Recent Trends in Wireless and Mobile Networks

Mesh routers in wireless mesh networks (WMNs) utilize multiple radio interfaces in order to improve the performance of wireless network, resulting in transmitting multiple packets simultaneously on different orthogonal channels without interference. For the purpose of achieving the goal, routing protocols should be designed to improve the utilization of the network resources when selecting routes between source and destination nodes. However, single path routing protocols cannot explore the path providing sufficient bandwidth if the source traffic requires more bandwidth than the available bandwidth of the route. In this paper, we therefore propose a novel multipath routing algorithm which can explore multiple paths with sufficient bandwidth and transmit packets along disjointed paths simultaneously. we also introduce the path metrics which can estimate the available bandwidth of the path, considering the interface utilization information of each node, the channel diversity of route, and the intra-flow of the traffic. Using NS-2 simulations, we observe that our proposed multi-path routing protocol contributes to the performance improvements in terms of the end-to-end delay and the throughput in WMNs.

Third Generation (3G) cellular network technology allows the transmission of information and voice at higher data rates. 3G security model achieves confidentiality and integrity by using KASUMI block cipher algorithm as a standard. However, a number of serious weaknesses in the KASUMI cipher have been identified. For this reason, users may need other encryption applications. If phone user wants to be sure that the data transmitted can only be decrypted by the intended receiver, the user should use an encryption application. This work looks for the answers of the question if it is possible to make efficient and stable encrypted video calls and this paper presents the performance results of RSA encryption of a video call on mobile internet devices. To measure the process, an application is developed which captures live images from webcam then encrypts the images with RSA and then sends it over TCP-IP protocol to receiver.

Educational and Business MANETs are a different class of networks that are instantiated for the cause and get abandoned as the task is over. Such for-the-purpose networks have a limited span and may occasionally be interfacing with any surrounding networks. The proposed protocol, Energy Efficient Integrated Routing Protocol (E

2

IRP), is a layer 2 enhancement that suggests an upgradation to IEEE 802.11 to support service oriented routing for minimizing the energy consumption over the network. Enhancing layer 2 for these tasks relieves higher layers to implement other features like security and application specific control structures. The protocol can perform Unicasting, Multicasting and Broadcasting in a seamlessly integrated fashion. E

2

IRP does not require any delivery infrastructure and has provision for non-participating nodes to sleep. Thus the protocol is expected to perform well in terms of bandwidth utilization and energy efficiency for a low set-up time.

Sending a single video stream over multiple TCP-Friendly Rate Control (TFRC) connections is a promising lightweight way of coping with wireless channel losses

and

traffic congestion in a concatenated network (one consisting of a broadband wireless link and a wired all-IP network). Multi-connections introduce the need for data re-ordering at the receiver. This paper considers the potential delay at an uplink destination on the Internet. It also considers the impact on video quality of packet drops due both to channel loss and router buffer overflow, when the TFRC congestion controller is applied. Results for an IEEE 802.16e (mobile WiMAX) link show a worthwhile gain in video quality from using three or more connections over a single connection but with start-up delay n the multi-connection case due to the need to avoid possible buffer underflow.

In a wireless network, all information about users must be stored in one or more databases. Since a user location in cellular networks is not fixed and can change along the time, the information of a user that moves to a new location must be updated. Besides, when user x wants to communicate with user y, the location of user y must be extracted from the relevant database. Therefore, the database must be accessed for updating, recording, deleting, and searching information of a user. Thus, the most important criterion of an algorithm is to have a small database access time. In this paper, we compare the number of database accesses required for updating, deleting, and searching under different approaches that have been proposed in wireless networks.

