Search

For network life time, load balancing and scalability, wireless sensor networks have to configure the network based on local management. Clustering algorithm is for satisfying above requirements. A cluster head selection of this algorithm requires to solve additional communication cost and unequal clustering compared with previous clustering algorithm. So, this paper proposes novel cluster head selection algorithm revised previous clustering algorithm.

In this paper, we propose a middleware based on the OSGi framework for home robots in Home-Network. The home-network services make us convenient and cozy. In addition, it improves quality of our life and includes a function which is to interoperate each device. In this paper, home robots are composed dynamically in the network and we can use a lot of services by the home robots.

For reducing railroad accidents, we developed monitoring system which is providing railroad information to a locomotive engineer. Especially, monitoring system apply to railroad crossing. In this paper, we proposed new architecture what based on web technology. This architecture will be made more efficient to the monitoring system.

Recently, one of the main emerging issues in the robotics area is the robotics middleware. We are developing robotics middleware for intelligent mobile robots. In this paper, we provide our architecture model and component interfaces. Our software architecture is concentrated on the intelligent services. Among many research issues in the robotics middleware, component model is an important factor. Component-based software engineering is able to guarantee not only reusability, maintainability, fault isolation, and so on, but also non-functional properties. With component-based software engineering, robotics middleware provides intelligent service model.

Wireless sensor networks are networks for collecting environment information in specific area using sensor nodes with low power. The process of sensor networks is a multiphase process compromised of sensing environment, transmitting the sensing data to neighbor nodes, collecting it from nodes for aggregating, and processing it for a request. As sensor networks have a feature which can have many similar data between adjacent nodes, they use a clustering method which sets up a cluster heads, collects many similar data from member nodes in a local cluster and sends the collected data to sink by cluster heads. A clustering method of sensor networks needs to equally distribute member nodes to a cluster head because of affecting the number of member nodes in a local cluster to the energy consumption of a cluster head, a local cluster and sensor network. So, this paper proposes how to set up a local cluster with optimal member nodes in wireless sensor networks.

Existing system architecture model cannot guarantee isolation of applications or system components from system faults that occur in system services. If an error causes the failure by operating system services, the entire system can be corrupted or components which are dependent that services are affected. Fault tolerant architecture model can be achieved by decoupling state information from the server. In this paper, we compare four kinds of architecture models. For constructing stateless server model, we use simple lightweight middleware. It can support fault tolerant architecture with low performance overhead.

The goal of wireless sensor networks is to collect sensing data on a specific region using wireless communication. The sink node gathers all local sensing data, processes them, and transmits them to users who use sensor networks. Generally, sensor nodes are low-cost, low power devices with limited sensing, computation and wireless communication capabilities. A sensor network applies to multi-hop communication on a large-scale network. As neighboring sensor nodes have similar data, clustering is an effective technique for ‘data-aggregation’. In the multi-hop based cluster formation technique, it is necessary that the number of cluster member nodes should be distributed equally due to the balance of cluster formation. Therefore, this paper proposes a method of cluster head selection to form equal cluster sizes in wireless sensor networks

The technologies for the digital broadcasting over a network are an emerging technology that provides a more function-rich, user interaction form of TV to the customer. This represents a change from low interactivity to moderate interactivity and high interactivity. The important thing is a IPTV set-top box(STB) which is a embedded system providing two-way communications using Internet Protocol over a network, and decoding the video streaming media. Recently, due to the rapid development on the hardware and software, it is available for not only video services but also data services such as web, e-mail or network game that is similar with a PC.However, these applications are often too heavy and complex. These are also subject to spamming, spoofing, content theft, and other hacking attacks, like many other IP-based solutions. The result of these can cause serious distress to the STB. In addition, because most of STB has its own RTOS, reusability and migration of various applications are restricted. With our increasing IPTV services, their reliability and security become more and more of an issue.Therefore, in this paper, we will propose a design of L4 microkernel based broadcasting middleware which supports the IPTV application platform and linux-like common environment.

