Distributed Computing and Networking

With the rise of service-oriented computing, applications are more and more based on coordination of autonomous services. Envisioned over largely distributed and highly dynamic platforms, expressing this coordination calls for alternative programming models. The chemical programming paradigm, which models applications as chemical solutions where molecules representing digital entities involved in the computation, react together to produce a result, has been recently shown to provide the needed abstractions for autonomic coordination of services.

However, the execution of such programs over large scale platforms raises several problems hindering this paradigm to be actually leveraged. Among them, the atomic capture of molecules participating in concurrent reactions is one of the most significant.

In this paper, we propose a protocol for the atomic capture of these molecules distributed and evolving over a large scale platform. As the density of possible reactions is crucial for the liveness and efficiency of such a capture, the protocol proposed is made up of two sub-protocols, each of them aimed at addressing different levels of densities of potential reactions in the solution. While the decision to choose one or the other is local to each node participating in a program’s execution, a global coherent behaviour is obtained. Proof of liveness, as well as intensive simulation results showing the efficiency and limited overhead of the protocol are given.

Synchronization in distributed systems is expensive because, in general, threads must stall to obtain a lock or to operate on volatile data. Transactional memory, on the other hand, allows speculative execution so that it can hide the latencies that are inherent to distributed systems.

In this paper, we discuss how transactional memory can carry over to code that uses Java’s synchronization means i.e. monitors and volatiles. We show that we can guarantee correct execution according to the Java memory model (JMM) without having to stall at synchronization points. To this end, we use a multi-version software transactional memory system that executes JMM synchronization operations asynchronously. If any such execution has violated the JMM, the transaction rolls back. As a result, only blocking operations require immediate synchronization barriers.

Modern systems often need to address changing environment and/or faults. The economic and practical issues dictate that the existing models and/or programs be reused while providing fault-tolerance in the presence of faults.

Our paper proposes a framework of automated revision of existing program design modeled in UML to add fault-tolerance. Our framework starts with program design modeled in UML state diagram, and then automatically transforms design model to the corresponding underlying computational model. Subsequently, automated revision algorithms are applied to the underlying computational model. Finally the revised program model is converted into an UML model that provides the desired fault-tolerance property. We illustrate the whole work-flow with a case study from automotive systems.

In this paper, we consider the message forwarding problem that consists in managing the network resources that are used to forward messages. Previous works on this problem provide solutions that either use a significant number of buffers (that is

n

buffers per processor, where

n

is the number of processors in the network) making the solution not scalable or, they reserve all the buffers from the sender to the receiver to forward only one message. The only solution that uses a constant number of buffers per link was introduced in [1]. However the solution works only on a chain networks. In this paper, we propose a snap-stabilizing algorithm for the message forwarding problem that uses a constant number of buffers per link as in [1] but works on tree topologies.

The aim of a Software Transactional Memory (STM) system is to discharge the programmer from the explicit management of synchronization issues. The programmer’s job resides in the design of multiprocess programs in which processes are made up of transactions, each transaction being an atomic execution unit that accesses concurrent objects. The important point is that the programmer has to focus her/his efforts only on the parts of code which have to be atomic execution units without worrying on the way the corresponding synchronization has to be realized.

Non-trivial STM systems allow transactions to execute concurrently and rely on the notion of commit/abort of a transaction in order to solve their conflicts on the objects they access simultaneously. In some cases, the management of aborted transactions is left to the programmer. In other cases, the underlying system scheduler is appropriately modified or an underlying contention manager is used in order that each transaction be (“practically always” or with high probability) eventually committed.

This paper presents a deterministic STM system in which (1) every invocation of a transaction is executed exactly once and (2) the notion of commit/abort of a transaction remains unknown to the programmer. This system, which imposes restriction neither on the design of processes nor or their concurrency pattern, can be seen as a step towards the design of a deterministic universal construction to execute transaction-based multiprocess programs on top of a multiprocessor. Interestingly, the proposed construction is lock-free (in the sense that it uses no lock).