Widespread usage of broadband Internet connections has allowed the birth of a new threat against service providers and subscribers as well. Botnets are vast networks of compromised hosts under the control of single masters who possess the ability to launch crippling denial of service attacks, send vast quantities of unsolicited e-mail messages and infect thousands of vulnerable systems with privacy-violating spyware and other forms of malicious software. Our goal is to propose a distributed architecture and introduce novel algorithms for malicious (potential botnet) activity recognition based on network traffic statistics generated by NetFlow. Scalability and robustness were the main principles during the design of the architecture. In this paper, we demonstrate that we are able to reduce the number of NetFlow records significantly with an own aggregation scheme. Furthermore, we are able to detect botnet participant computers (zombies) with the help of aggregated samples originating from various local networks, while the algorithms provide utmost anonymity to network operators.

This study proposes cognitive methods for managing the multimedia packet traffic of 802.16e system providing the dynamic spectrum sharing by TDM & OFDMA under heavy traffic conditions. The cognitive techniques are used in order to maximize the throughput of a system having different signal shapes such as video, voice or data packets. All real time voice and video packets transmissions have been provided with full success where the loss rate of data packets has been minimized using the cognitive methods. By these methods buffers are managed successfully for the given parameters, optimum buffer sizes have been formulized and it’s shown that we brought the throughput up to its calculated theoretical limits for multi-hop mobile wimax (802.16e).

Wireless mobile Ad-Hoc networks (MANETs) are networks of mobile nodes with no fixed infrastructures. Their peer-to-peer communication architectures over shared wireless medium with no clear line of defense pose a number of challenges to security design. Some previous works have used a form of digital signature for protection. However, there are still limitations in the implementation. This paper proposes a variant of the Merkle Signature Scheme (MSS) to protect an Ad-Hoc On-demand Distance Vector (AODV) protocol, a common routing protocol for MANETs. The new protocol is named Extra Secure AODV or ESAODV in short. ESAODV provides hop-count protection, digital signature and protocol enforcement. Through these features several security benefits on the routing packets have been established, namely fast authentication, integrity, and non-repudiation. Several experiments have been conducted using different digital signatures of the MSS variant signatures. The result shows that ESAODV is effective for combating routing attacks, and works properly although the malicious are 20% of the nodes total number in MANETs. In terms of features, ESAODV are also far superior than secure AODV (SAODV) and Authenticated Routing for Ad hoc Networks (ARAN).

Cognitive radios (CRs) are proposed as a solution to inefficient utilization of the overcrowding spectrum and opening space for the new wireless communication technologies. CRs have the ability to monitor the spectrum activities and decide on the best transmission channel depending on the measurements; thereby provide agility in spectrum access. However, new abilities of CRs are exposed to new security threats in addition to the existing security threats experienced in wireless networks. In this paper, we briefly evaluate these new security threats and provide cryptographic and non-cryptographic countermeasures to them. We also present the proposed security solution in [1] and provide the security evaluation of this proposed architecture. Moreover, we propose a security architecture to mitigate the threats to spectrum sensing and decision making accordingly.

In this paper we present a robust and dynamically load bal- anced clustering scheme (DRLC) for wireless sensor networks. Firstly, we introduce a simple but effective dynamic and robust cluster head elec- tion protocol that makes decision based on the nodes’ energy level and the number of neighbors. Since re-clustering is heavy and costly process, providing survivability of clusters in case of the cluster head (CH) failure is an important issue. To overcome this problem, we introduce a dynamic backup CH selection method that doesn’t need global topology knowl- edge or extra nodes in the network. In case of CH failures this backup CH undertakes the CH’s role and recovers cluster immediately. Secondly, we introduce a dynamic load balancing method to prevent overloading of CHs in network. When a CH’s load increases, the CH sends a message to inform its members about the load so that some of these members may change their CHs. This decreases the loads of overloaded CHs and increases the lifetime of that cluster. The experimental results verify the achievement of our model.

Wireless sensor networks are expected to operate in an unattended manner for long periods of time. As a result, they should be able to tolerate faults and maintain a reasonable performance level. Therefore, we propose two fault tolerant clustered De Bruijn based multi layered architectures, a fault tolerant routing scheme and a distributed fault diagnosis algorithm. The performance of the proposed work was analyzed according to the end-to-end delay and the data success rate. Also, the performance was compared to that of mesh networks. Our simulation results show that De Bruijn based networks perform better in both fault free and faulty situations.