During the past decade, we already experienced the wide spreading of mobile devices such as a mobile phone. In the first stage, most mobile phone has its own and fixed embedded applications, mobile phone users could not executed additional applications. To execute additional applications, the mobile virtual machines are introduced and embedded in most mobile phones. However, due to the restrictions of mobile device and commercial reasons, most mobile phone includes a single application execution platform. Therefore, mobile phone users only can run applications which are implemented in the same platform. To ensure reusability of applications and to reduce development costs of the applications, we designed L4 microkernel based lightweight mobile middleware which supports multiple application execution platforms in a single device. In this paper, we will review our design and implement a mobile middleware which support multiple virtual machines to the actual mobile device. By applying our mobile middleware, we expect that it can reduce porting effort of virtual machines and also offers efficient and flexible application execution environment in the mobile devices.

Applied fields that use a wireless sensor network is increasing gradually, and it's used to a fire watch monitoring system, a theft prevention system, and the environmental monitoring system that used image processing technology etc.By low power and low costs, this wireless communication and hardware miniaturization make it possibile in technical developments that proceeded are search regarding applicability in actual physical environments and use of a security camera is increasing rapidly so that a function of a security camera met by surrounding environment while we can be got image information.These paper implement a security monitoring system by using Zigbee Module(IEEE 802.15.4 Protocol) and CMOS Image Sensor. This security monitoring system make it possible to reduce inconvenience of loss of installation charge and control.

WSNs middleware is software positioned between applications and the operating system in sensor nodes. This middleware should have a communication function between the sensor nodes, to support good interactive functions between the user's applications. It must balance the resource use between the client and server system. WSNs comprise server-side middleware and in-network middleware. Server-side middleware supports application services efficiently in the server system. In-network middleware, located between sensor nodes, is in charge of managing the inner sensing information. We choose an event-driven network, the publish and subscribe network model, to support WSN. Using this model, sensor networks could reduce unnecessary energy consumption and respond to a variety of user requests.

An important issue in a wireless sensor network is to use a nodepsilas limited energy efficiently. The clustering technique can reduce the energy consumption of wireless sensor networks. The cluster head collects sensing data from sensor nodes neighbors, aggregates sensing data and sends the aggregated data to a sink node. Electing the cluster head is important to reduce the networks' energy consumption. This paper proposes adaptive cluster head selection that reorganizes the clustering, considering the position between a cluster head and the nodes in a cluster.

Reputation has recently received considerable attention within a number of disciplines such as distributed artificial intelligence, economics, evolutionary biology, and among others. Most papers about reputation provide an intuitive approach to reputation which appeals to common experiences without clarifying whether their use of reputation is similar or different from those used by others. DF provides a Yellow Pages service. Agents in the FIPA-compliant agent system can provide services to others, and store these services in the DF of the multiagent system. However, existing DF cannot detect the fake service which is registered by malicious agent. So,a user may search these fault services. In this paper, we analyze the DFpsilas problem and propose the solution. We describe the Reputation mechanism for searching these fake services. Reputation function assumes the presence of other agent who can provide ratings for other agents that are reflective of the performance or behavior of the corresponding agents.

Wireless sensor networks are attractive means to monitor environment and to utilize for military. Sensor networks are increasingly improving. Operating system and routing are especially researched in times past. On the other hand, the aspect of middleware that sits between the operating system and the application did not be sufficiently discussed. The importance in sensor networks is power saving. And middleware is located between operation system and application. Middleware in sensor networks is responsible to process data effectively to minimize spending energy of sensors. Therefore this paper proposes to effectively process data using data fusion technique.

Today, the railway is very important in modern society means of public transport. Many people use the railway each day. On the other hand, with regard to railway accidents every year due to damage caused many casualties. In this paper, we propose railroad crossing accident prevention system for monitoring the situation based on OSGi for safe railway operations, about the same railroad crossing a dangerous place. This system, communication between the train and railroad crossing. So the information from the train crossing while a train driver was monitored in real time with more information than can help prevent accidents on the crossing. Using the OSGi framework, can be used widely in many hardware.

The agent systems can register, deregister and search the information of agent and services. Amount of the agent systems and the information according to agent and service are existed. The differences of the information specification make not easy to the relevancy about them. Thus, we propose the agent service system (ASS) to find the relevancy among the agents information and to manage them. The ASS can provide the various services to the other multi-agent systems or general systems that may be not the agent platform.

