model. Analytical results show the relation between the cost efficiency and delay of broadcast schemes and configuration of network and broadcast parameters. It is proved that the new model can be used to choose proper broadcast scheme and parameters for optimal performance.

In this paper joint power control and channel allocation algorithms are proposed for wireless ad hoc networks. The performances of the algorithms are evaluated in both FIFO queue system and Priority queue system. To increase network capacity with time-sensitive data transmission, we incorporate the back off scheme into these systems. Through in-depth simulations, the network throughput and the probability of successful transmission of time-sensitive traffic are examined by varying several parameters. We conclude by presenting the advantages and disadvantages of different scenarios.

Major researches in MANET have emphasized on the design of efficient routing protocols such as DSDV, AODV, etc. The majority of routing protocols assume that mobile nodes in ad hoc networks are configured with IP addresses before they begin communications in the network. Thus, Auto-configuration is desirable in implementing MANET. In this paper, we investigate a MANET topology that all the nodes in MANET want to connect to the Internet through a special node called the Internet Gateway and we propose a new IPv6 address auto-configuration mechanism in MANET. In order to validate the scheme, we present several simulation results.

Fairness is an important design purpose for shared channel contention based MAC protocols, such as IEEE 802.11 in wireless networks. It is a complex problem due to its many dimensions that include consideration of location-dependent contention, spatial reuse of channels, and desire to achieve fully distributed scheduling in the wireless communication systems. This paper presents a distributed fairness supported scheduling algorithm (DFSS), which accounts for the received service of each flow and adjusts it’s backoff time to ensure getting a fair service. In DFSS, each node exchanges the flow’s average service index with neighbors as one parameter of determining backoff time. The main advantage of DFSS is that nodes only need to exchange little information to achieve global and local fairness.

In order to obtain better performance in wireless ad hoc networks, the majority of existing work focuses on the multi-priority based packet scheduling mechanism under multiple service classes, admission control and rate policing, according to application requirements and network conditions. In addition, no studies have reported on the feasibility and scalability of adapting the IEEE 802.11 EDCA scheme in such networks. This paper presents a new QoS provisioning model called Dynamic Hop Service Differentiation (DHSD), which performs dynamic class selection based on estimation of each service class (AC) in 802.11e based ad hoc networks. This model is designed for Soft QoS provisioning. Performance evaluation is performed by simulation using OPNET. It is demonstrated that this QoS model outperforms existing service models in wireless ad hoc environments.

The broadcast operation in ad hoc wireless networks is essential, but expensive in terms of power consumption. In DP (dominant pruning) algorithms, the number of forward nodes is treated as a criterion to measure the consumed power. However, this principle is not suitable anymore if the power-adaptive characteristic is supported and each node can detect the strength of a received packet and tune its transmission power level. Thus, in this paper, we propose a modified version of the DP algorithm with the objective being to minimize the total power consumption of forward nodes. Comparing with original DP algorithms, the sender in our algorithm not only chooses its forward nodes, but also designates the transmission ranges of its forward nodes. As a result, the designated transmission ranges not only maintain the same coverage as DP algorithms, but also reduce the total power consumption. We propose a new forward nodes selection process where the weight of a node is dependent on the incremental transmission cost rather than the effective degree. Although the resultant number of forward nodes in the proposed approach is often greater than DP algorithms, the power consumption shown in simulations is less than DP algorithms.

Real-time video transmission in mobile ad hoc networks is an important yet challenging problem. A Bandwidth-aware Multi-path Routing Protocol is proposed to support QoS for real-time video delivery under ad hoc network environments. We incorporate the on-demand node-disjoint multi-path routing scheme with the bandwidth estimation method. In this approach, multiple node-disjoint paths are formed during the route discovery process and are actively maintained. The detector packets measure the available bandwidth of each hop along the paths. The approximate bandwidth of a node is estimated based on the bandwidth consumption information indicated in the modified HELLO messages. The estimated bandwidth is used as the metric to choose the primary route. Simulation results show that packet delivery rate increases a lot, and end to end delay and jitter decrease significantly, while the overhead is not increase too much, compared with original single path routing protocol.

One of the most critical issues in wireless ad hoc networks is represented by the limited availability of energy within network nodes. Most of the researches focused on the problem of routing issues rather than energy efficiency or prolongation of network lifetime. In this paper, we proposed a multicast power greedy clustering algorithm (termed as MPGC) with the mesh scheme in the multicasting protocol of ad hoc wireless networks. The greedy heuristic clustering partitions a large-scale ad hoc network into a hierarchical cluster structure. Nodes in a cluster determine adaptively their power levels so as to be power efficient. The clusterheads acting as the agents of transmit- ters/receivers can reduce efficiently bandwidth consumption and complexity of mesh structures. Besides, the mechanism of cluster maintenance can remarkably prolong the network lifetime. The power aware multicasting protocol based on ODMRP executes suitably on the super-nodes topology formed by clusterheads. The results of the simulation show that our scheme achieves better performance for ad hoc networks, in terms of network lifetime and network scalability.

Reconstruct a new route for network services during the handoff process is a fundamental issue of wireless communication. This paper proposes a new rerouting algorithm to achieve a fast handoff based on

k

-hop backtracking mechanism. The algorithm can dynamically decide the backtracking hops according to the velocity of wireless devices and the current network bandwidth through iterative strategy. During the backtracking process, our algorithm is able to find out an optimized route for the handoff network services and require all intermediate nodes which has received the

Location Update

information forward all received packets to Mobile Terminal with their optimal route to the destination subscriber station. This will greatly reduce the cost of packet forwarding during the handoff process.

This paper describes a novel control-theoretic distributed multicast congestion control scheme, called self-tuning proportional integrative plus derivative (SPID) controller. The control parameters can be designed to ensure the stability of the control loop in terms of source rate. The distributed explicit rate SPID overcomes the vulnerability that suffers from the heterogeneous multicast receivers. The SPID controller is located at the multicast source to regulate the transmission rate. Simulation results demonstrate the efficiency of the proposed scheme in terms of system stability and fast response, low packet loss, and high scalability.

Recently, wireless personal area network (WPAN) that is the main part of ubiquitous network has been spotlighted. Even though the space-time code (STC) is a good answer for overcoming the intensive fading of indoor channel, it needs larger size and higher cost for additional antennas. The cooperative scheme is another good solution which gives transmit diversity only with the existing hardware. However, classical cooperative schemes cannot provide full rate. In this paper, a new cooperate diversity scheme for code division multiple access (CDMA) system that provides full rate is presented. Moreover, because cooperative system supplies bad performance when the channel condition between the source and relay is poor, a technique that intelligently selects mode according to it is adopted. We apply the proposed scheme to the direct sequence ultra wideband (DS-UWB) system and evaluate its performance.

The current QoS multicast routing model aims to solve a one-objective optimization problem with one or more bounded-constraints, such as delay, delay jitter, bandwidth, etc. To satisfy the individual requirement for users in multiple QoS networks, we analyze the limitation of the current model and propose a new QoS multicast routing model that supports multi-objective optimization. The new model considers the QoS guarantee as QoS optimization objectives rather than QoS constraints. It overcomes the limitations that exist in the traditional multicast routing model. Furthermore, a new routing algorithm to deal with the new model based on immune principles and Pareto concepts is given. In this algorithm, a gene library is introduced to speed up the algorithm to satisfy the real-time requirement of the routing problem. The initial experimental results have shown that the new algorithm can effectively produce more than one Pareto optimization solution compromising all QoS objectives within one single running.