We consider here the Byzantine agreement problem in synchronous systems with

homonyms

. In this model different processes may have the same authenticated identifier. In such a system of

n

processes sharing a set of

l

identifiers, we define a

distribution of the identifiers

as an integer partition of

n

into

l

parts

n

1

,…,

n

l

giving for each identifier

i

the number of processes having this identifier.

Assuming that the processes know the distribution of identifiers we give a necessary and sufficient condition on the integer partition of

n

to solve the Byzantine agreement with at most

t

Byzantine processes. Moreover we prove that there exists a distribution of

l

identifiers enabling to solve Byzantine agreement with at most

t

Byzantine processes if and only if

$l > \frac{(n-r)t}{n-t-min(t,r)}$

where

$r = n \bmod l $

.

This bound is to be compared with the

l

 > 3

t

bound proved in [4] when the processes do not know the distribution of identifiers.

Due to their increasing complexity, today’s SoC (System on Chip) systems are subject to a variety of faults (e.g., soft errors, component crash, etc.), thereby making fault tolerance a highly important property of such systems. However, designing fault tolerance is a complex task in part due to the large scale of integration of SoC systems and different levels of abstraction provided by modern system design languages such as SystemC. Most existing methods enable fault injection and impact analysis as a means for increasing design dependability. Nonetheless, such methods provide little support for designing fault tolerance. To facilitate the design of fault tolerance in SoC systems, this paper propose an approach where fault tolerance is designed at the level of inter-component communication protocols in SystemC Transaction Level (TL) models. The proposed method includes four main steps, namely model extraction, fault modeling, addition of fault tolerance and refinement of synthesized fault tolerance to SystemC code. We demonstrate the proposed approach using a simple SystemC transaction level program that is subject to communication faults. We also provide a roadmap for future research at the intersection of fault tolerance and hardware-software co-design.

Network virtualization is an important concept to overcome the ossification of today’s Internet as it facilitates innovation also in the network core and as it promises a more efficient use of the given resources and infrastructure. Virtual networks (VNets) provide an abstraction of the physical network: multiple VNets may cohabit the same physical network, but can be based on completely different protocol stacks (also beyond IP). One of the main challenges in network virtualization is the efficient admission control and embedding of VNets. The demand for virtual networks (e.g., for a video conference) can be hard to predict, and once the request is accepted, the specification / QoS guarantees must be ensured throughout the VNet’s lifetime. This requires an admission control algorithm which only selects high-benefit VNets in times of scarce resources, and an embedding algorithm which realizes the VNet in such a way that the likelihood that future requests can be embedded as well is maximized.

This paper describes a generic algorithm for the online VNet embedding problem which does not rely on any knowledge of the future VNet requests but whose performance is competitive to an optimal offline algorithm that has complete knowledge of the request sequence in advance: the so-called competitive ratio is, loosely speaking, logarithmic in the sum of the resources. Our algorithm is generic in the sense that it supports multiple traffic models, multiple routing models, and even allows for nonuniform benefits and durations of VNet requests.

We present and analyze a wait-free deterministic algorithm for solving the at-most-once problem: how

m

shared-memory fail-prone processes perform asynchronously

n

tasks at most once. Our algorithmic strategy provides for the first time nearly optimal effectiveness, which is a measure that expresses the total number of tasks completed in the worst case. The effectiveness of our algorithm equals

n

 − 2

m

 + 2. This is up to an additive factor of

m

close to the known effectiveness upper bound

n

 − 

m

 + 1 over all possible algorithms and improves on the previously best known deterministic solutions that have effectiveness only

n

 − log

m

·o(

n

). We also present an iterated version of our algorithm that for any

$m = \mathrm{O}(\sqrt[3+\epsilon]{n/\log n})$

is both effectiveness-optimal and work-optimal, for any constant

ε

 > 0. We then employ this algorithm to provide a new algorithmic solution for the Write-All problem which is work optimal for any

$m=\mathrm{O}(\sqrt[3+\epsilon]{n/\log n})$

.