Many casualties are produced by railway cross accidents every year. Since railway cross currently consists of traffic light and crossing gate, there are difficulties in preventing the accidents. To solve the problem, the system in which train entering railway cross can estimates in advance and handle the situation is emerging and many solution are proposed. However, locomotive engineers control the trains by own decision in the current systems. Therefore, accidents generated by locomotive engineers' error cannot be prevented. In this paper, the auto control system using OSGi and JESS is proposed. The proposed auto control system immediately deals with information of railway cross and train by using interactions of the context aware service components which are equipped on OSGi platform as bundles. Also, information management is easy due to the functions of OSGi. And state of railway cross is estimated more correctly based on collected information through JESS rule engine. If locomotive engineers determine the situation late, alarm message services are provided and trains can be controlled automatically. By using this system, we can deal with urgent situations at the railway cross and prevent railway cross accidents in advance.

In ubiquitous computing environments, the basic information such as types or locations of services should be provided in advance to use distributed services and devices efficiently. Using a combination of developed agent technology and service discovery to efficiently and flexibly manage distributed services, devices, and users is a trend. In this paper, we analyze the DF's searching method and designed ubiquitous location-based service in context-aware environment, because context-aware environment provides a suitable method to gather requested information in an appropriate way.

Although Web services include execution of distributed applications and technology based on service-oriented architecture (SOA), its research is not enough to provide the active services. Specially, from the viewpoint of the distributed domains (e.g. UDDI, Semantic) Web services still has considered different approaches such as QoS, performance, and service management, etc. The goal of this paper proposes the method supporting QoS for Web services in the distributed domains. The existing approaches are shown the research supporting the basic model on the Web services architecture or only one public service registry. In this paper, we propose the approach considered the models which can support QoS of Web service through case-study in multi-domain.

Mobile agent is important in mobile agent based distributed environment. The mobile agent performs a purpose through accesses to resources of the remote host by migrating from one host to another host or interactions between agents. Therefore communications between the mobile agents is important. However, there are communication failures problem and message forwarding errors due to the mobile agent's properties. Hence, communication protocol requires location-transparency, reliability, high-efficiency, asynchrony and self-adaptability and there are mechanisms to satisfy the requirements. In this paper, we propose mobile agent communication mechanism using Home Server and Sub Server. By using the mechanism, we can provide reliable message delivery services that solve the problems in the previous methods for communication failures and message forwarding.

Mobile IPv6 (MIPv6), which is proposed by IETF(Internet Engineering Task Force), will be a main mobility management protocol in the future IP-based wireless and mobile networks. But MIPv6 has a long handover delay relative to a link layer handover delay. As handover latency increases, mobile nodes (MN) experience performance degradation. It significantly impacts on the performance of real-time applications. Various protocols that are based on MIPv6 are proposed by IETF to reduce the handover latency. In particular, hierarchical MIPv6 (HMIPv6) is designed to reduce the handover latency by introducing micro and macro mobility management. In HMIPv6, binding update costs can be lessened when handover occurs in the intra domain area. In this paper, we propose the network-based handover scheme based on HMIPv6 that uses indication from link layer. Handover latency and packet losses during intra domain handover are significantly reduced compared to HMIPv6 by using the proposed scheme. We evaluate the proposed mechanism by simulating it over IEEE 802.16e in ns-2. The simulation results quantitatively show that the proposed scheme outperforms HMIPv6 in terms of handover latency and packet losses.

The Foundation for Intelligent Physical Agents (FIPA) is a non-profit organization aimed at producing standards for the interoperation of heterogeneous software agents. The FIPA Semantic Language (SL) is a communication language that uses the redefined 'content' parameter of the FIPA Agent Communication Language (ACL). The FIPA SL is also a most widely used standard language between the FIPA-based agent platforms. JADE, the famous agent platform in agent paradigms, also applies the FIPA SL. The agent platforms should follow the FIPA specifications such as FIPA ACL and FIPA SL. However there are no communication methods between the heterogeneous platforms that employ the different languages at present. For communication between the heterogeneous platforms that employ the different languages we should fix the source codes directly. Thus, we propose the FIPA SL-to-XML converting module. The converting module makes the FIPA SL-based agent platforms communicate with the XML-based web applications and platforms. We have the employed converting module in XML-based the Component -based Autonomic Layered Middleware (CALM). We also propose the starting point of the integration module that can combine the heterogeneous language-based platforms.