A novel collaborative strategy employing two interacted relays is developed for the multihop inter-vehicle communication (IVC) networks in this paper, which has great complexity degradation. Instead of the traditional terminal collaboration or distributed space-time coding (DSTC), a relay tier interaction is proposed. Considering time division multiple access (TDMA) scheme, two spatial adjacent vehicles are conjoined by an interaction radio link to create a virtual antenna array (VAA). Then the collaborative two-relay is integrated into a multihop IVC network mentioned above, which enables a single-antenna terminal to share its antenna with other terminals easily, and avoids the synchronization or other complicated requirements in terminal collaboration or DSTC systems. Furthermore, except the low complexity in hardware implementation, the proposed method also can enhance the performance of the IVC systems over both the quasi-static fading channel and time varying Rayleigh fading relay channel, which is verified through the Monte-Carlo simulations.

IEEE 802.11e enhanced distributed coordination function (EDCF) provides a priority scheme by differentiating contention window sizes, medium occupancy limits and arbitrary interframe spaces (AIFS). In this paper, Gibichini’s model has been extended to depict EDCF mechanism. The proposed model analyses channel throughput under finite loads, which employs two states to accurately depict the backoff scheme. The proposed model also takes packet arrival rate, contention window size, maximum backoff state and AIFS into consideration. Simulation results proved the accuracy of the proposed model and the enhancement of the EDCF scheme.

This paper introduces an opportunistic packet scheduling method and medium access control (MAC) scheme for controlling the throughput in wireless local area networks (WLANs). The proposed method takes advantage of the multi-user diversity in time-varying wireless channel while the asymmetric traffic load problem between the uplink and the downlink is alleviated. The proposed method can be implemented without the modification of the deployed IEEE 802.11 nodes. The performance of the proposed method is compared with IEEE 802.11 Distributed Coordination Function (DCF) by computer simulations.

The nodes in wireless sensor networks are energy-constrained and have limited bandwidth. In addition, communication links between nodes are unpredictable and unstable. These factors make the design of routing protocols for such a network very challenging. We propose a scalable, efficient and reliable routing protocol called EAR that provides reliable data delivery from nodes to hubs in a wireless sensor network. Simulation results have shown EAR outperforming existing routing protocols.

In this paper, we proposed an approach for Quality of Service (QoS) routing algorithm of Wireless Sensor Networks (WSNs) based on Ant Colony Optimization (ACO). The special characteristics of WSNs need to reduce the computational complexity and energy consumption of the QoS routing algorithm especially. We note that ACO algorithm using collective intelligence of artificial ants as intelligent agents is very appropriate to solve the combinatorial optimization problems in a fully distributed way, so in this paper we use modified ACO approach to solve Delay Constraint Maximum Energy Residual Ratio (DCMERR) QoS routing problem of WSNs. The QoS routing solution proposed in this manuscript, which is named as ACO based QoS routing algorithm (ACO-QoSR), searches for the best paths, which are satisfied with the QoS requirements with intelligent artificial ants. To overcome the problem of limited energy in WSNs, there are some modifications to enhance ACO’s convergence rate. ACO-QoSR algorithm is the tradeoff between a certain guaranteed QoS requirements and acceptable computational complexity. The simulation results verify that ACO-QoSR algorithm can reduce the selected paths’ delay and improve the selected paths’ normalized energy residual ratio at the similar levels of routing overhead.

LEACH is one of the most popular hierarchical routing algorithms for sensor networks. In LEACH, the cluster number per round takes an important effect on the network lifetime. Thus the goal of the cluster head selection algorithm is to ensure that the expected number of clusters per round equals a pre-determined optimal value

k

. However, a slight inaccuracy is spotted in the computation of the node self-selected probability per round. As a result, the number of clusters produced by the algorithm is distributed in a large range around the target value. We propose an improved clustering scheme (I-LEACH) to correct this inaccuracy. The experimental results show that this scheme can ensure a more stable number of clusters.

Wireless Sensor Network (WSN) is expected to have a significant impact on the efficiency of military and civil applications such as target field imaging, intrusion detection, weather monitoring. Sensors with low cost, low power and multifunction are expected to last until their energy drains, so energy-conserving forms of communication and computation are essential to the sensor node lifetime. Given the unreliable nature of the wireless channel and the high failure rate of the individual sensors, a fault tolerant routing protocol with energy-efficiency is expected to overcome these problems. In this paper, we propose an energy efficient multipath routing algorithm (EMRA) for wireless sensor networks. EMRA can efficiently find a disjoint multipath and provide a protection for routing failure. From the observed results, EMRA performs well in terms of average dissipated energy and delay to set up backup path.

Wireless sensor networks play a key role in monitoring remote or inhospitable physical environments. One of the most important constraints is the energy efficiency problem. Power conservation and power management must be taken into account at all levels of the sensor networks system hierarchy. Especially, DPM (Dynamic Power Management) technology, which shuts down the devices when not needed and wake them up when necessary, has been widely used in sensor networks. In this paper, we modify the sleep state policy developed by Simunic and Chdrakasan in [1] and deduce a new threshold satisfies the sleep-state transition policy. Nodes in deeper sleep states consume lower energy while asleep, but require longer delays and higher latency costs to awaken. Implementing dynamic power management with considering the battery status and probability of event generation will reduce the energy consumption and prolong the whole lifetime of the sensor networks. The sensor network consumed less energy in our simulation than that in [1].

The lifetime optimization is a key challenge of sensor networks. Since data transmission is the main energy consumer, it is important to make use of energy efficient communication protocols to prolong the network lifetime. We formalize the lifetime optimization problem to a max flow problem in a directed graph with capacity powers on arcs and vertices. Then we propose a fast algorithm to solve this problem. The method gives the value of maximum lifetime exactly. The time complexity is

O

(|

V

|

2

.|

A

|)

Coverage has been an active research area in mobile sensor networks. For a randomly placed large-scale sensor network, sensor nodes would most probably be distributed asymmetrically, and it requires the coverage algorithm to do with the diffusion and contraction of the network. Most of the existed algorithms are on the assumption that sensor nodes are initially densely distributed or the states of the network coverage are known to all the nodes, which does not meet all application scenarios. This paper proposes a new adaptive coverage algorithm based on the combination of boundary contraction and random repulsion. It works well on the scenarios of the asymmetrical initial distribution, the isotropic sensor nodes, and that only the coverage states in communication range being known by nodes. Simulation results show that the algorithm realizes both the diffusion and contraction of the sensor network, and that the deployed nodes tend to be uniformly distributed.

In wireless sensor networks, it is important to efficiently disseminate information from each source to a sink node. In particular, in mobile sink applications, due to the sink mobility, a stationary dissemination path may no longer be effective. The path will have to be continuously reconfigured according to the current location of the sink. In this paper, an Adaptive Reversal Tree (ART) protocol, based on the Adaptive Reversal algorithm and dynamic Root change mechanism, is proposed. Data dissemination from each source to a mobile sink can be easily achieved along the ART without additional control overhead, because the ART proactively performs adaptive sink mobility management. In addition, the ART can maintain a robust tree structure by quickly recovering the partitioned tree with minimum packet transmission. Finally, the simulation results demonstrate that the ART is a considerably energy-efficient and robust protocol.