The demand for Voice over IP (VoIP) services over wireless local area networks (WLANs) and Quality of Service (QoS) support in WLANs have grown substantially in recent years. IEEE 802.11e QoS enhancement standard has an important role on this expansion as well. It provides a suitable solution for delay sensitive real-time multimedia applications. Under the circumstance of any deterioration of wireless link conditions, IEEE 802.11 PHY/MAC specifications allow mobile nodes to select a proper transmission rate to optimize bit transmission. However, if a node selects lower transmission rate, packet losses and delays increase due to multi-rate effect. Therefore, all active calls have a general degradation. In this paper, we propose a fast, efficient, adaptive, proactive codec and frame size modification algorithm for the optimization of maximum throughput and voice quality on multi-rate WLANs.

Fourth generation networks are the result of the convergence of the different existing technologies, taking benefit from their complementary properties. In this context, recent research has addressed Digital Video Broadcast (DVB) networks, and has tried to couple them with complementary networks. In our work, we couple DVB-SH with a 3G network, thus providing a bidirectional interactivity path. DVB-SH benefits from a very high bandwidth capacity that allows unidirectional IP-TV channels broadcast. A residual bandwidth in the DVB-SH path may still be available because of the variable bit rates of served flows. In our work, we focus on this residual bandwidth. We realize an efficient switching of some 3G popular services, to the residual bandwidth of DVB Networks. The goal is to provide interactive low cost services over DVB networks.

In this paper, we propose a new algorithm to efficiently schedule our 3G switched services over DVB residual bandwidth.

One of the most fundamental management aspects of the Future Internet is the representation of management and operational data. The vast majority of languages and data structures used by network device manufacturers, such as SNMP-based designs and Command Line Interfaces, are data-oriented and are not conducive to representing semantic knowledge. Furthermore, such languages have no ability to represent business concepts, such as a Service Level Agreement, or higher-level concepts, such as the ability to maximize revenue for all users. To solve this problem, we draw inspiration from semantic routing, which traditionally has been used to connect users and applications with desired content and services based on intent, meaning, and other semantic qualities. We define a type of semantic routing that can be used for both semantic querying and network management tasks, and use it to create a semantic overlay network that enables routing to be done on the meaning of the packets. This can be used to provide valuable insight into how best to implement personalized and context-aware services, as well as choose between multiple reconfiguration alternatives for a given scenario.

Several strategies have been proposed for routing in the Low Earth Orbit (LEO) satellite networks. The multi-layered routing approaches are envisioned as promising because they use Middle Earth Orbit (MEO) satellite to extend the LEO satellite network’s communication capabilities. The previously proposed multi-layered routing approaches, however, still assume that the satellites in the same layer share similar characteristics. This assumption is not true in the future satellite networks. This is because the satellites in the future will be heterogeneous with various computation, communication and power capacities that lead to more complicated route construction challenges. In order to solve this problem, we propose the usage of cross-layer designs that can collect information from the neighboring satellites and evaluate their capacity during route construction and maintenance phases. This paper first analyzes the advantages and disadvantages of different satellite routing approaches. Then a multi-layered routing scheme called Mirrored Routing with Cross-layer optimization (MRCL) is introduced. In order to reduce overhead caused by the routing scheme, a hop count limitation, instead of a strict grouping policy, is used to direct packets to the MEO layer. According to our simulations, the end-to-end delay can be reduced 15% when a proper hop count limitation is selected. The novel routing scheme also significantly reduces the packet loss and the routing overhead (in terms of bytes of routing information) compared to the routing with out cross-layer optimization and hop limitation. We also simulate and careful investigate the performance of MRCL using various hop count limitation configuration.