For unconditionally reliable message transmission (URMT) in synchronous directed networks of

n

nodes, a subset of which may be malicious, it is well-known that the minimum connectivity requirements for zero-error (perfect) protocols to exist is strictly higher than those where a negligible yet non-zero error probability is allowed (

Monte Carlo

protocols) [12]. In this work, we study the minimum connectivity requirements for the existence of (a) synchronous

Las Vegas

, (b) asynchronous Monte Carlo, and (c) asynchronous Las Vegas protocols for URMT.

Interestingly, we prove that in any network, a synchronous Las Vegas URMT protocol exists if and only if an asynchronous Monte Carlo URMT protocol exists too. We further show that in any network, an asynchronous Las Vegas URMT protocol exists if and only if a synchronous perfect protocol exists as well. Thus, our results establish an interesting interplay between (im)perfectness, synchrony and connectivity for the case of URMT.

Paxos is probably the most popular state machine replication protocol. Two optimizations that can greatly improve its performance are batching and pipelining. Nevertheless, tuning these two optimizations to achieve high-throughput can be challenging, as their effectiveness depends on many parameters like the network latency and bandwidth, the speed of the nodes, and the properties of the application. We address this question, by first presenting an analytical model of the performance of Paxos that can be used to obtain values for tuning batching and pipelining. We then present results of experiments validating the model and investigating how these two optimizations interact in a WAN. Results for LAN are also mentioned. The results show that although batching by itself is usually sufficient to maximize the throughput in a LAN environment, in a WAN it must be complemented with pipelining.

Simulation is often used for the evaluation of new network protocols and architectures. In order to perform more realistic simulations, modern simulators such as ns-3 integrate more detailed models and even support direct execution of real protocol code. However, such complex models require more computational and memory requirements. In this paper, we study the feasibility of a hybrid approach based on distributing a complex simulation scenario on several nodes in a grid network. We show that by exploiting the real time operation of the ns-3 simulator, it is possible to map such complex scenarios on grid nodes. We run experiments to define the operational zone in which the obtained results are accurate. We also propose a basic mapping algorithm to distribute a simulation scenario in several nodes.

Acyclic flow networks, present in many infrastructures of national importance (e.g., oil & gas and water distribution systems), have been attracting immense research interest. Existing solutions for detecting and locating attacks against these infrastructures, have been proven costly and imprecise, especially when dealing with large scale distribution systems. In this paper, to the best of our knowledge for the first time, we investigate how mobile sensor networks can be used for optimal event detection and localization in acyclic flow networks. Sensor nodes move along the edges of the network and detect events (i.e., attacks) and proximity to beacon nodes with known placement in the network. We formulate the problem of minimizing the cost of monitoring infrastructure (i.e., minimizing the number of sensor and beacon nodes deployed), while ensuring a degree of sensing coverage in a zone of interest and a required accuracy in locating events. We propose algorithms for solving these problems and demonstrate their effectiveness with results obtained from a high fidelity simulator.

Today’s large scale distributed systems are characterized by strong dynamics caused by the inherent unreliability of their constituting elements (e.g. process and link failures, processes joining or leaving the system). This continuous dynamism has a strong negative impact on distributed algorithms designed to work on such systems. Regular registers [1], Replication [2], in-network aggregation[3], are all examples of such problem.

Directional sensor network is composed of many directional sensor nodes. Unlike conventional sensors that always have an omni-angle of sensing range, directional sensors may have a limited angle of sensing range due to technical constraints or cost considerations. Therefore, it is possible that when a directional sensor node is randomly deployed and scattered in the environment, some interested targets cannot be covered even if these targets are located in the sensing range of the sensor. We propose a Maximum Coverage with Rotatable Sensors (MCRS) problem in which coverage in terms of the number of targets to be covered is maximized whereas the rotated angles of sensors are minimized. We present two distributed greedy algorithm solutions for the MCRS problem. Simulation results shows that to apply angle adjustment algorithm can enhance the coverage rate of the directional sensor network.