Without congestion control techniques specifically designed for sensor network applications, the system may not function properly due to data transmission failure. Moreover, most many-to-one communication schemes adopted by sensor network applications cause a practical congestion problem called the funneling effect. With this problem, the hottest area is confined to the connection between the sink and its neighbors. In order to solve the bottleneck problem, we propose a congestion-control technique which uses adaptive time-slot scheduling and a service-differentiated technique. The simulation results show that the proposed method outperforms the conventional technique.

In this paper, an information-driven sensor selection algorithm is proposed to select sensors to participate in Kalman filtering for target state estimation in sensor networks. The mutual information between the measurements of sensors and the estimated distribution of the target state is considered as the information utility function to evaluate the information contribution of sensors. And only those sensors with larger mutual information are selected to participate in the Kalman filtering iterations. Then the geographic routing mechanism is utilized to visit these selected sensors sequentially and set up a path to transport the state estimation information to the sink node. Simulation results show that compared with the shortest path tree algorithm, the information-driven sensor selection algorithm involves smaller participated sensors, and shorter total communication distance, while the estimation performance approaches the same bound.

A distributed sensor network with mobility provides an ideal system platform for surveillance and for search and rescue applications. We consider a system design consisting of a set of autonomous robots communicating with each other and with a base station to provide image and other sensor data. A robot-mounted sensor which detects interesting information coordinates with other mobile robots in its vicinity to stream its data back to the base station in a robust and energy-efficient manner. The system is partitioned into twin sub-networks in such a way that any transmitting sensor will pair itself with another nearby node to cooperatively transmit its data in a multiple-input, multiple-output (MIMO) fashion. At the same time, other robots in the system will cooperatively position themselves so that the overall link quality is maximized and the total transmission energy in minimized. We efficiently simulate the system’s behavior using the Transaction Level Modeling (TLM) capability of SystemC. The simulation results demonstrate the utility of our design and provide insights into performance of the system.

Typical applications of wireless sensor networks require scalability because they are deployed in a large area. Conventional centralized systems for acoustic source localization experience practical difficulties since they focus on high accuracy in a small area rather than scalability in a large region. In this paper, we present a scalable and light-weight acoustic source localization system suitable for use with resource constrained hardware. The distributed mechanism has low complexity and an error-tolerant algorithm for construction of a feasible system in a real environment. The proposed system has been implemented and validated by experiments in a real environment.

The wireless sensor network (WSNs) was extensively deployed and researched for many applications in recently. By taking the advantage of smaller dimension, lower cost and simple structure of the sensor node, the more restrictions come together with sensors. Therefore, the limited power of sensor nodes is the most direct and difficult problem we meet. The limitation on the energy of sensor node makes the bottlenecks for designing the suitable routing protocols. In order to solve the problem of limited energy, the loading of nodes have to be distributed as possible as it can. If the energy consumption can be shared averagely by most nodes, then the lifetime of sensor networks will be enlarged. Thus we propose the routing protocol called Reactive Energy Decision Routing Protocol (REDRP) for sensor networking by considering several representative routing protocols in different structures. This protocol will create the routes in reactive routing method to transmit the data node gathered and the REDRP use the residual energy of nodes as the routing decision for energy-aware. As the results of simulation shows that the more fairness usage of sensor nodes, the total energy consumption of entire network will be distributed fairly by our protocol and the lifetime will also be increased.

Pipelining the messaging between sensor nodes increases the overall throughput of the querying system, however at the cost of extra communication. But for long running queries, the messages communicated in pipelined architecture are even less than the normal count of messages in any query processing methodology in sensor networks, as also pointed out in previous work. In this paper we device a novel methodology to process aggregation queries in sensor networks by using the systolic architecture. We explicitly define and stipulate the use of systolic message communication as aggregation query processing technique to yield increased response time with the saving of energy by reduced message communication when considering long running queries. We show through simulation the two-fold gain using the proposed technique as compared to methods without pipelining.

Research activity in the area of medium-access control protocols of wireless sensor networks (WSN) has grown dramatically in the past few years. A number of MAC protocols are proposed to achieve energy efficiency upon the limitation of WSN which are usually deployed in a special environment, assigned with long-term work, and supported by limited battery. It is found that there is a trade-off between energy efficiency and transmission delay, thus duty cycle of fixed sleep/listening MAC protocols has to be adjusted carefully to achieve the best performance. Taking the challenge to design an adapted listening mechanism, A-MAC is designed to adaptively and dynamically adjust the duty listening time based on traffic load. With A-MAC, it can achieve more power efficiency in low traffic load and much less transmission latency comparing to existing MAC protocols. We simulate A-MAC, and the simulation result shows that A-MAC can significantly prolong the lifetime of network when traffic is low and reduce packet delivery latency.

In this paper, we propose a relay shift based approach to solve uneven sensor distribution problem due to the initial random dropping or the existence of faulty sensors. The distinguishing feature of our work is that the sensors in our model have limited mobility. After determining the optimal cluster head positions by particle swarm optimization (PSO) method, we use proposed Relay Shift Based Algorithm (RSBA) for movement assisted sensor deployment. Dijkstra’s algorithm is applied to find a shortest path from a redundant sensor to a virtual node point in an uncovered area, and each sensor moves along this path by relay shift based on the principle that evenly distributed sensors can provide better coverage. Simulation results show that our approach can provide high coverage within a short time and limited movement distance as well as ensuring connectivity and energy efficiency.

In this paper, a novel data dissemination scheme in wireless sensor network is proposed. A global location-based structure called

transfer posts

are adopted to act as the data forwarding stations among sink nodes and source nodes with mobility. Multiple source nodes share the transfer posts. The sink nodes can easily request or collect data through transfer posts in which store the location information of source nodes. Evaluation results show that our approach consumes less energy compared to recently results such as two-tier data dissemination (TTDD) [3] and Railroad [6]. We have resolved the problem of query forwarding and data delivering among mobile sinks and multiple sources with the remarkable performance.

The sensor network environment will become the standard daily environment in the next generation. The technology of the sensor network can combine with computing and daily space. Combinations of computing and sensor networks happen only inside a computer. In other words, a user cannot understand what happens inside the computer. In particular, our system carries out reasoning. There is no research on the visualization of reasoning processes. The visualization of reasoning processes is a complex process. Therefore, we need to develop visualization of reasoning processes. The main technologies to visualize reasoning are to resolve motion and to use 3D space expression. This paper proposes a method of visualization of reasoning processes in a sensor network environment.

Development of energy-efficient, traffic-adaptive MAC algorithms that provide both high reliability and easy implementation property is the current major focus in wireless sensor network research. The operational characteristic of the station in packet transmission is completely different from those of deferring stations in steady state. In this paper, the fast collision resolution MAC algorithm is combined with traffic-aware queue status managements to achieve high energy efficiency and high performance at the same time. Through the implementations and various performance studies, the proposed algorithm shows significant performance improvements in wireless sensor networks.