In multi-hop cognitive radio networks, the communication links among the cognitive nodes fail easily with the dynamic appearances of the licensed users. We propose an approach to compute the weight of the available channels intersections. The cognitive nodes compare the weights of different routing paths and update the route-table to reduce the re-routing times caused by the network changing, in order to reduce packet loss rate in cognitive radio networks. Simulation results show that in a multi-hop cognitive radio network with frequent change of the licensed users and uneven distribution of the available spectrum, our protocol provides less packet loss rate in the communication channels and proper overhead in the control channel.

Mobile ad hoc networks (MANETs) are characterized by high dynamics in particular with respect to the formation of network partitions. The presence of unconnected partitions makes the deployed services inaccessible to some network participants. Service replication is employed as an approach to overcome this problem and to ensure higher service availability. Several protocols and algorithms for service replication in MANETs have been proposed. Most of these approaches apply topological analysis (like partitioning prediction) schemes to produce the required replication decisions. One approach, SDP, the Service Distribution Protocol, bases its decisions on analyzing interest in the service, an application layer concept. In this paper, we compare the performance of approaches based on these two criteria. First, we analyze protocols based on topology prediction and choose two typical representatives of this category, namely PSRP and SSRP. We then compare SDP as the only candidate using service interest to these two approaches. This comparison is based on an extensive set of simulation runs which are discussed in detail in the paper.

With recent advances in the development of wireless communication networks, Vehicular Ad hoc Networks (VANETs) have been receiving considerable research interest. One of the major applications of VANETs is Intelligent Transportation Systems (ITS). To exchange and distribute messages, geocast protocols have been proposed for ITS. Almost all of these protocols evaluate network performance level, instead of evaluating the protocol impact on the vehicular system. Nowadays, many drivers are becoming increasingly concerned with rising fuel cost. Therefore, it is desirable to create new “economical” geocast (EG) protocols. The main goals of this paper are to motivate communications researchers to design EG protocols, demonstrate the ability to integrate fuel consumption models with VANETs, and illustrate the necessity of transmitting information to vehicles in order for drivers to choose the economical path. Simulation results demonstrate that significant amounts of fuel will be saved if such an EG protocol is used.

Realtime resource sharing between operators is an efficient approach to deal with the ever increasing wireless traffic. Simple and closed form formulas relating cooperation terms to the network performance measures are needed, so that the operators can take cooperate/not cooperate decisions, and track the benefits/losses of sharing. Analytical solutions to resource sharing the among access networks of a single operator exist. However the fact that operators will not share network internal data calls for solutions that are separable from each other. In this paper we provide such a closed form formulation, validate it with simulations and propose a simple negotiation mechanism between two operators utilizing this model. This simple model can be extended to model more complex interactions between operators, interaction between more than two operators, or can be used to evaluate long term cooperation policies.

Smart or adaptive antenna promises to provide significant increases in system capacity and performance in wireless systems. In this paper, a simplified, near-optimum array receiver is proposed, which is based on the angular gain of the spatial filter. This detection is then analyzed by calculating the exact error probability. When there is spread of the Direction of Arrival (DoA) of the incident radio waves, the proposed detector significantly gains over an adaptive antenna array (which forms a weighted sum of the received antenna signals). Also, for this adaptive array, a simple way to account the multi-access interference can be exploited to evaluate the average probability of error when the users are randomly distributed within an angular sector. The proposed model confirms the benefits of adaptive antennas in reducing the overall interference level (intercell/intracell) and to find an accurate approximation of the error probability. We extend the method that has been proposed for propagation over Nakagami-m fading channels, the model shows good agreements with simulation results.

We propose an algorithm to compute zone-disjoint multi-paths for mobile ad hoc networks and compare its performance with that of the single-path minimum-hop routing algorithm as well as with the node-disjoint multi-path routing algorithm proposed in our earlier work. Simulation results indicate that the number of zone-disjoint paths per multi-path set can be at most 2, which is far lower than the number of node-disjoint paths per multi-path set. Also, the time between zone-disjoint multi-path discoveries is far lower than the time between node-disjoint multi-path route discoveries and can be at most 10% more than the time between single minimum-hop path route discoveries. However, there is no appreciable difference in the average hop counts per zone-disjoint multi-path set and node-disjoint multi-path set and they can be only at most 12% more than the average minimum hop count determined using single-path routing.

