In parallel computing systems using the master/worker model for distributed grid computing, as the size of handling data grows, the increase in the data transmission time degrades the performance. For divisible workload applications, therefore, multiple-round scheduling algorithms have been being developed to mitigate the adverse effect of longer data transmission time by dividing the data into chunks to be sent out in multiple rounds, thus overlapping the times required for computation and transmission. However, a standard multiple-round scheduling algorithm, Uniform Multi-Round (UMR), adopts a sequential transmission model where the master communicates with one worker at a time, thus the transmission capacity of the link attached to the master cannot be fully utilized due to the limits of worker-side capacity. In the present study, a Parallel Transferable Uniform Multi-Round algorithm (PTUMR) is proposed. It efficiently utilizes the data transmission capacity of network links by allowing chunks to be transmitted in parallel to workers. This algorithm divides workers into groups in a way that fully uses the link bandwidth of the master under some constraints and considers each group of workers as one virtual worker. In particular, introducing a Grouping Threshold effectively deals with very heterogeneous workers in both data transmission and computation capacities. Then, the master schedules sequential data transmissions to the virtual workers in an optimal way like in UMR. The performance evaluations show that the proposed algorithm achieves significantly shorter turnaround times (i.e., makespan) compared with UMR regardless of heterogeneity of workers, which are close to the theoretical lower limits.

We present a TCP flow level performance evaluation on error rate aware scheduling algorithms in Evolved UTRA and UTRAN networks. With the introduction of the error rate, which is the probability of transmission failure under a given wireless condition and the instantaneous transmission rate, the transmission efficiency can be improved without sacrificing the balance between system performance and user fairness. The performance comparison with and without error rate awareness is carried out dependant on various TCP traffic models, user channel conditions, schedulers with different fairness constraints, and automatic repeat request (ARQ) types. The results indicate that error rate awareness can make the resource allocation more reasonable and effectively improve the system and individual performance, especially for poor channel condition users.

The monitoring of performance in VoIP traffic has become vital because users generally expect VoIP service quality that is as high as that of PSTN services. A lightweight method of processing by extracting VoIP flows from Internet traffics is proposed in this paper. Estimating delay variations and the packet loss ratio using knowledge about specific features and the characteristics of VoIP flows, i.e., the inter-packet gap (IPG) which is constant in VoIP flows, is also proposed. Simulation with actual traffic trace is used to evaluate the method, and this revealed that delay variations (IPG variance) can be accurately estimated by monitoring only a few percentage of all flows. The proposed method can be used as a first-alert tool to monitor large amounts of flows to detect signs of degradation in VoIP flows. The method can be used by ISPs to estimate whether VoIP flow performance is adequate within their networks and at ingress from other ISPs.

Wireless LANs (Local Area Networks) are currently spreading over diverse places such as hotels and airports, as well as offices and homes. Consequently, they provide convenient and important ways to access the Internet. Another type of communication model, Peer-to-Peer (P2P) communication on the Internet, has also attracted much attention, and P2P over wireless LANs will soon be very common. There are concerns about the capability of wireless stations (STAs) to send a large amount of traffic on an uplink. In this paper, we first clarify some issues that arise in this context by examining the feature of the Access Point(AP). Furthermore, we consider the role of the AP and propose, as a solution, ways of enabling both efficient and fair transmission over both the downlink and uplink. We evaluate the proposed schemes through simulations and show that communications over the uplink and downlink can share the wireless LAN access resource effectively. Furthermore, we show that coordinating our solution with the 802.11e Enhanced Distributed Channel Access (EDCA) can meet the different requirements of various types of traffic.