In this paper, bidirectional data aggregation (BDA) scheme is proposed for wireless sensor networks. Traditionally, data aggregation bas been performed in backward direction (from source to sink) where each node in the network combines data from its child nodes. BDA algorithm, however, aggregates sensory data in both directions (sink to sources and sources to sink) when the sink is interested in gathering singular aggregates such as

MAX

and

MIN

. In forward aggregation (sink to sources), each node tags its sensor reading to the ongoing query only if its local reading is not redundant. Node receiving the tagged query suppresses its response if its local reading is redundant. By doing so, we can limit a number of redundant and unnecessary responses from the sensor nodes, saving energy. Performance evaluation shows that BDA algorithm significantly improves energy-efficiency as well as provides an accurate response for a given singular query in the presence of time-varying sensor readings.

IEEE 802.16 PMP mode recommends using truncated Binary Exponential Backoff (BEB) algorithm to resolve the contention when multiple Subscriber Stations (SSs) compete for the connection and resource allocation from the Base Station (BS). The BEB approach may not be effective and transmission opportunities (TOs) may be wasted due to independent contention and backoff of each SS. This paper proposes an efficient coordinated backoff algorithm (COB) through BS coordination so that a global window can be assigned by the BS and the TOs can be effectively consumed by the SSs. Theoretical analysis and simulation results show that COB is adaptive to the dynamic change of active SSs and outperforms the BEB approach.

A grid-like environment may be constructed from ad hoc processing devices, including portable battery-powered devices. Battery lifetime is a current limitation here. In this paper we propose policies for minimizing power consumption using voluntary collaboration between the autonomously controlled nodes. We exploit the quadratic relationship between processor clock-speed and power consumption to identify processing devices which can be slowed down to save energy while maintaining an overall computational performance across a collaboration of nodes.

Recommendation systems, which automatically understand user preferences and make recommendations, are now widely used in online shopping. However, so far there have been few attempts of applying them to real-world shopping. In this paper, we propose a novel real-world recommendation system, which makes recommendations of shops based on users’ past location data history. The system uses a newly devised place learning algorithm, which can efficiently find users’ frequented places, complete with their proper names (e.g. “The Ueno Royal Museum”). Users’ frequented shops are used as input to the item-based collaborative filtering algorithm to make recommendations. In addition, we provide a method for further narrowing down shops based on prediction of user movement and geographical conditions of the city. We have evaluated our system at a popular shopping district inside Tokyo, and the results demonstrate the effectiveness of our overall approach.

Energy is a vital resource in pervasive computing. Remote execution, a static approach to energy saving of mobile devices, is not applicable to the constantly varying environment in pervasive computing. This paper presents a dynamic software configuration approach to minimizing energy consumption by moving or/and replicating the appropriate components of an application among the machines. After analyzing three types of energy costs of the distributed applications, we set up a math optimization model of energy consumption. Based on the graph theory, the optimization problem of energy cost can be transformed into the Min-cut problem of a cost graph. Then, we propose two novel optimal software allocation algorithms for saving power. The first makes use of component migration to reasonably allocate the components among the machines at runtime, and the second is to replicate some components among machines to further save more energy than component migration. The simulations reveal that the two proposed algorithms can effectively save energy of mobile devices, and obtain better performance than the previous approaches in most of cases.

Applications that can follow mobile users when they change to a different environment are in high demand by pervasive computing. In this paper, we describe a mobile agent based paradigm for enabling an application to migrate with the user in pervasive computing environments. Compared with existing efforts on application mobility, our approach has the following distinctive features: (1) Applications are supported by a middleware with a reflective architecture that helps separate business functions from context-awareness logic; (2) Mobile agent is used to manage the mobility of an application and help the application adapt to its new context; (3) The advantages of mobile agent, such as reactivity, autonomy and intelligence, are naturally incorporated into the pervasive computing environment. Our experience shows that mobile agent is a promising technology for pervasive and mobile computing where mobile agents can act as a bridge connecting the cyber world with the physical world.

Ubiquitous computing systems are connected with a number of sensors and devices immersed in the environment, spread throughout providing proactive context aware services to users. These systems continuously receive tremendous amount of information about their environment, users and devices. Such a huge amount of information deserves special techniques for efficient modeling, storage and retrieval. In this paper we propose the modeling of context information as time series and applying the time series approximation techniques to reduce the storage space requirements and for faster query processing. We applied an algorithm based on non-linear interpolation to approximate such data and evaluated the approximation error, storage space requirements and query processing time.

Pervasive computing paradigm is based on much more dynamic environment than the existing computing paradigm. Therefore, we need a middleware framework for efficiently integrating the pervasive objects. In this paper we propose an adaptive bridging approach for the interoperability of heterogeneous middleware. Here the objects are classified into hot or cool objects according to the requested frequency reported by the portable interceptors, and the cool spot objects and hot spot objects are assigned dynamic bridge and static bridge, respectively. The proposed approach is evaluated using a test bed comprising of three different platforms – JADE, MICO, and CALM – on nine servers. The experiment reveals that the proposed adaptive bridging approach significantly reduces the response time compared to the case of using only either static bridge or dynamic bridge, especially when the number of activated objects is relatively large.

This paper presents a comparison of version 1.0 Protocol Specification for 900MHz Class 0 RFID Tag with that of Class-1 Generation 2 UHF RFID Protocol for Communications at 860 MHz – 960 MHz for large number of tags. Although the Generation 2 protocol has been released for Class 1, it is intended to be used by all classes of RFID tags. Using detailed simulation we compare their performance and security features. If security is a lower priority over speed in cases where one can be sure that the risk of presence of an eavesdropper is low, Class 0 draft protocol should be used as it provides a definite advantage over Generation 2 protocol. In application areas where the risk of consumer identity/privacy theft is high (consumer goods area), Generation 2 provides the security that eliminates the vulnerability of the RFID EPC structure.

This paper presents a novel wireless projecting and cooperating approach named FreeSpeech based on Ad Hoc network, which makes conference exchange more convenient and ubiquitous. All attendees can access this system by their laptops or other mobile computing devices all around the conference venue and make the projections of their presentations without connecting to the projector physically. Moreover, they can make cooperative display simultaneously with others. FreeSpeech takes a mixed

Virtual Networking Computing

(VNC) C/S mode based on some improvements on the traditional ones. A novel IP multicast-based transfer and control mechanism is presented to make large numbers of attendees obtain the wall screen contents by their laptops locally without aggravating the overhead of bandwidth remarkably. An efficient remote screen synchronization method is employed to transfer screens of some attendees to the projector-connected computer, which also can save the consumed bandwidth significantly. All above methods use the limited wireless bandwidth more efficiently. Performance evaluation shows that this approach can work perfectly for ubiquitous conference occasions.

The use of data broadcasting in wireless environment has been of much interest especially to deal with the exponential increase of mobile users due to its scalability. In this paper, we present a unified broadcast model in multi-channel wireless databases and its comprehensive performance analysis. This model aims to minimize query access time, tuning time and power consumption of mobile users when obtaining broadcast database items. This scheme also concerns with single and multiple data items request. A prototype and simulation-based experiment has been developed to evaluate the performance of the broadcast model. We compare the performance of the proposed model against the conventional scheme and we found that the proposed unified model provides substantially better performance in every aspect of the evaluation. It is also shown that the results of our simulation are very close to those obtained from the prototype system.