Solar cells use the natural sunlight for a variety of purposes including electrical power generation. In order to maximize the power obtained from a solar cell, it has to operate at what is known as the Maximum Power Point (MPP). Intelligent algorithms that can track MPP for a given cell are crucial for the success of a large scale deployment of solar panels. This paper investigates the performance of two well-known Maximum Power Point Tracking (MPPT) algorithms. A variable irradiance environment has been considered in this work, where sunlight varies significantly during the day. Towards the end, frequency stability analysis is conducted to evaluate the relevant impact of MPPT algorithms on frequency stability of system.

Multi-agent technologies for ubiquitous environment became recently attractive for many peoples. According to these interests, agent technologies have been researched by various researchers and laboratories. In this multi-agent environment, each of these agents has its own services and stores the service descriptions in the service repository of the multi-agent system. From this repository, the user agent can search the specific descriptions which the user wants to find. In this paper, we proposed the efficient service repository architecture which can improve the existing directory facilitator as service repository. An advantage of using this architecture is that searching for suitable services can be faster than the old one.

Medical institutions are developing devices for monitoring patients in hospitals, and also for remote healthcare, by which patients can be treated at their home. In this paper, we propose an improved context-aware healthcare system that has evolved by collaboration between the JADE and OSGi frameworks. The existing healthcare systems provide very passive services and suffer from incompatibility among systems, because of the differing standards between the systems. In this research, we explain why by applying the JADE to the healthcare system, we provide effective management and automated services. Also, our proposed healthcare system is not limited by platform or application program, because of the installation of the OSGi framework onto the surrogate system. To gain the advantages provided by different frameworks, we apply two frameworks to our healthcare system, and develop interoperability technology between the OSGi and JADE to share services among the frameworks.

Web service technology is based on service-oriented architecture (SOA) as new computing paradigm and an intelligent agent platform provides service-discovery, communication, and interaction between heterogeneous platforms as middleware. As these trends looked promising, we propose an advanced SOA model using Web service by the intelligent agent platform. We aim to support the interoperability of service-oriented and large scale domain through integration of two technologies in this paper.

Nowadays, the field of healthcare becomes more grew and sophisticated. The purpose of these u-healthcare technologies is to provide convenient healthcare service to both caregivers and patients, and to make it easy to diagnose user's health condition by caregiver. In this paper, we design a ubiquitous-healthcare system that can consider user's context information dynamically by rule engine.

This paper discusses a multi-agent system (MAS) based efficient healthcare system, which is comprised of medical sensor modules in conjunction with wireless communication technology. In particular, the proposed system supports a wide range of services including mobile telemedicine, patient monitoring, emergency management and information sharing between patients and doctors or among the healthcare workers involved. For this purpose, the proposed system provides not only a systematic computing environment between the BAN (body area network) and the subsystem within the hospital, but also participates in an agent network with the BAN's PDAs through peers. Information within the hospital is shared under a MAS environment. The tool used to develop the MAS is called JADE, a FIPA-compliant development framework. The ultimate goal of our system is to manage and monitor the patient's medical information and to support efficient medical service through the medical staff's cooperation.

The purpose of service discovery techniques is to minimize the cost of detecting services and provide users with convenience, even though various devices and services exist. For more dynamic and useful service discovery, middleware for context-aware service discovery is required. In this paper, a middleware system is designed and implemented, based on the agent platform for context-aware service discovery. When a service is detected, context information relating to the user and environment is used. As a policy-based system, our middleware does not only use context information, but also use predefined policy. In other words, user preference can be considered. Also, it has authentication module, so users having authority can only access the middleware. Near the conclusion of this paper, a hospital scenario is composed and implemented, by applying the proposed middleware solution.

The traditional server-centric design for P2P service discovery is not suitable for ad-hoc environments. In this paper, in order to facilitate the inter operation of ad-hoc environments with other platforms, we introduce D/sup 2/HT, an algorithm which enables directory federation using DHT, in order to support open scalability. We also introduce the lightweight service discovery (LSD) protocol based on the FIPA compatible agent framework. Our approach overcomes the reconfiguration problems associated with other DHT variants (e.g. Chord (Stoica et al., 2001), CAN (Ratnasamy, 2001) and Pastry (Rowstron and Druschel, 2001), which arise when peers join/leave the network, as well as taking into consideration device limitations. We also present an initial implementation of LSD using iPAQs and the D/sup 2/HT-enabled FIPA agent platform.