High-speed Internet provides unprecedented opportunities in the information domain. The wealth of information in the web-based knowledge banks brings about innumerable business prospects. This relentless rush for knowledge lowers the information ethics. The rewards had brought a stream of successful dot-com corporations in the 1990s. The dangers have also brought bankruptcy for most of these corporations. The plunderers are the knowledgeable and skilled few and the victims are the ignorant who are not able to see through the games that Internet exploiters can play. The boom and bust of the dot-com opportunists is grim reminder of how knowledge cycle has taken the ignorant few to the depth of despair. In this paper, we suggest that intelligent knowledge processing software should curtail the abuse of the IP networks, much as the network security programs that block the spread of spam and viruses.

The interconnection of Mobile ad hoc networks (MANETs) and the Internet is supported by a Gateway. The gateway is responsible for informing about some configuration parameters as well as for facilitating the creation of the routes to the Internet in the MANET nodes. For these tasks, several control messages are generated. The way in which these messages are originated differentiates the integration supports for MANETs. In particular, in the hybrid Global Connectivity support the Gateway generates periodic Modified Router Advertisements (MRA) which are broadcast in an area close to the Gateway. The optimum values to define the periodicity of these messages and the area in which they are propagated depend on the network conditions. Therefore, an automatic and dynamic algorithm is recommended to be implemented in the Gateway to adjust these two parameters. In this sense, this paper presents a technique by which the interval of emission of the MRA messages is controlled by a fuzzy system. The fuzzy system captures several network conditions such as the link stability or the number of sources. The simulation results show that the proposed scheme outperforms other adaptive approaches for the gateway discovery in MANETs.

There are many restrictions for Wireless Sensor Networks that enforce us to use low power batteries as their source of energy. Moreover since we often use these networks in rough and inaccessible environments, normally there is low possibility to change or recharge dead nodes. Today, Dynamic Power Management approaches with purpose of reduction of energy consumption in sensor nodes, after deployment and designing of the network draw attentions of many research studies. Therefore, there was a strong interest to use intelligent and capable tools such as Neural Networks in recent years, due to their simple parallel distributed computation, distributed storage, data robustness, auto-classification of sensor readings, dimensionality reduction and sensor data prediction obtained simply from the outputs of the neural-networks algorithms which lead to lower communication costs and energy conservation. This paper aims to present the most important applications of neural networks in reduction of energy consumption of WSNs, up to now.

We propose a dynamic, distributed tree based tracking algorithm for very fast moving targets in wireless sensor networks, with speeds much higher than reported in literature. The aim of our algorithm is to decrease the miss ratio and the energy consumption while tracking objects that move in high speeds. In order to do this, the root node which is determined dynamically in accordance with the node’s distance to the target, forms lookahead spanning trees along the predicted direction of the target. As the miss ratio decreases, the usage of recovery mechanisms which are employed to detect a target again that is moving away from the predicted trajectory also decreases. This decrease reduces the energy consumption and increases the network lifetime. We describe all the phases of the algorithm in detail and show by simulations that the proposed algorithm performs well to track very fast moving targets. We also compare the algorithm with the generic cluster, generic tree and dynamic multi cluster based tracking algorithms in terms of miss ratio and energy consumption.