The ubiquitous smart home is the home of the future that takes advantage of context information from the human and the home environment and provides an automatic home service for the human. Human location and motion are the most important contexts in the ubiquitous smart home. We present a real-time human tracker that predicts human location and motion for the ubiquitous smart home. We used four network cameras for real-time human tracking. This paper explains the real-time human tracker’s architecture, and presents an algorithm with the details of two functions (prediction of human location and motion) in the real-time human tracker. The human location uses three kinds of background images (IMAGE1: empty room image, IMAGE2: image with furniture and home appliances in the home, IMAGE3: image with IMAGE2 and the human). The real-time human tracker decides whether the human is included with which furniture (or home appliance) through an analysis of three images, and predicts human motion using a support vector machine. A performance experiment of the human’s location, which uses three images, took an average of 0.037 seconds. The SVM’s feature of human’s motion recognition is decided from pixel number by array line of the moving object. We evaluated each motion 1000 times. The average accuracy of all the motions was found to be 86.5%.

This article proposes the Augmented Library (AL), which is a library system augmented by ubiquitous computing (UbiComp) technology. In UbiComp, users interact with computers without being aware that they are actually using a computer. We applied the UbiComp concept to a library system. In the AL, the action of removing and returning books, which are the natural actions in a library, become the library system’s input. This function is called a ”real-time book database update.” The system uses a combination of infrared sensors and CCD cameras to identify which book is moved from the shelf. After developing a prototype a prototype system, we conducted experiments that investigated the effectiveness and effectiveness and efficiency of AL.

A pervasive environment needs to take account of a user’s context and preferences in determining which services to provide to the user. Moreover, one of the important features of a pervasive service environment is its dynamic nature, with the ability to adapt services as the context of a user changes, e.g. as he/she moves around. This paper describes how these requirement changes can be sufficiently accounted for by using a personalisation component to ‘decide’ what a user needs, and a composition component to continuously monitor services and the changes associated with them. The paper presents how services can be recomposed dynamically if the changes in context require it. This approach has been incorporated into a platform to support pervasive services. The service composition process used is described, and the way in which personalisation is incorporated into this process is shown. Finally the paper provides a brief account of two prototypes built as a proof of concept for these ideas.

In this paper, we present an Ontology-based Context-aware multi-Agent Service System(OCASS) architecture to provide these context-aware services in a smart home. To model various contexts, we design Ontology which supports to share context knowledge, detect and resolve inconsistency of the knowledge. In addition, we classify context-aware services into three layers – Session, Task, Subtask – to make sure definition of service conflict problems and to solve the problems easily. With our context model and the classification of service conflicts, our system supports autonomic tasks including recognizing and learning user’s formal/informal activity pattern, and resolving conflicts between services in different situations of users invisibly.

Ubiquitous systems use context information to select and adapt multimodal user interfaces and appliances for individual users in certain situations. However, in order to enable true reactive environments, context information has to be adequately collected, filtered, and processed and combined with user, device and other profiles. In this article, we present how an XML-based transcoding system can be applied for advanced profile processing and evolution. We demonstrate how to encode domain knowledge into sets of rules, which perform adaptations of user, device and context profiles for smart environments.

For a smart home in ubiquitous computing environments, execution of all the home services must be dependent on user’s situation contexts, which are dynamically generated in ubiquitous environments. In this paper, we propose a home-network service system that can support home services appropriate to user’s situation information in ubiquitous computing environments. The suggested system uses a uWDL workflow service scenario [2] describing user’s situation information as service execution constraints and supports a context-aware home service through comparing contexts described in the service scenario with user’s situation information generated from ubiquitous computing environments. To do that, the suggested system represents contexts described in a uWDL document as a context subtree, which expresses not only context data but also relation information among services into the fields of its node. The suggested system uses a algorithm for context comparison between context subtrees and user’s situation information. The algorithm selects a context that has all together values and types entirely equal to those of user’s contexts. Therefore, the suggested system will be useful in the development of a context-aware home-network workflow service application based on uWDL in the ubiquitous computing environment.

Context-aware computing is widely researched in recent years, but we lack for a powerful context-aware middleware which supports a uniform programming model. So developing context-aware applications is still complex and time-consuming. We introduce a workflow embedded middleware called FollowMe. It supports pluggable context-aware applications. FollowMe includes a workflow engine and sustains applications described with pvPDL, which is a workflow definition language proposed specially for context-awareness. The employment of workflow makes the development of applications simplified and the maintenance much easier. We testify the improvement by realizing an example and the related evaluation.

In ubiquitous computing environments, the need of contents adaptation and delivery is growing to support multiple target platforms for single source. Since 2D/3D graphics contents deal with a large data set and a high performance, a service adaptation for context changing is required to manipulate graphics contents with a more complicated method in multiple devices such as desktops, laptops, PDAs, mobile phones, etc. In this paper, we suggest a new notion of service adaptation middleware based on service rendering algorithm, which provides a flexible and customized service for user-centric 2D/3D graphics contents. The service adaptation middleware consists of

Service Adaptation

(SA) for analyzing environments,

Service Rendering

(SR) for reconfiguring customized services and

Intelligent LOD Algorithm

for generating an adapted service by processing customized data according to the level of detail. These adaptation services are able to intelligently and dynamically support the same computer graphics contents with good quality, when user environments are changed.

The key feature of ubiquitous computing services or applications is that they should be highly adaptive to events and context information. These factors are essential in ubiquitous computing environments. The services and applications must communicate with each other through fixed or ad-hoc networks. In ubiquitous computing, the event and context managers must be provided at the middleware level, for convenient development of associated applications. In this paper, ubiquitous computing middleware supporting event and context management services are proposed. The JavaSpaces service in Jini network technology is proposed to modify and develop the event manager, because JavaSpaces contains various interfaces that can be used to implement event management services, such as write, read and notify. Due to excellent GUI support, Macromedia Flash was used to represent the virtual ubiquitous computing environment, with communications through XMLSocket. It is demonstrated that the developed event and context managers can make it straightforward to efficiently develop ubiquitous computing applications.

With the development of wireless network and embedded system, the Ubiquitous Computing technology has been applied in learning domain, which is called Ubiquitous Learning (U-Learning). We proposed the Frame-based Interaction and Learning Model (FILM) framework to model the context-aware applications of U-Learning for the construction, maintainability and extensibility issues. The FILM contains context-aware frame layer (lower layer) and U-Learning activity script (higher layer). In lower layer, the context information and event process are handled by frame structure and the slot attachment, and the relations of context frames are managed by the context ontology. In higher layer, the sequence of frames are combined as the context-aware services and represented by the XML-based U-Learning Activity Script and High Level Petri Net is applied for events flow control and activities sequence construction. Finally, a context-aware scenario is described and modeled by FILM.

Steganography is the technology of hiding messages in such a way that no one except the authorized recipient knows the existence of the messages. In steganography, a message is hidden in some cover message. The larger the cover message is relative to the hidden message, the easier it is to hide the latter. When the hidden message is an image, it is difficult to hide the message into another image unless the size (in number of bits) of the hidden image is much smaller than that of the cover image. To solve this problem, this paper proposes a novel steganographic technique based on image morphing. The basic idea is to transform the hidden image into a morphing image, and use the morphing image as the stego message. The authorized recipient can recover the hidden image from the morphing image through demorphing. The morphing image can also be used directly for certain purposes.