The coverage problem in wireless sensor networks addresses the problem of covering a region with sensors. Many different definitions of coverage exist depending on the goal of the coverage. In this paper, we address the problem of covering lines in a bounded rectangular region

R

by a set of sensors. We first introduce two new metrics,

smallest k-covered line segment

and

longest k-uncovered line segment

, for measuring the quality of line coverage achieved by a sensor deployment. Polynomial time algorithms are then proposed for finding the

smallest k-covered axis-parallel line segment

and the

longest k-uncovered axis-parallel line segment

in a bounded rectangular region

R

.

Urban Sensing is an emerging paradigm that combines the ubiquity of smartphones with measurement capabilities of sensor networks. While this concept is still in development, related security and privacy concerns become increasingly more relevant. In this paper, we focus on a number of scenarios where nodes of an Urban Sensing system are subject to individual queries. We address the problem of protecting

query privacy

(i.e., hiding which node matches the query) and

data privacy

(i.e., hiding sensed data). We introduce a realistic network model and two novel adversarial models:

resident

and

non-resident

adversaries. For each of them, we propose a distributed privacy-preserving technique and evaluate its effectiveness via analysis and simulation. To the best of our knowledge, this is the first attempt to define and address both query and data privacy in the context of Urban Sensing. Our techniques are tunable, trading off the level of privacy assurance with a small overhead increase. We additionally provide a relevant improvement of data reliability and availability, while only relying on standard symmetric cryptography. The practicality of our proposals is demonstrated both analytically and experimentally.

Target tracking in Wireless Sensor Networks is often not limited to the target discovery and computation of the target trajectory. In some cases, it may be required to intercept the target. The tracking information is used to guide a vehicle towards the target. This is referred to as the “Guided Navigation” problem. This paper presents a mechanism that uses Kalman filter based target state prediction and adaptive velocity of the guided object to intercept the target, using a hierarchical clustered wireless sensor network architecture. The use of adaptive velocity is different from earlier mechanisms, such as the Current Location based Guidance (CLG) and

α

-

β

Predicted Proportional based guidance (ABG) algorithms, which use fixed velocity of the guided object. Simulation modeling based experiments were conducted for varying network field sizes with a random sensor deployment. The results show that the proposed algorithm can intercept the target earlier (approx. 25% reduction in intercept time) when compared to the existing algorithms.

Wireless Sensor Networks (WSNs) pose a few unique security challenges due to the fact that they (often) run unattended, do not rely on tamper-resistant hardware, and are severely resource constrained—to name a few. In this context, a particularly dreadful attack is the replica attack. That is, sensors are captured and their state seized, and replicated in reprogrammed sensors eventually performing some rogue activities. While some solutions to this problem do exist in static WSNs, mobile WSNs lack of solutions that are both effective and efficient, due to the complexity added by sensor mobility.

In this paper, we propose a novel solution against the replica attack in mobile WSNs. In particular, we provide the following contributions: we first introduce a novel realistic attacker model that can be used to assess the quality of the solutions provided. Later, we detail a distributed, efficient, and cooperative protocol to detect replica. Leveraging just local (one-hop) communications and node mobility our solution enforces the emergent property of replica detection. Finally, our solution is tested against the introduced attacker model. Simulation results show that our solution is effective and efficient—providing high detection rate while incurring limited overhead.

The design of large-scale critical infrastructures demands for innovative data dissemination services, able to jointly provide reliability and timeliness guarantees. Current middleware solutions do not address both these aspects. Indeed, fault tolerance is typically achieved at the cost of severe performance fluctuations, or timeliness is always obtained by softening the fault-tolerance requirements. In this paper we propose to fulfill this lack by combining two different approaches, namely coding and gossiping. We provide a theoretical model to evaluate the potential benefit of coding on the information delivery performance. These results are also confirmed by an experimental analysis conducted on a real air traffic control workload, which evidences how coding mitigates latency and overhead penalties to ensure reliable event notification.