In order to prevent the degradation of TCP performance while traversing two WLANs, we present an implementation design of an inter-domain TCP handover method based on cross-layer and multi-homing. The proposed handover manager (HM) in the transport layer uses two TCP connections previously established via two WLANs (multi-homing) and switches the communication path between the two connections according to the handover trigger and the comparison of new/old APs. The handover trigger and comparison are conducted by assessing the wireless link quality using the frame-retry information obtained from the MAC layer (cross-layer). In a previous study, we proposed a preliminary concept for this method and evaluated its functional effectiveness through simulations. In the present study, we design an implementation considering a real system and then examine the effective performance in a real environment because a real system has several system constraints and suffers from fluctuations in an actual wireless environment. Indeed, depending on the cross-layer design, the implementation often degrades the system performance even if the method exhibits good functional performance. Moreover, the simple assessments of wireless link quality in the previous study indicated unnecessary handovers and inappropriate AP selection in a real environment. Therefore, we herein propose a new architecture that performs cross-layer collaboration between the MAC layer and the transport layer while avoiding degradation of system performance. In addition, we use a new assessment scheme of wireless link quality, i.e., double thresholds of frame retry and comparison of frame retry ratio, in order to prevent handover oscillation caused by fluctuations in the wireless environment. The experimental results demonstrate that the prototype system works well by controlling two TCP connections based on assessments of wireless link quality thereby achieving efficient inter-domain TCP handover in a real WLAN environment.

We propose two types of autonomic and distributed cooperative behaviors of peers for peer-to-peer (P2P) file-sharing networks. Cooperative behaviors of peers are mediated by query trails, and allows the exploration of better trade-off points between file search and storage load balancing performance. Query trails represent previous successful search paths and indicate which peers contributed to previous file searches and were at the same time exposed to the storage load. The first type of cooperative behavior is to determine the locations of replicas of files through the medium of query trails. Placement of replicas of files on strong query trails contributes to improvement of search performance, but a heavy load is generated due to writing files in storage to peers on the strong query trails. Therefore, we attempt to achieve storage load balancing between peers, while avoiding significant degradation of the search performance by creating replicas of files in peers adjacent to peers on strong query trails. The second type of cooperative behavior is to determine whether peers provide requested files through the medium of query trails. Provision of files by peers holding requested files on strong query trails contributes to better search performance, but such provision of files generates a heavy load for reading files from storage to peers on the strong query trails. Therefore, we attempt to achieve storage load balancing while making only small sacrifices in search performance by having peers on strong query trails refuse to provide files. Simulation results show that the first type of cooperative behavior provides equal or improved ability to explore trade-off points between storage load balancing and search performance in a static and nearly homogeneous P2P environment, without the need for fine tuning parameter values, compared to replication methods that require fine tuning of their parameters values. In addition, the combination of the second type and the first type of cooperative behavior yields better storage load balancing performance with little degradation of search performance. Moreover, even in a dynamic and heterogeneous P2P environment, the two types of cooperative behaviors yield good ability to explore trade-off points between storage load balancing and search performance.

In a Peer-to-Peer (P2P) network, in order to improve the search performance and to achieve load balancing, replicas of original data are created and distributed over the Internet. However, the replication methods which have been proposed so far focus only on the improvement of search performance. In this paper, we examine the load on the storage systems, which is due to writing and reading, and propose two replication methods for balancing the load on the storages distributed over P2P networks while limiting the degradation of the search performance within an acceptable level. Furthermore, we investigate the performance of our proposed replication methods through computer simulations, and show their effectiveness in balancing the load.

In this paper we propose a file replication scheme inspired by a thermal diffusion phenomenon for storage load balancing in unstructured peer-to-peer (P2P) file sharing networks. The proposed scheme is designed such that the storage utilization ratios of peers will be uniform, in the same way that the temperature in a field becomes uniform in a thermal diffusion phenomenon. The proposed scheme creates replicas of files in peers probabilistically, where the probability is controlled by using parameters that can be used to find the trade-off between storage load balancing and search performance in unstructured P2P file sharing networks. First, we show through theoretical analysis that the statistical behavior of the storage load balancing controlled by the proposed scheme has an analogy with the thermal diffusion phenomenon. We then show through simulation that the proposed scheme not only has superior performance with respect to balancing the storage load among peers (the primary objective of the present proposal) but also allows the performance trade-off to be widely found. Finally, we qualitatively discuss a guideline for setting the parameter values in order to widely find the performance trade-off from the simulation results.