JXTA is a Peer-to-Peer application development infrastructure that enables developers to easily create service oriented software. This paper presents a low-cost, patient-friendly JXTA-based healthcare system, which is comprised of medical sensor modules in conjunction with wireless communication technology. In particular, the proposed system supports a wide range of services including mobile telemedicine, patient monitoring, emergency management and information sharing between patients and doctors or among the healthcare workers involved. For this purpose, the proposed system provides not only a systematic computing environment between the BAN (Body Area Network) and the subsystem within the hospital, but also participates in a JXTA network with the BAN's PDAs through peers called JXTA relays. Information within the hospital is shared under a P2P environment by means of JXTA grouping technology. The proposed system is shown to be adequate for a ubiquitous environment which offers effective service management as well as continuous, remote monitoring of the patient's status.

The Simple Object Access Protocol (SOAP) is a novel protocol which supports information exchange in a decentralized, distributed environment, which brings flexibility and extensibility to the communication and invocations among remote hosts. However, SOAP requires a high performance for collaboration such as service delivery and invocation. Furthermore, SOAP itself is not well suitable for mobile devices due to network bandwidth or resource-constraints problems. In this paper, we describe SOAP-based remote data access scheme which is comprised of XML and RPC technology. Especially, we focus on the scheme which is optimal over service discovery and invocation based on SOAP technology. Also, we apply the proposed scheme to LSD (Lightweight Service Discovery) which is a peer-to-peer, cache-based service discovery protocol for ad hoc environments.

In mobile communications, to overcome the highly nonlinear nature of time delay estimation, an iterative nonlinear filtering algorithm, called the scaled unscented filter (SUF) has recently been proposed in the field of nonlinear filter theory. Furthermore a newly introduced method, called simplex unscented filter (SimplexUF) uses significantly fewer points than other unscented filter methods previously published, while showing comparable performance (accuracy) of mean and covariance to SUF. In this paper we apply a SimplexUF-based algorithm for the estimation of closely-spaced path delays and related coefficients in CDMA environments. The simulation results also show that this filter performs as good as SUF. We also show that it provides a more viable means for tracking error for time-varying amplitudes and delays in CDMA communication systems than SUF

Service discovery protocol is the main element that determines the efficiency in middleware platform of ubiquitous networks. Recently, a large number of middlewares supporting scalability which focus on the distributed computing and data sharing among tremendous peers has been proposed. However, due to the distributed nature of ad-hoc networks, peers may not be able to find other peers and corresponding resources persistently. Furthermore, current service discovery engines do not provide open scalability that can make them to interoperate with each other. In this paper, we propose a simple mechanism, which provides cross-platform interoperability through directory federation combined with DHT mechanism to support open scalability and lightweight service discovery in ad-hoc network based on FIPA compatibility agent framework.

Traditional service discovery mechanisms propose efficient and fast searching schemes for service discovery. And their schemes do not consider interoperability among discovery protocols. In this paper, we design an agent platform (AP) modifying FIPA-OS (Foundation for Intelligent Physical Agents Open Source) based on FIPA specifications. FIPA (Foundation for Intelligent Physical Agents) is a nonprofit organization aimed at producing standards for the interoperation of heterogeneous software agents. Until now FIPA specifications have not been considering mobile ad hoc networks. In such environments, agents can 7 find services offered by other service discovery engines in heterogeneous networks. We can guarantee scalability, by adding DM (discovery middleware) modules in the AP. The DM helps agents find and use services exploiting existing service discoveries under ad hoc networks. We chose UPnP (universal plug and play) and LSD (lightweight service discovery protocol) as DMs among various service discovery technologies. UPnP is a set of protocols for service discovery under development by the UPnP forum and LSD is a new discovery protocol for mobile ad hoc networks emphasizing on cache management that we designed. On our AP, we can use the services which devices of UPnP and LSD provide.