In this paper, we propose a protocol for downstream communication in a wireless sensor network (WSN) based on the Postorder Numbering (PN) scheme in a tree. The existing solutions are either based on mesh routing or on full dissemination. Mesh routing is not suitable for large collection networks, while dissemination based routing can not address individual nodes. PN routing requires parent information, and postorder numbering for addressing individual nodes for routing the packets downstream. We combine PN routing with the concept of NATing to provide a simple light weight transport protocol independent of application. The transport layer of the PN routing is merged with that of a tree based collection protocol for upstream routing to provide a seamless integration of IP network with WSN.

Secure dynamic IP addressing is a prime requirement in mobile ad hoc networks (MANETs), as the node cannot participate in unicast communications or routing until it is assigned a unique address. Recently, several approaches have been proposed for dynamic address allocation in MANET and most of the these approaches rely on broadcasting for address solicitation and/or duplicate address detection. As a result, several types of security threats can be observed at the time of address allocation. In this paper, we present an ID based distributed dynamic IP configuration scheme that securely allocate IP addresses to the authorized nodes for a mobile ad hoc network without broadcasting throughout the network. The scheme is distributed among MANET nodes and therefore each node can generate unique IP addresses from its own IP address and can assign that addresses to the new nodes. The proposed scheme provides authentication for address allocation without the help of a trusted third party while taking care of the security-threats associated with dynamic IP allocation protocol. Performance analysis shows that the proposed addressing scheme has low control overhead and fairly good addressing latency with added security mechanisms compared to the similar existing configuration schemes. Moreover, the proposed scheme is efficient and secure to solve the problem of network partitions and mergers along with the arrival and departure of a node.

Wireless sensor networks (WSNs) have many promising applications for monitoring critical regions, such as in military surveillance and target tracking. In such applications, privacy of the location of the source sensor is of utmost importance as its compromise may reveal the location of the object being monitored. Traditional security mechanisms, like encryption, have proven to be ineffective as location of the source can also be revealed by analysis of the direction of traffic flow in the network. In this paper, we investigate the source-location privacy issue. We first propose a semi-global eavesdropping attack model which we show as being more realistic than the local or global eavesdropping attack model discussed in literature. In this model, we use a linear-regression based traffic analysis technique and show that it is effective in inferring the location of the data source under an existing source-location preserving technique. To measure source location privacy against this semi-global eavesdropping, we define an

α

-angle anonymity model. Additionally, we adapt the conventional function of data mules to design a new protocol for securing source location privacy, called the Mules-Saving-Source (MSS) protocol, which provides

α

-angle anonymity. We analyze the delay incurred by using data mules in our protocol and examine the association between privacy preservation and data delay in our protocol through simulation.

Stanford University Interim (SUI) channel model has been proposed for simulations, design, development and testing of technologies suitable for IEEE 802.16 networks. SUI channel model proposes a set of six empirical time-dispersive channels for three typical terrain types. Most of the simulation studies for IEEE 802.16 networks involving Multi-Input Multi-Output (MIMO), either use a flat fading channel model or adopt an existing analytical/standard model, such as Kronecker model, 3GPP, IEEE 802.11 Broadband wireless models, and Pedestrian model A-B. Although this reduces the complexity of the channel models, it results in lower accuracy. This paper presents the evaluation of Bit Error Rate (BER) performance of various MIMO techniques for 2×2 and 4×4 antenna configurations, over SUI channel models. The encoding and decoding equations for Space-Time Block Code (STBC)

$\mathcal{G}4$

and Spatial Multiplexing (SM) for frequency selective channels are also presented.

In wireless networks, the planar embedding of the communication network is often used in applications such as topology control and routing. Typically, wireless networks are nonplanar and hence algorithms which use local information are often applied to create planar graphs. These algorithms provably yield a connected planar graph as long as the network obeys certain assumptions like Unit Disk Graph (UDG). In this paper we propose a new

explicit

localized graph planarization algorithm to planarize graphs that are more general than UDGs, i.e. graphs satisfying

redundancy property

and a new property introduced by us called

coexistence property

. This algorithm detects intersections locally and planarizes networks by removing them. A link addition phase ensures that the network remains connected after the planarization algorithm. Theoretical analysis shows that our algorithm provably creates planar graphs without disconnecting the network graph if it has redundancy and coexistence properties. We also show these planar graphs are weak

c

-spanners at least in UDG modeled networks. Empirical analysis shows that they are as good as the best state of the art localized planarization algorithm with respect to the spanning ratio.