In QoS networks, routing algorithms for QoS traffic have to provide the transmission path satisfying its QoS requirement while achieving high utilization of network resources. Therefore, server-based QoS routing algorithms would be more effective than distributed routing ones which are very common on the Internet. Furthermore, we believe that rerouting function enhances the advantage of their algorithms in which an already accepted flow with established path is replaced on some other path in order to accept newly arriving transmission request if it can not be accepted without doing so. Thus in this paper, we will propose a rerouting algorithm with the server-based QoS routing and evaluate its performance in terms of the blocking probability by computer simulation. In addition, we will investigate the impact of the amount of traffic with high-priority on the performance in some network topologies. Through some simulation results, we also discuss some issues arising in improving the effectiveness of rerouting.

Recent improvements in the performance of end-computers and networks have made it feasible to construct a grid system over the Internet. A grid environment consists of many computers, each having a set of components and a distinct performance. These computers are shared among many users and managed in a distributed manner. Thus, it is important to focus on a situation in which the computers are used unevenly due to decentralized management by different task schedulers. In this study, which is a preliminary investigation of the performance of task allocation schemes employed in a decentralized environment, the average execution time of a long-lived task is analytically derived using the M/G/1-PS queue. Furthermore, assuming a more realistic condition, we evaluate the performance of some task allocation schemes adopted in the analysis, and clarify which scheme is applicable to a realistic grid environment.

In this paper, we discuss configurations of photonic ATM (Asynchronous Transfer Mode) switches and their advantages in terms of the number of optical switching devices to be implemented on the system, the number of wavelengths, throughput, broadcast function etc. In particular, we focus on photonic ATM switch architectures which can be built in the near future; that is, with presently available optical and electrical devices. For example, we assume the optical devices such as optical gate switches with 40 dB on/off ratio. In this context, we evaluate 17 types of photonic ATM switches; they are 6 types of input buffer type switches, 6 types of output buffer type switches, 4 types of shared buffer switches, and 1 proposed type. From our evaluation, for cell switching, wavelength division switching technologies are desirable compared with space division switching technologies in the sense that the former enables us to build a photonic ATM switch with the less number of optical gate switches. Furthermore, we propose a switch architecture equipped with optical delay line buffers on outputs and electric buffers on inputs. We show that our switch architecture is superior in the number of required optical gate switch elements under the given conditions.

As the Internet grows, various types of traffic, such as voice, video and data, are transmitted. Therefore, the Internet should provide the Quality of Service (QoS) required by each type of traffic as well as end-to-end connectivity, and routing decisions should be based on the utilization of links/routers and/or the application types of traffic. This kind of routing is called Traffic Engineering (TE), and its objective is to improve such performance factors as flow loss probability for users and the utilization of links for networks, simultaneously. Some studies claim that the Multi-Protocol Label Switching (MPLS) technique can easily implement TE. So far, some experimental results show that TE is effective on a MPLS network; however, its performance has not been theoretically and quantitatively analyzed. Thus, in this paper, we will investigate the basic and preliminary performance of MPLS networks with TE by analyzing flow loss probability and Smoothness index of link utilization in the queueing system.

The recent evolution on the network tomography have successfully provided principles and methodologies of inferring network-internal (local) characteristics solely from end-to-end measurements, which should be followed by deployment in practical use. In this paper, two kinds of user-oriented tools for inferring one-way packet losses based on the network tomography are proposed. They can infer one-way packet loss rates on paths or path segments from/to a user-host (a client) to/from a specified target host (an application server or a router) without any measurement on the target, and thus can find the congested area along the path between the client and an application server. One is a stand-alone tool running on the client, and the other is a client-server style tool running on both the client and a proxy measurement server distributed in the Internet. Prototypes of the tools have been developed and evaluated by experiments in the actual Internet environment, which shows that the tools can infer the loss rates within 1% errors in various network conditions.

The layer3 switch enables us to fast transmit IP datagrams using the cut-through technique. The current layer3 router would become bottleneck in terms of delay performance as the amount of traffic injected into high speed networks gets relatively large. Thus, the layer3 switch should be an important element constructing the next generation Internet backbone. In this paper, we analyze the cut-through rate, the datagram waiting time and the mis-ordered rate of a layer3 switch in case of flow-driven connection setup. In the analysis, by using 3-state Markov modulated Bernoulli process (MMBP), we model the arrival process of IP flow and IP datagram from each source. Furthermore, we investigate impacts of the arrival rate and the average datagram length on the performance.