In wireless communications, a joint channel coefficient and time-delay tracking technique are a critical issue. Due to the highly nonlinear nature of time delay estimation, particle filter (PF) and sigma point particle filter (SPPF) can be employed. The SPPF algorithm consists of a particle filter that uses a sigma point Kalman filter (SPKF) to generate the important proposal distribution. The SPKF allows the particle filter to incorporate the latest observations into a prior updating routine. In addition, the SPKF generates proposal distributions that match the true posterior more closely. We propose an SPPF based on algorithm for the estimation of closely-spaced path delays and related channel coefficients in CDMA environments. We show that the parameter estimation using this filter structure is very effective even in the non-orthogonality of spreading codes and under imperfect power control in the CDMA environments. The simulation results show that this filter outperforms the PF including conventional Kalman filters.

The integration of CDMA cellular network and wireless LAN (WLAN) has drawn much attention recently. In this integrated architecture, it is required to support a vertical handoff from the WLAN to the CDMA system when a user moves out of the WLAN coverage area and vice versa. We propose a context based handoff procedure and the corresponding mechanism from WLAN to CDMA system, and vice versa, based on wireless channel assignment. In other words, this paper focuses on the handoff decision which uses context information such as dropping probability, blocking probability, GOS, the number of handoff attempts and velocity. As a decision criterion, velocity threshold is determined to optimize the system performance. The optimal velocity threshold is adjusted to assign available channels to the mobile stations with various handoff strategies. The proposed scheme is validated using computer simulation. Also, the overflow traffic with take-back technique is evaluated and compared with non-overflow traffics in terms of GOS.

In this paper, we propose SDRP: a new service discovery protocol combined with routing policies in MANETs. It performs the protocol acting upon a backbone and access network as well as any kind of distributed networks. We describe a service by a unique ID number and use a group-cast routing policy in advertising and requesting. In addition, the nodes in the reply path also cache the advertisement information. We compare SDRP with Flood and AODV in terms of overload, demand throughput and average delay. Simulation results show SDRP can save response time and accommodate even high mobility network environments

Service discovery middleware allows users to find and use services through service discovery protocols without previous knowledge of the locations or characteristics of the services with minimum manual efforts in heterogeneous and ubiquitous environments. For this reason, many researchers have carried out studies related to service discovery middleware and many papers dealing with this field have been published. However, when a number of service consumers request services from middle agents (e.g. matchmaker, broker, yellow page, blackboard, etc.) within the service discovery middleware, the middle agent does not guarantee efficient and rapid matching results because they only use a simple matching algorithm. In this paper, we address the issues of existing matching algorithms, and then propose a new matchmaking algorithm based on the marriage matching algorithm of ATM networks (Gusfield and Irving, 1989) to improve middle agents' performance, complementing shortcomings of existing matching algorithms. We also add priority based matching to the new algorithms. Through this priority, important service request messages are processed faster than request messages that have low priorities. We expect the proposed matchmaking mechanism to increase users' satisfaction in dynamic environments.

The integration of WLANs and CDMA networks has recently evolved into hot issue. We propose a vertical handoff algorithm between WLAN and CDMA network. In this algorithm, a vertical handoff is classified into handoff decision making and MIP (mobile IP) registration process. Then, a handoff initiation is decided by the received signal strength (RSS). To reduce the unnecessary handoff probability, we consider a hand-off initiation model based on the criteria of RSS. As a mechanism of network selection, we also propose a context based network selection and the corresponding algorithms between WLAN to CDMA network, based on wireless channel assignment. We focus on a handoff initiation criteria which uses both the RSS and the network selection method which uses context information such as dropping probability, blocking probability, GOS (grade of service), velocity. As a decision making criterion, velocity threshold is determined to optimize the system performance. The optimal velocity threshold is adjusted to assign available channels to the mobile stations. The proposed scheme is validated by computer simulation.

As the number of Internet services grows, it is becoming more important for the network users to be able to locate and utilize those services that are of interest to them. As a result, a service discovery protocols are coming to play an increasingly important role in highly dynamic networks. In this paper, we describe a new architecture for service discovery called LSD (Lightweight Service Discovery). The LSD is a peer-to-peer cache-based service discovery protocol for ad hoc environments. Through the LSD protocol, small handheld mobile devices with wireless connectivity can detect each other in their neighboring devices, and perceive whether particular shared services are available or not. In addition, the LSD solves the overhead problems which arise when traditional discovery protocols are applied to mobile ad hoc networks.