The iTrust system is a decentralized and distributed information publication, search and retrieval system, whose objective is to prevent censorship and filtering of information accessed over the Internet. In iTrust, metadata describing information are randomly distributed to multiple participating nodes. Similarly, requests containing keywords are randomly distributed to multiple participating nodes. If a participating node receives a request and the keywords in the request match the metadata it holds, the participating node sends the URL for the information to the requesting node. The requesting node then can retrieve the information from the source node. In this paper, we present the iTrust messaging and membership protocols. We establish lower bounds for the probabilities of a match if all of the participating nodes are operational and if a proportion of the participating nodes are non-operational or subverted. We provide probabilistic results for

n

participating nodes, where the metadata and the requests are distributed to a multiple of the square root of

n

nodes. These results show that distribution of the metadata and the requests to relatively few nodes suffices to achieve a high probability of a match, even if some of the nodes are non-operational or subverted.

In this paper, we present a MANET experimentation framework developed at Poznan University of Technology within the EU FP7 OPNEX project including its main part called the

wnPUT

testbed. The presentation is complemented by description of the state-of-the art in research on wireless testbed development and experimentation methodology. The main goal of the wnPUT testbed environment is to support the first stage of experimental research on backpressure-based wireless networks. From this perspective, wnPUT is a small-scale testbed with the aim of ‘preliminary’ experimentation supporting target activities realized subsequently in large-scale testbed environments. The paper contributes with a description of the testbed development process, tested functionalities, as well as hardware and software characteristics. In particular, we analyze experiment execution phases such as network configuration, node monitoring and results evaluation, as well as a set of methods used to gather and visualize statistics.

Cloud computing is a paradigm that has the potential to transform and revolutionalize the next generation IT industry by making software available to end-users as a service. A cloud, also commonly known as a cloud network, typically comprises of hardware (network of servers) and a collection of softwares that is made available to end-users in a

pay-as-you-go

manner. Multiple public cloud providers (ex., Amazon) co-existing in a cloud computing market provide similar services (software as a service) to its clients, both in terms of the nature of an application, as well as in quality of service (QoS) provision. The decision of whether a cloud hosts (or finds it profitable to host) a service in the long-term would depend jointly on the price it sets, the QoS guarantees it provides to its customers , and the satisfaction of the advertised guarantees. In this paper, we devise and analyze three

inter-organizational

economic models relevant to cloud networks. We formulate our problems as

non co-operative

price and QoS games between

multiple

cloud providers existing in a cloud market. We prove that a

unique

pure strategy Nash equilibrium (NE) exists in two of the three models. Our analysis paves the path for each cloud provider to 1) know what prices and QoS level to set for end-users of a given service type, such that the provider could exist in the cloud market, and 2) practically and dynamically provision appropriate capacity for satisfying advertised QoS guarantees.

MIMO techniques supported by IEEE 802.16 networks improve either throughput or reliability in the network. But these MIMO techniques do not always perform optimally, especially in the presence of high mobility. In this paper, we propose a cross layered mapping technique that exploits multiple antenna available at each MS. An optional error correction mechanism is proposed at the receiver to correct erroneously received signal. Finally, using extensive simulations we show that the proposed technique achieves higher throughput compared to the existing techniques while providing the same reliability. We also show that the proposed technique can be a stand alone technique and adaptive switching of MIMO techniques is not required.