In this paper, a modified signature scheme with (

t, n

) threshold group-oriented for applications in an ubiquitous environment is proposed. In the proposed scheme, any group of more than

t

out of

n

members can cooperate to generate the group signature. The message recovery property is also added into the scheme. By employing the public key of the verifier, the transmission of the signature will be secured. Finally, to detect a removal by an eavesdropper, the linkage property between signed ciphertexts of message blocks is also added into the scheme.

Modern network attacks range from fully automated to multilayer attacks. Vulnerabilities in a system are exploited by an intelligent attacker to facilitate to do anything from denial of service (DoS) attacks to the system takeover. This paper addresses the development of an architecture that includes the use of fault tolerance and honeypot technology to provide layered protection to avoid a single point of failure.

Security schemes of pairwise key establishment play a fundamental role in research on security issue in wireless sensor networks. The properties of weak connectivity in subsection based on hypercube model are addressed, for purpose of achieving inner-area pairwise key establishment. Also a hybrid hypercube model based on location-aware deployment is proposed, focusing on dramatic varieties of network size. Corresponding algorithms on inter-area pairwise key establishment are presented, as well as security and performance analysis.

Sensor networks are widely used in a variety of commercial and military applications due to their self-organization characteristics and distributed nature. As a fundamental requirement for providing security functionality in sensor networks, key predistribution plays a central role in authentication and encryption. In this paper, we describe the hexagon-based key predistribution scheme and show how it can improve the performance of key predistribution in sensor network through the use of bivariate polynomials in a hexagonal coordinate system based on the deployment information about the expected locations of the sensor nodes. More specifically, we show that the hexagon-based key scheme can improve the probability of establishing pairwise keys between sensor nodes by more than 40% over previous schemes.

A prerequisite for secure communication between two sensor nodes is that the nodes should share a session key to bootstrap their trust relationship. The open problems are how to set up a session key between communicating nodes, and how to minimize any information about the keys disclosed to the other side during key agreement. Almost schemes one of the existing protocols cannot perfectly solve above problems due to some drawbacks. Accordingly, we propose a strong key pre-distribution scheme having the following merits. First, it supports authentication services. Second, as each node can only find

w

(=⌈

W

/2⌉) keys of secret keys (i.e.

W

) that shared with other side, without revealing the unshared key information, it substantially improves resilience of network against node capture. Our scheme is based on the MRS scheme. Performance and security analyses have proven that our scheme is suitable for sensor networks in terms of performance and security aspects.

Recent measurements for Public Key Cryptography (PKC) protocols on 8-bit wireless sensor nodes showed optimistic results. It has been shown that Elliptic Curve Cryptography (ECC) is quite applicable to WSN. Still, PKC is much expensive in terms of computation and memory compared by the Symmetric Key Cryptography (SKC). In addition, in PKC, each public key needs to be authenticated before it’s used. We believe that sooner or later, PKC will be widely deployed in WSN. Therefore, we present a cooperative distributed public key authentication scheme that does not require any cryptographic overhead. In our scheme, each node is let to store a few number of hashed keys for other nodes. When a public key authentication is required, nodes who store this key help in authenticating it in a distributed and cooperative way. We consider the constrained resources of the sensor node. Additionally, we extend our scheme to fit with small range of authentication error.

Similar to hand-written signatures, digital signatures are designed to provide authenticity, integrity and non-repudiation. In the arena of ubiquitous computing, the ability of convincing

any

third party should be restricted. In a service such as an Internet trial-browsing service, a user is allowed to access the service of

t

times without any charge, but will be charged on the

t

+1 count of access. In this paper, we introduce the notion of restricted universal designated verifier signature. In this notion, a signature holder can convince up to

t

verifiers, and the convincing statement is designated to these verifiers. However, when the signature holder uses the signature for

t

+1 times, then the signature will become publicly available. We note that this type of signature schemes has many applications in practice.

Security schemes of pairwise key establishment, which enable sensors to communicate with each other securely, play a fundamental role in research on security issue in wireless sensor networks. A new kind of pairwise key pre-distribution model for sensor networks is proposed. And in addition, based on which, an efficient dynamic key path establishment algorithm is designed. Theoretic analysis and experimental figures show that the new algorithm has better performance than those previous related works.

In intelligent ubiquitous home environments, access control should be able to accommodate dynamic mechanism that was considered various access control components including temporal and spatial information. In this paper, we propose dynamic role based access control model can support smart and secure services suitable for intelligent ubiquitous home. The proposed model is appropriate for user-oriented service based on ubiquitous computing. In the proposed model, a permission is assigned by context information. In addition, this paper observes detailed access control procedures for service in intelligent ubiquitous home, providing use case scenarios and prototype design.

As the portable devices are becoming popular, the demand for ubiquitous Internet access services has grown extremely. However the increasing demand for ubiquitous services imposes more security threats to communications due to open mediums in wireless networks. The further widespread deployment of Wireless IP networks, however, depends on whether secure networking can be achieved. We propose an efficient authentication framework for wireless mobile network, which is based on one-time password (OTP) mechanism. Further we have simulated the implementation of proposed scheme and EAP-TLS and analyzed the performance in terms of various performance metrics.

Low-cost Radio Frequency Identification (RFID) tags affixed to consumer items as smart labels are emerging as one of the most pervasive computing technologies in history. This presents a number of advantages, but also opens a huge number of security problems that need to be addressed before its successful deployment. Many proposals have recently appeared, but all of them are based on RFID tags using classical cryptographic primitives such as Pseudorandom Number Generators (PRNGs), hash functions, or block ciphers. We believe this assumption to be fairly unrealistic, as classical cryptographic constructions lie well beyond the computational reach of very low-cost RFID tags. A new approach is necessary to tackle the problem, so we propose a minimalist lightweight mutual authentication protocol for low-cost RFID tags that offers an adequate security level for certain applications, which could be implemented even in the most limited low-cost tags as it only needs around 300 gates.

Infrastructureless networks are becoming more popular with the increased prevalence of wireless networking technology. A significant challenge faced by these infrastructureless networks is that of providing security. In this paper we examine the issue of authentication, a fundamental component of most security approaches, and show how it can be performed despite an absence of trusted infrastructure and limited or no existing trust relationship between network nodes. Our approach enables nodes to authenticate using a combination of contextual information, harvested from the environment, and traditional authentication factors (such as public key cryptography). Underlying our solution is a generic threshold signature scheme that enables distributed generation of digital certificates.

In general, privacy can be viewed as the right to be left alone when desired (solitude), the right to remain anonymous (anonymity), and the right to confidentiality (secrecy of information). More specifically, location privacy is “the ability to prevent other parties from learning one’s current or past locations”. In this paper, we focus on two primitives that make up location privacy:

identity information

and

location information.

Identity information has to do with the static attributes and characteristics that uniquely identify a person. Information about an individual’s identity can also be inferred based upon their location at various times (in other words, their

activities

can give away identity information). The other type of information upon which we will focus,

location information

, deals specifically with the whereabouts of an individual or group. We will also describe

location-aware applications and services

and their relationship with location privacy.