Routing in Wireless Sensor Networks contains challenges, including limited energy constraints, network density, wireless channel errors. Different approaches exist in literature to overcome these challenges, such as data centric, location based and hierarchical routing. Most routing protocols in Wireless Sensor Networks are dealing with energy efficiency and network lifetime. In this paper, we present a shortest path routing algorithm based on Chandy-Misra’s distributed shortest path algorithm regarding both node weight and edge weight. X percent of edge’s weight and (100 - X) percent of node’s weight form a total cost between neighbor and source node which is used in order to generate the shortest paths and construct a spanning tree. Variation of X percent, node weight and edge weight provide resilience for shaping needed paths and change the spanning tree’s structure. When at least one node is close to critical energy level or a fault occurs, the routing algorithm is re-executed and new paths are generated. In order to obtain energy efficient paths, high network lifetime and finding out the overheads, we analyze the simulation results by assigning the battery level to node weight, communication cost to edge weight and %10, %30, %60 and %80 to X separately.

As the demand for uninterrupted Internet access grows, the popularity of wireless communication increases. However, wireless communication has some problems compared to conventional wired communication. Especially, if widely used Wireless Local Area Network (WLAN) applications are taken into consideration, it becomes an important issue to balance the load among available access points. It is impossible to balance the load when wireless hosts associate with an access point by using the classical approach of Received Signal Strength Index (RSSI). Some solutions containing a central server, requiring a specific brand of access point or protocol revisions have been proposed previously, but none of them has been favored as a generally accepted solution. In this study, a proposal which is central server free and requires no modifications to the existing infrastructure is presented. The proposed model is based on a dynamic determination of the least loaded access point to associate with, in order to balance load and maximize throughput.

High reliability and low power consumption are the critical objectives in wireless networks and the network topology is an effective issue in these objectives. This paper investigates these two objectives in the wireless multi-hop networks simultaneously. For this purpose, a connected dominating set CDS-based topology control approach is proposed. In this approach a distributed topology control algorithm with different power adjustment measures is suggested. Our goal is to self-organize this network with minimum interference and power consumption subject to connectivity preservation. Unlike many reliability enhancement algorithms, the proposed mechanism does not compromise power consumption. Experimental results show that the proposed topology control mechanism improves lifetime, reliability as compared to the traditional CDS-based mechanism significantly.

The next generation wireless networks are expected to support different applications within a heterogeneous network environment. In such networks, mobile users can use the advantage of diverse networks by switching between them. The most challenging issue in heterogeneous wireless networks is the Always Best Connected (ABC) concept in order to support best connectivity to mobile users at anywhere anytime from any network. Recently, many vertical handoff decision schemes have been proposed in the literature in order to provide solution to ABC. As traditional handoff decision is mainly based on received signal strength (RSS), vertical handoff decision must take into account other factors like user mobility, network conditions, pricing issues of diverse networks, and user preferences in addition to RSS. In this paper, we first outline the main characteristics needed for vertical handoff decision, and then we present an overview of the most recent and interesting vertical handoff decision schemes. We classify these strategies and present their main characteristics.

Browser games are nowadays a very popular type of computer games. Those games are played directly in web browser and do not need software installations; therefore those are available for more players. Even those have existed rather long time, there were no comprehensive study about browser games. This paper targets at analyzing browser games and categorizing of those. We also discuss financial opportunities involved in browser game genre.

The Internet has never been more insecure than ever before and scientific research projects are not immune from these rising risks, threats, and vulnerabilities. In this paper, we have aimed to identify possible threats, threat sources, and vulnerabilities of a distributed, interdisciplinary, scientific research project. We have selected our current NSF-EPSCoR funded research project as a case study. We have also briefly discussed security vulnerability trends for the year of 2010 and beyond. Finally, some countermeasures have been recommended to protect or recover assets from these attacks and vulnerabilities.

The knowledge domain is more encompassing than that of wealth and materials. For dealing with knowledge utility, all factors (its scarcity, its total utility, its marginal utility, its utilitarian value, its exchange value, etc.) that influence its evaluation must be considered. From a communication perspective, knowledge can be traced backward and extrapolated forward, much like scientific parameters. From a structural perspective, we propose that processing of knowledge be based on the most basic and fewest truisms. These truisms are based on reality and permit characterization of information and knowledge. To this extent, computational processing does not depend on the philosophic writings of earlier economists. However, the truisms are validated from a longer-term philosophic interpretation of how these truisms have survived so that they can be expanded and reused in computational environments. This approach permits machines to process knowledge based on content of a piece of information and to enhance content.