A generally agreed upon requirement for correctness of concurrent executions in Transactional Memory systems is that all transactions including the aborted ones read consistent values. We denote this as all-reads-consistency. Opacity is a correctness criterion that satisfies the above requirement. A relevant property, which we call as non-interference, is that an aborted transaction should not affect the consistency of the transactions that are executed subsequently. This property is desirable in general and critical for closed nested transactions in the sense that the read steps of an aborted sub-transaction (that were executed before aborting) may make committing its top-level transaction impossible. Recently we proposed a new correctness criterion, Abort Shielded Consistency, that satisfies both all-reads-consistency and non-interference. In this paper, we present an efficient on-line scheduler that implements Abort Shielded Consistency. The scheduler is based on the notion of conflicts which have been appropriately defined for optimistic executions. The scheduler maintains a conflict-graph based on the events executed so far. An event is allowed to be executed only if it does not cause a cycle in the graph. The conflict-graph has separate components for each (parent) sub-transaction. Each component can be maintained autonomously at the site executing the sub-transaction and the checking can be done in a distributed manner.

In this paper the problem of designing optimal logical topologies for WDM networks using the tabu search meta-heuristic has been considered. For a given physical topology and requests for data communication at specified rates between different pairs of nodes in the network, our objective is to obtain a logical topology, using the tabu search, to minimize the network congestion.

This paper presents the design and implementation of the Dynamic Transport Library (DTL), a UDP-based reliable transport library, initially designed for - but not limited to - peer-to-peer applications. DTL combines many features not simultaneously offered by any other transport library including:

i

) Wide scope of congestion control levels starting from less-than-best-effort to high-priority,

ii

) Prioritization of traffic relative to other non-DTL traffic,

iii

) Prioritization of traffic between DTL connections,

iv

) NAT-friendliness,

v

) Portability, and

vi

) Application level implementation. Moreover, DTL has a novel feature, namely, the ability to change the level of aggressiveness of a certain connection at run-time. All the features of the DTL were validated using a controlled environment as well as the Planet Lab testbed.

Mobile data services have become very popular, but there is still no widely deployed solution supporting handovers between heterogeneous networks. Existing approaches usually require support of the infrastructure, which prevents a widespread use.

In this paper, a new solution for mobility based on Datagram Transport Layer Security (DTLS) is introduced, which can be implemented in a library as part of an application and does not require any specific support of the operating system or network infrastructure. DTLS can be used by applications directly or by tunneling solutions for providing security and mobility. Additionally, the Heartbeat Extension for DTLS is presented, which is required for mobility and used to provide the necessary keep-alive functionality as well as to realize a Path MTU Discovery.

Much effort has been put into making P2P applications ISP-friendly, i.e. finding a solution that reduces inter-ISP traffic while maintaining the high file-sharing efficiency of existing P2P applications. Related works have analyzed different ISP-friendly P2P solutions by simulations. However, most of these simulation models have two common assumptions; 1) the peer arrival process follows a flash-crowd scenario, 2) the distribution of peers follows a uniform distribution. Most of the time, both these assumptions are not true for a P2P network. In this paper we show that the peer arrival process and the distribution of peers have a significant effect on the simulation results. We also show that solutions perform differently, e.g. in terms of the amount of inter-ISP traffic and download time, in a flash crowd scenario than in a steady state, limiting the applicability of the results and conclusions reported in previous works. Based on this new insight, we propose a Preemptive Neighbor Selection (PreeN) that makes an ISP-friendly P2P application efficient in steady state scenarios.

abstract

environment.

Information dissemination in a decentralized network is a challenging issue as no single node can have knowledge of the whole network for reasons like security, access control, memory. Most importantly - a centralized service provider is missing which can cater to user requests. In this paper we propose and investigate an algorithm for information dissemination in decentralized networks which only uses local information - that is information associated with a node and its neighbors for disseminating a piece of information with a goal of maximum coverage and minimum spamming. The approach emulates multiagent system using particle swarms, and explores multiple relations that exist among nodes while selecting potential forwarders in case members having same interests are not well connected. The approach is generic enough to be used in various kinds of scenarios: for example in job market or network of emails, decentralized recommendation and contextual advertisement systems. We measure various metrics such as recall, precision and message duplication to analyze the quality of our protocol. The experimental results on real and synthetic dataset show the effectiveness of our approach.