Because sensor networks consist of devices with weak physical security, they are likely to be compromised by an attacker. So, it is very important to establish a pairwise key securely between the communicating nodes. However, utilizing one hop local keys are known to be very vulnerable to threats caused by compromised nodes. This paper proposes a scheme where each node establishes three hop local keys at network boot-up time and employs them for a later pairwise key establishment. When any two nodes agree a pairwise key, all nodes on the route between two nodes contribute to the agreement of the pairwise key. Here, the initial three hop local keys are employed for encrypting a secret key delivered from a node to other nodes. Therefore, the proposed scheme bothers attackers to compromise much more nodes than the scheme using one hop local keys only. The simulation results showed that the proposed scheme provides better performance and higher security than the one hop local key based scheme.

Signcryption is a new cryptographic primitive that simultaneously fulfills both the functions of signature and encryption. The definition of

Generalized Signcryption

is proposed in the paper firstly. Generalized signcryption has a special feature that provides confidentiality or authenticity separately under the condition of specific inputs. Based on ECDSA, a signcryption scheme called ECGSC is designed. It will be equivalent to an AtE(

OTP

$

, MAC) encryption scheme or ECDSA when one of party is absent. A third party can verify the signcryption text publicly in the method of ECDSA. Security properties are proven based on Random Oracle mode: confidentiality (CUF-CPA), unforgeability (UF-CMA) and non-repudiation. Compared with the others, ECGSC presents a 78% reduction in computational cost for typical security parameters for high level security applications.

Neighbor Discovery (ND) protocol is very important in ubiquitous networks because it can provide IP auto-configuration and address resolution. However, a malicious user can make access router of local area network (LAN) generate useless ND protocol messages by sending it abnormal data packets with fictitious destination IP address. If a malicious user sends the access router the enormous volume of abnormal traffic, this may result in network congestion and degrade quality of service (QoS) not only for ND-requested normal traffic, but also for ND-free normal traffic. In this paper, we propose a scheme that is able to effectively control ND congestion by rate-limiting ND protocol messages generated by abnormal data packet. In our scheme, when an access router receives a ND-requested packet, it checks if the destination IP address of the packet exists actually on the target LAN. If yes, it sends out the ND message for the packet using good QoS in packet forwarding service. Otherwise, it uses bad QoS. To learn topology of the target LAN, the router monitors all traffic from the target LAN. Through simulation, we show that our scheme can guarantee not only QoS of ND-requested data traffic, but also QoS of ND-free data traffic irrespectively of the degree of attack strength.

The increased heterogeneity and dynamism of new computing paradigms and especially of ubiquitous computing models is boosting the need for auto-configurable systems. In these new scenarios, heterogeneity and dynamism are inherent properties and applications are built by aggregating distributed information and services that are not under the control of a single entity. The current trend towards distributed computing poses important problems related to the transmission of large amounts of data between the nodes of the computing system; the control over the information; and the flexibility to adapt to heterogeneous client requirements. These characteristics, difficult to manage by traditional computing models, are making the mobile agent paradigm to gain momentum and increasing the interest of researchers and industry in this paradigm. In this paper we present a solution to provide a secure and auto-configurable environment for mobile agents in ubiquitous computing scenarios, based on two main building blocks: trusted platforms and profiles.

Secure key establishment in wireless sensor networks has been shown to be efficient. For sake of computation power, pre- distribution of symmetric keys is recommended. In such schemes, a set

S

of keys is used and each deployed node knows a subset of

S

. However, the capture of nodes can quickly lead to the corruption of a significant part of the network and solutions like multi-deployment have been proposed. In multi-deployment schemes, sets of nodes – using different key sets – are sequentially deployed that limits the impact of the capture of nodes. In this paper, we consider the problem of connectivity of single deployment in regards to the size of key sets. We show that connectivity guaranteed solutions lead to network vulnerability when nodes are captured. Then we evaluate network robustness of different schemes.

The (

t

,

n

) threshold signature scheme can delegate the signing capability to all

n

participants in a group, and at least

t

participants can cooperatively sign a message on behalf of the group, where

t

≤

v

. For the group communication in the real society, the verification site also needs to be restricted in its associated access privileges. Therefore, this paper proposes a practical solution to the (

t

,

n

) threshold untraceable signature with (

k

,

l

) verification scheme, which requires that

k

out of

l

verifiers or more can verify the threshold signature on behalf of the verification group. Compared with the previous works, such as Wang et al.’s scheme and Chang et al.’s scheme, our proposed scheme is more practical and expansible. Our scheme allows each group to be both a signing party and a verification party, and the shadows of all group members are no need to be redistributed after the initialization has been finished. In addition, the share distribution center (SDC) is not required in our scheme.

The virtual region-based P2P architecture was provided in this paper, which extended the proximity concept in overlay network. By virtual region it meant various related factors should be integrated into the design of P2P topology, such as peer interest, peer locality, and heterogeneity in peer capacity etc. But, the shared contents in P2P are vast and miscellaneous, it is very difficult to recognize the interest similarity among peers, dynamically form the interest region and direct peer to appropriate interest region. In this paper, the Jensen-Shannon Divergence (JSD) was used to characterize interest/semantic similarity among peers, and the interest region formation and location mechanism based on Dynamic Interest Landmark (DIL) was offered to facilitate to dynamically guide peers to join the appropriate interest region corresponding to peers’ content semantic. Considering that interest region is composed of geographically sporadic peers, each interest region was organized according to peer proximity in underlying infrastructure. The theoretical and simulated results illustrated that our architecture can significantly reduce the query overhead, and improve the search performance.

The fairness of scheduling resources are important to improve the whole performance. In this paper, we study the economy-based approach,

i.e.

, portfolio selection, to realize the dynamic allocation of distributed and heterogeneous resources. The portfolio selection method emphasizes the mean-variance model, which can evaluate the final return and help the scheduler to adjust the allocation policy. We present the practical algorithms for network nodes and Bloom filter-based surveillance, which can support the efficient adjustment of a scheduler.

Most portal systems provide simple web services in existing web environment but cannot support semantic web environment. Moreover, current portal systems are not adequate to compose personalized services. We also need semantic portlet search methods using portlet ontology, which can support ubiquitous environment, as well as portlet API that supports users’ access to u-Portals through various types of terminal equipment. This paper constructs a user-centered customized web service u-portal system that supports semantic web service search and composition through a ubiquitous portal engine.

The video services of most Access Grid (AG) toolkits still only employ outdated Video Conferencing Tool to transmit real-time video, and hardly have efficacious way to mix the 3D graphics model of grid results and the live videos of AG session into one scene for grid users’ collaborations. This paper introduces the research work about augmented video services (AVS) of CGAG (ChinaGrid Access Grid) based on the AG framework. It shows up a reasonable way on providing augmented video services in an advanced Access Grid environment, and also a new approach to share grid resources in Grid applications. At the end of this paper, a typical sample is given to demonstrate the capabilities of this AVS, and the result proves that it yields superior quality compared to the traditional webpage of grid portal and usual interaction modals of AG.