Central switches are expected to operate at the rate of Terabit per second in high speed networks, like the B-ISDN. Photonic switches using lightwave technology based on wavelength division multiplexing (WDM) and frequency division multiplexing (FDM) are promising ones for high speed switching. Such lightwave networks are mainly divided into two groups, according to the number of hops required for packets to arrive at their destinations: single-hop networks such as networks using star coupler and multihop networks such as Manhattan Street Network and ShuffleNet. In this paper we focus our attention on multihop networks and propose a mesh network, referred to as RookNet, for high speed communication. The average transmission delay time and maximum throughput of RookNet is approximately analyzed. It is shown that, as the number of nodes goes to infinity, the maximum throughput aproaches 0.433 and 0.485 when each node is equipped with no internal buffer and internal buffers of infinite capacity for relayed packets, respectively.

We propose an anomaly detection method for finding patterns in network traffic that do not conform to legitimate (i.e., normal) behavior. The proposed method trains a baseline model describing the normal behavior of network traffic without using manually labeled traffic data. The trained baseline model is used as the basis for comparison with the audit network traffic. This anomaly detection works in an unsupervised manner through the use of time-periodic packet sampling, which is used in a manner that differs from its intended purpose – the lossy nature of packet sampling is used to extract normal packets from the unlabeled original traffic data. Evaluation using actual traffic traces showed that the proposed method has false positive and false negative rates in the detection of anomalies regarding TCP SYN packets comparable to those of a conventional method that uses manually labeled traffic data to train the baseline model. Performance variation due to the probabilistic nature of sampled traffic data is mitigated by using ensemble anomaly detection that collectively exploits multiple baseline models in parallel. Alarm sensitivity is adjusted for the intended use by using maximum- and minimum-based anomaly detection that effectively take advantage of the performance variations among the multiple baseline models. Testing using actual traffic traces showed that the proposed anomaly detection method performs as well as one using manually labeled traffic data and better than one using randomly sampled (unlabeled) traffic data.

We have focused on the RIO queueing mechanism in statistical bandwidth allocation service, which uses AF-PHB. We have studied the parameterization of RIO to achieve both high throughput and low delay. We were able to parameterize RIO for that purpose in terms of both minth and maxp used in dropping OUT packets. Furthermore, we have also examined the parameterization regarding EWMA (Exponential Weighted Moving Average), i.e., weight factor wqout, and have shown that dropping OUT packets should depend upon the queue length without much delay unlike in RED. From our simulation results, we could see that our parameterization provided high throughput performance and also limited the queue length in a narrow range more effectively.

LSR (Label Switching Router)s in MPLS (Multiprotocol Label Switching) networks map arriving IP flows into some labels on Layer 2 switching fabric and establish LSP (Label Switching Path)s. By using LSPs, LSRs not only transmit IP datagrams fast by cut-through mechanism, but also solve traffic engineering issue to optimize the delay of some IP datagram flows. So far, we have analyzed the performance of LSR focusing only on the maximum number of LSPs which can be set on Layer 2. In this paper, we will also consider the bandwidth allocated to each LSP and analyze the IP datagram transmission delay and the cut-through rate of LSR. We suppose the label mapping method as the data-driven scheme in the analytical model, so that the physical bandwidth of LSR is shared by both the default LSP for hop-by-hop transmission and the cut-through LSPs. Thus, we will investigate the impact of the bandwidth allocation among these LSPs on the performance.

A basic and inevitable problem in estimating flow duration distribution arises from "censoring" (i.e., cutting off) the observed flow duration because of a finite measurement period. We extended the Kaplan-Meier method, which is used in the survival analysis field, and applied it to recover information on the flow duration distribution that was lost due to censoring. We show that the flow duration distribution from a short period of actual traffic data with censoring that was estimated using a Kaplan-Meier-based method can approximate well the flow duration distribution calculated from a sufficiently long period of actual traffic data.