Optimized Link State Routing (OLSR) protocol is a leading proactive routing protocol for Mobile Ad-hoc Networks (MANETs). Since the OLSR protocol in its standard version does not support multi-path packet forwarding, we have developed and implemented its multi-path extension. As a result of a slightly modified path computation algorithm and the application of backpressure Max-Weight Scheduling (MWS) policy, the MANETs can utilize the extended functionality based on OLSR. The experiment results presented in this paper compare the performance of the standard and extended OLSR versions in a large congested MANET (i.e., in large-scale DES-Testbed located at Freie Universitat Berlin).

Buffer dimensioning of nodes is essential to design a practical and efficient Delay-Tolerant Network (DTN). The existing literature on DTN assumes either infinite or finite (arbitrary) buffer size of the nodes in the system model; however, it does not quantify the buffer size. In this paper, we propose a large deviations framework to quantify the buffer size of DTN nodes moving according to Random WayPoint (RWP) mobility model and investigate the effect of buffer size in terms of its impact on the performance of underlying message forwarding protocol. Our extensive simulation results show that the performance of the proposed dimensioned buffer model is statistically equivalent to that of the infinite buffer model.

The

store, carry, and forward

paradigm of the Delay- Tolerant Network (DTN) architecture enables a node to carry messages for a long period of time. This long-term storage is supported by the DTN architecture with the usage of persistent storage; however to the best of our knowledge, the routing/scheduling framework that incorporates support for persistent storage has not been addressed much in the DTN literature. In this paper, we investigate the impact of persistent storage on the routing performance over different buffer scheduling policies. Our extensive simulation studies demonstrate that they exhibit an improvement in delivery ratio, but with a compromise on delivery delay. This shows the pressing need for a new scheduling policy to tap the complete potential of the persistent storage. To this end, we propose a Time in Primary Scheduling (TiPS) policy with two variants (one using local information and the other using global information) that outperforms the contemporary buffer scheduling policies with respect to the persistent storage framework.

Wormhole-switching and virtual channel flow control are two critical techniques in networks-on-chip (NoCs). In an NoC adopting these two techniques, a packet may hold several virtual channel (vc) resources spanning multiple routers. These vcs constitute a vc chain to the packet. Through observation, we find that the lengths of the vc chains play an important role in the performance of an NoC, and it helps to improve the performance of the network to cut short the vc chains. In this paper, we propose a novel input selection function (ISF) which allows packets spanning in the network in a more compact and consecutive manner, thereby lowering the delay while simultaneously boosting throughput. Owing to the simplicity of the novel ISF, we can implement it with a practical design, incurring a minimal hardware overhead with an additional requirement of storage less than 3.6%. We simulate and evaluate the proposed input selection approach in terms of average delay and throughput. Our experimental results indicate that the proposed ISF is effective in NoC design compared to other ISFs in previous literatures. Though we assume a two-dimensional mesh topology throughout this paper, the proposed ISF can be readily extended to other topologies. Furthermore, it can be coupled with any OSF and any routing algorithm.

The peer-to-peer (P2P) technology has been well developed over the Internet. BitTorrent (BT) is one of the most popular P2P sharing protocols. BT network traffic detection has become increasingly important since ISP and enterprise networks often want to detect and limit P2P traffic for other critical applications. We propose a new detection method that is based on an intelligent combination of Deep Packet Inspection (DPI) and Deep Flow Inspection (DFI) with semi-supervised learning. Comparing with existing methods, the new method has achieved equally high accuracy with shorter classification time.

Proxy signatures allow a proxy signer to sign messages on behalf of an original signer within a given context. They are widely used in distributed systems, grid computing, mobile agent applications, distributed shared object systems, global distribution networks, and mobile communications. Multi-proxy signature is a variation of the proxy signature primitive. In such a scheme, an original signer delegates his signing power to a group of proxy signers, and then only the cooperation of all the proxy signers can generate valid proxy signatures, referred to as multi-proxy signatures, on behalf of the original signer. Recently, a multi-proxy signature scheme is presented in certificateless public key cryptosystem. We show this scheme is insecure.