Using peer-to-peer overlay network to provide video-on-demand service has been a promising solution due to its potential high scalability and low deployment cost. However, it remains a great challenge to construct an efficient overlay network for peer-to-peer video-on-demand systems owing to their inherent dynamicity caused by frequent VCR operations or joining/leaving operations. In this paper, we propose a

ring based overlay network

to handle this problem, called

RINDY

, in which each peer maintains a

gossip-ring

to explore appropriate data suppliers and several

skip-rings

with power law radius to assist the quick relocation of VCR operations. Our simulation results show that RINDY achieves better load balance in the control overhead than tree based overlay. As compared with the traditional client/server model, it saves more server bandwidth and achieves lower start-up latency when lots of users watch a same video simultaneously.

Despite a variety of service scenarios and prototype systems that adopt ubiquitous computing technology, sophisticated methodologies for assessing the level of ubiquitous service and systems are still very rare. Hence, we propose a multi-layered model to assess levels of ubiquitous computing services.

As multimedia contents are becoming widely used in ubiquitous computing environments among many application fields, e.g. educational content management, entertainment, and live surveillance, the demand of personalized access to these contents has increased dramatically. Delivering ubiquitous personalized multimedia services (UPMSs) is a challenging task, which relies on many different functions. In this work, we propose a three-layer software platform, called UPmP to support efficient development and deployment of UPMSs. It fulfills the core functions for UPMS including service management, multimedia recommendation, adaptation, and delivery. We adopt component-oriented approach in building the platform. Therefore the configurability of the platform is inherently achieved. A representation model is introduced to hierarchically organize components and describe meta-level information about components. We also present a visual configuration tool together with a XML-based language for the purpose of platform configuration. The experimental results show the UPmP is flexible to be configured under different settings, and the overheads are acceptable.

We present a variant of end-user programming targeting ubiquitous computing environments that allows non-technical users to create “programs” to customise their personal living spaces. Using this end-users do not need to write program code, or follow a rigid sequential list of actions in order to achieve results Rather they only need to show the system the required behaviour via physical interactions with the environment. Finally, we report on a user evaluation that indicates end-users find this approach to be a useful and enjoyable experience.

Although many users want to access their personal data from anywhere in ubiquitous computing environments, distributed storage systems with data replication have not yet been accepted widely. We point out that one current problem of existing systems is the difficulty of determining which data is important and should be replicated in mobile devices. We propose a new personal storage system that uses RDF for metadata descriptions and for selections of replication candidates. We present a basic design and a prototype implementation of the system and confirm it’s feasibility for large amount of accumulated data.

The emergence of information technology the last decade is leading to changes in the way people work and learn. Computer networks and especially ubiquitous intelligent mobile learning environments offer new strategies the dissemination of course work. Following the current trend towards ubiquitous computers, networks, information and services, a new innovative mobile e-learning intelligent world is created, where software components embedded with intelligence assist the users in education and provide functionalities such as context awareness, automation, adaptivity and proactiveness. Within this paper a case study of the similarity and differences between an object oriented solution and an agent-based solution of ubiquitous intelligent mobile learning environment is discussed. The results of this research are presented in terms of preliminary user tests for the design of the Mini-ME Interface and performance testing for both implemented architectures.

Ubiquitous power includes one important factor that is the coordination among small instances of individualized smart programs. With the coordination, intelligent strategies can be established to realize computing topology and to serve different purposes of computing goals. With the development of cellular phones and internet, these individualized programs exist in the digital world where a new ecosystem can be formed. These smart entities create groups and consume natural resources in the digital world. Competition and collaboration thus exists among different groups of autonomous entities. We propose a model, based on a concept called food web in ecosystem. Food web exists in natural words for minions of years and serves as a balancing algorithm among different types of species. We define the food web model for autonomous entities and propose algorithms for formation and communication. As a consequence, the proposed model can be used to consider the development of intelligent strategies and the underlying communication topology such that ubiquitous intelligences can be evolved on the digital food web.

Towards an ideal ubiquitous computing environment, M.Yokota et al have already presented the concept of Distributed Intelligent Robot Network (DIRN) consisting of one brain node and numerous sensor and actor nodes with human-friendly interfaces. In order for well-coordinated DIRNs, it is essential to develop a formal knowledge representation language such as is universal for any kinds of devices and enlightens them enough to communicate with ordinary people naturally and to compute about and act upon their environments appropriately. For this purpose, the author has developed a formal language

L

md

and been applying it to integrated multimedia understanding and various cross-media operations in simulation of DIRN-world interaction.

This paper proposes a noble RFID system architecture, named uRFID, to support application domain-independent semantic interoperability under ubiquitous environment. The objective of the uRFID is to let ubiquitous RFID applications independently use information of tags in a consistent way in a variety of application domains. Although many researches have been studying to support the RFID applications, they do not consider the aforementioned semantic inconsistency issue. The uRFID is based on the metadata registry to achieve the goal. It provides an infrastructure support for the semantic consistency and application domain-independent information utilization. As a result, it enables high quality RFID applications to be developed for ubiquitous environment.

To cope with the fact of high competitiveness in tourism industry, some travel agencies are adopting the strategy of providing relatively more abundant and higher quality services to their clients. Since prior research has found that lack of information is one of the main barriers for people to travel, provide location-based information to people visiting and staying at a new location should be useful and valuable. This research attempts to find out travelers’ attitude toward location-based service (LBS). “What factors will affect the travelers’ willingness to adopt this new service” is what we tried to find out in this research. Our findings shows that perceived usefulness, perceived ease of use, and security & privacy are associated with travelers’ attitude toward using LBS, but location information, context awareness, and device functions are not. Our findings can be referenced by tourism industry for the purpose of the design and development of successful business applications to catch the revolutionary opportunity and benefit of LBS.

This paper describes ongoing work aimed to explore how to augment a person’s work environment with information which enables her to feel the presence of intimate companions. The vehicle we deem to be appropriate for this situation is indirect communication (in some ways related to peripheral communication). The presence we intend is based on the activities of these intimate people. In short, with a certain periodicity the user is presented, on a peripheral user interface (windows desktop or digital picture frame), with a new collage composed of pictures indirectly triggered by their loved ones. In particular the pictures are triggered by the text they write and read, and the songs they listen to, while working on their PCs.

In order to adapt the teaching in accordance to individual students’ abilities in the distance learning environment, more research emphasis on constructing personalized courseware. The new version of SCORM 1.3 attempts to add the sequence concept into this course standard. The sequencing describes how the sequencing process is invoked, what occurs during the sequencing process and the potential outputs of the sequencing process. However, the related research of sequence trap is lack. Sequence trap results from improper sequence composing. The more complex course is the higher trap-probability arises. When the sequence trap occurs, it will block any learning activities and cannot go on any course object. As a result, we apply the valuable features of Petri net to decrease the complexity of the sequencing definition model in the SCORM 1.3 specification and process the input sequencing information to detect the sequencing trap in advance.

The present business environment is characterized by the use of groups, which work in distributed environments and have to deal with uncertainty and rapidly changing information. In this work we propose an architecture for a ubiquitous group decision support system able to support persons in group decision processes. The system considers the emotional factors of the intervenient participants, as well as the argumentation between them. Particular attention will be taken to one of components of this system: the multi-agent simulator, modelling the human participants, considering emotional characteristics, and that allowing the exchanges of hypothetic arguments among the participants.