The integration of WLANs and CDMA networks has recently evolved into a very hot issue. In order to support a vertical handoff, we propose a context based vertical handoff decision process and the corresponding algorithms from WLAN to CDMA system, and vice versa, based on wireless channel assignment. We focuses on the handoff decision which uses context information such as dropping probability, blocking probability, GoS (Grade of Service), the number of handoff attempts and velocity. As a decision criterion, velocity threshold is determined to optimize the system performance. The optimal velocity threshold is adjusted to assign available channels to the mobile stations with various handoff strategies. The proposed scheme is validated using computer simulation. Also, the overflow traffic (a vertical handoff) is evaluated and compared with non-overflow traffics in terms of GoS.

Service discovery is an important technology for ad hoc communications and collaboration in ubiquitous computing environments. However, such an environment will be vastly different from the traditional infrastructure-based environment. For this reason, we design a new discovery protocol for mobile ad hoc networks. LSD (Lightweight Service Discovery) is a peer-to-peer cache-based service discovery protocol for ad hoc environments. Through the LSD protocol, small handheld mobile devices with wireless connectivity can detect each other in their neighboring devices and also perceive that shared services are available or unavailable. Besides, LSD solves the overhead problems when traditional discovery protocols are applied to the mobile ad hoc networks. Finally, we show an implementation of LSD and do a tentative protocol test for our system using LG IBM Thinkpad and WinCE based iPAQ's

The integration of WLANs and CDMA networks has recently evolved into hot issue. We propose a vertical handoff algorithm between WLAN and CDMA network. A handoff initiation is decided by the received signal strength (RSS). To reduce the unnecessary handoff probability, we also consider a distance. As a mechanism of network selection, we propose a context based network selection algorithms between WLAN to CDMA network, based on wireless channel assignment such as dropping probability, blocking probability, GoS (Grade of Service), the number of handoff attempts. As a decision making criterion, velocity threshold is determined to optimize the system performance. The optimal velocity threshold is adjusted to assign available channels to the mobile stations. The proposed scheme is validated by computer simulation.

Wireless LANs (WLANs) are one of the fastest growing wireless access technologies. Unfortunately, since they were developed to closely match existing wired LANs, the popular IEEE 802.11 standards have several problems in providing Quality of Service (QoS) to stations. The IEEE 802.11e standard is being developed to overcome these drawbacks. One of most important function in 802.11e is the contention-based channel access mechanism called enhanced distributed coordination function (EDCF), which provides a priority scheme by differentiating the inter-frame space and the initial window size. In this paper, we first consider the enhanced DCF access method of IEEE 802.11e. We then introduced analytical model, which can be used to calculate the traffic priority and throughput corresponding to the configuration of multiple DCF contention parameters under saturation throughput. This enhanced DCF access method differs from the standard backoff schemes in that the backoff time is not uniformly chosen in the contention window interval that results in reduced collision probability under high loads. Numerical analysis shows that saturation throughput of the proposed scheme outperform the earlier approach.

P2P (Peer-to-Peer) applications allow flexible organization, distribution of role to participating peers and the ability to share information with the other peers for group collaborations. As a result, P2P systems are not only gaining importance, but also becoming ubiquitous media for information exchange. JXTA is a P2P application development infrastructure that enables developers to easily create service-oriented software. This paper presents a low-cost, patient-friendly JXTA-based healthcare system, which is comprised of medical sensor modules in conjunction with wireless communication technology. Among the most important activities associated with the healthcare system are the sharing of medical information and collaborative medical work. The JXTA grouping service facilitates to share data in a secure manner under P2P environments. Through JXTA grouping service, we implemented a prototyped system, which improves the ability of the healthcare workers to cope with dynamic situation, which in turn makes it possible to offer more efficient medical services.

As the number of Internet services is growing rapidly, the network users need a framework to locate and utilize those services that are available in the Internet. As a result, dynamic service discovery plays an important role in certain networks, such as P2P networks. Recently many P2P mechanisms have been proposed which focus on the naming and discovery protocols. However, because these mechanisms do not provide congestion management and priority-based scheduling in P2P networks, they cannot provide an efficient communication among the nodes. Based on these observations, in this paper, we propose a QoS-based agent framework, designed to resolve such problems, and demonstrate its effectiveness.