Graphical passwords are an alternative authentication method to alphanumeric passwords in which users click on images to authenticate themselves rather than type alphanumeric strings. This research aims to study the usability features of the recognition base graphical password methods available and extract the usability features of the existing methods. In this paper we study the recognition base graphical password type with the available methods from the usability point of view according to previous studies and surveys. Then we match the usability features (General usability features, existing usability features for existing graphical password methods, and ISO usability features) to the existing graphical password methods and make a comparison study between these methods and the usability features. We have found that there is no method has the most important usability features. Thus, by completing this study a set of usability features is suggested to be in one graphical password system. This set includes the easy of use, memorize, creation, learning and satisfaction. Moreover, this work proposes to build a new system of graphical password system that provides promising usability features.

In this paper we propose a bilinear pairing based hidden blind signature scheme. The proposed scheme allows the signer to append information to the hidden signed message. The requester cannot modify neither this information nor the signed message. This added information can be used to stamp the signature with a certain date and/or other relevant information which is an essential requirement in applications such as notary services and patent time proof. After issuing the signature by the signer, the requester can verify that the signature has the designated date and place. The security of the proposed scheme is analyzed against different cryptanalytic attacks.

Digital signatures are used for integrity of the signed media and authentication of the signers in digital environment. Beyond these the signature might provide an authorization service for a verifier to check whether the signer has license to sign the document in workflow. This verification is important in case of exchanging valuable documents between different organizational levels and also crucial in digital document archive. In this work, the necessity of signature authorization was described and also implementation of the authorization with X.509 based templates in hierarchical organizations was expressed. Also, a license structure for digital signature and a framework where signatures realize authorization of the signers on a document in hierarchical organization structure were proposed.

The role of intelligent communication systems in medical field is immediate and globally pervasive. Most of the modern hospitals and medical networks depend on the deployment of Intelligent Networks (IN) techniques for routing the calls and recipients to correct locations. Seeking the correct information from the appropriate medical knowledgebase is generally content-based. A decade earlier, the choice of knowledge base was entirely within discretion of the user. However, with the increasing dependability of network intelligence, the intelligent elements and agents embedded in the network can route the calls, services and agents to quickly generate the supporting data and procedures to attend to any emergency, medical or security issues. We extend these concepts to the use in hospitals, medical centers, and local communities.

This paper presents the work in progress of an on-demand software deployment system based on application virtualization concepts which eliminates the need of software installation and configuration on each computer. Some mechanisms were created, such as mapping of utilization of resources by the application to improve the software distribution and startup; a virtualization middleware which give all resources needed for the software execution; an asynchronous P2P transport used to optimizing distribution on the network; and off-line support where the user can execute the application even when the server is not available or when is out of the network.

Performance and other service quality attributes are crucial to user satisfaction of web services. Web Mining provides the key to un- derstanding web traffic behavior, which in turn explain the increasing interest in this domain and its high number of its possible applications. In this paper, we apply Web Usage Mining techniques to propose an intelligent caching solution with the goal of improving the quality of ser- vice of web sites. We found that empowering caching with a prefetching engine that predicates the components of pages to be used in the near future by users can enhance web sites performances. This is allowed by analyzing the historical of navigation of a web site reported in log files and by determining the set of components to be sollicitated in the future using frequent closed itemsets.

The aim of this paper is to address the problems associated with the current wireless ad-hoc networks, which are characterized by the lack of infrastructure. This characteristic makes it difficult to apply generic administrative approaches to solve problems such as intrusion detection. Due to the dynamic changes of topology in such networks, it is difficult to determine whether the node participating in the routing domain is Byzantine or indeed an authenticated node. The paper discusses also how mobile agent technology can be used to perform the authentication mechanism, and hence improve the utilization of the network bandwidth and provide an efficient profile management in mobile ad-hoc networks.