Zenghua ZHAO Yantai SHU Lianfang ZHANG Oliver YANG
Multi-Protocol Label Switching (MPLS) can efficiently support the explicit routes setup by the use of Label Switched Paths (LSPs) between an ingress Label Switched Router (LSR) and an egress LSR. Hence it is possible to distribute the network traffic among several paths to achieve load balancing, thus improving the network utilization, and minimizing the congestion. The packet-level traffic characteristics in the Internet is so complex that it is natural to do traffic engineering (TE) and control at the flow level. The emerging Multi-Protocol Label Switching (MPLS) has introduced an attractive solution to TE in IP networks. The main objective of this paper is to balance traffic at the flow level among the parallel Label Switched Paths (LSPs) in MPLS networks. We introduce a multipath load-balancing model at the flow level. In this model, each LSP is modeled as an M/G/1 processor-sharing queue. The load-balancing problem is then considered as an optimization problem. Based on the analysis of the model, we propose a heuristic but efficient mechanism that can make good use of the traffic characteristics at the flow level. Packet disorder is avoided effectively by dispatching packets belonging to one flow to the same path. This mechanism only need to be implemented in the ingress LSRs and the egress LSRs, while the intermediate LSRs only forward the packets. Apart from discussing the traffic allocation granularity, and the implementation issues in details, we have also performed extensive simulations using NS-2 with MPLS modules. The simulation results show that the load through the network is well balanced so that the network throughput is improved and the delay is decreased efficiently.
Atsushi SASAKI Tadashi ARARAGI Shigeru MASUYAMA Keizo MIYATA
We formally define the mobile agent allocation problem from a system-wide viewpoint and then prove that it is strongly NP-complete even if each agent communicates only with two agents. This is the first formal definition for scheduling mobile agents from the viewpoint of load balancing, which enables us to discuss its properties on a rigorous basis. The problem is recognized as preemptive scheduling with independent tasks that require mutual communication. The result implies that almost all subproblems of mobile agent allocation, which require mutual communication of agents, are strongly NP-complete.
Shigeyuki SAKAZAWA Yasuhiro TAKISHIMA Yoshinori KITATSUJI Yasuyuki NAKAJIMA Masahiro WADA Kazuo HASHIMOTO
This paper presents a novel data transmission protocol "SVFTP," which enables high-speed and error-free video data transmission over IP networks. A video transmission system based on SVFTP is also presented. While conventional protocols are designed for file transmission, SVFTP focuses on video data as a continuous media. In order to fit a flexible video transmission system, SVFTP achieves higher throughput on the long distance link as well as transmission interruption/resumption and progressive download and play back. In addition, a rate shaping mechanism for SVFTP is introduced in order to control greediness and burst traffic of multiple-TCP sessions. Laboratory and field transmission experiments show that SVFTP achieves high performance and functionality.
Feng-Tso CHIEN Ming-Hung LAI Shih-Tzung SU Kou-Way TU Ching-Ling CHENG
A new high ruggedness Power MOSFET structure with a planar oxide self align p+ implant structure is proposed and discussed. We compare the proposed self-align process with the conventional p+ MASK process and contact p+ implant process. It is shown that the self align implant structure with a wide p+ area can reduce the parasitic BJT effect and, therefore, improve the device's avalanche energy capability, which is required for inductive load circuits. Based on the unclamped inductive load switching measurement results, the proposed device avalanche energy with self align p+ implant process is improved about 355% as compared to the traditional one.
Yoji KISHI Takeshi KITAHARA Yujin NOISHIKI Akira IDOUE Shinichi NOMOTO
Broadband fixed wireless access (BFWA) systems with multi-hop mesh topologies have attracted considerable attention as a promising technology for next generation, high quality, high capacity, and high density access infrastructures. The primary advantages of mesh network topologies are an improvement of capacity by means of traffic engineering throughout the networks. This paper discusses an adaptive traffic load balancing method that maximizes the capacity for the mesh BFWA networks. Taking into account the variation of network conditions such as traffic demand distributions and qualities of wireless links, the adaptive traffic load balancing method attempts to equalize the utilization of capacity for each wireless link. To avoid deteriorating the performance of TCP communications, the proposed method implements flow-based traffic load balancing. Performance of the proposed adaptive traffic load balancing method is demonstrated and validated using the experimental mesh network environments with wired networks with up to sixteen nodes that emulates the variation of the wireless link capacity.
Ilseok HAN Wanyoung KIM Hagbae KIM
This paper presents an optimal load balancing algorithm based on both of the ANFIS (Adaptive Neuro-Fuzzy Inference System) modeling and the FIS (Fuzzy Inference System) for the local status of real servers. It also shows the substantial benefits such as the removal of load-scheduling overhead, QoS (Quality of Service) provisioning and providing highly available servers, provided by the suggested method.
Toshiyuki NAKANISHI Seiichi SAMPEI Hiroshi HARADA Norihiko MORINAGA
This paper proposes an OFDM based adaptive modulation scheme employing variable coding rate (VCR OFDM AMS), which selects optimum modulation and coding scheme (MCS) realized by combination of several modulation schemes and coding rates. The OFDM AMS with multilevel transmit power control (OFDM AMS/MTPC) can realize high data rate transmission in the dynamic parameter controlled-orthogonal frequency and time division multiple access (DPC-OF/TDMA). The employment of OFDM AMS/MTPC, however, makes transceiver design rather complex. To solve this problem, we propose to improve throughput performances of the OFDM AMS without employment of the MTPC. The simple OFDM AMS, however, does not fully utilize transmit power for throughput improvement because there is surplus transmit power which corresponds to power margin over required signal to interference plus noise power ratio (SINR). Thus, in order to improve transmit power efficiency for throughput increase, we reduce the required SINR gaps between adjacent MCSs by introducing many coding rates. Furthermore, this paper presents an effective bit loading algorithm when multiple coding rates as well as modulation schemes are used. Computer simulation confirms that the proposed VCR OFDM AMS gives sufficient throughput performances as an alternative to the OFDM AMS/MTPC.
Optimum wideband beam pattern synthesis methods are usually sensitive to antenna elements gain, phase and position errors. In this letter, these errors are taken into account in a constraint optimization process, and a generalized diagonal loading algorithm is obtained. Computer simulations indicate the robustness of this new method.
Kazuki IWASHITA Ikuo AWAI Hiroshi KUBO Atsushi SANADA
Unloaded Q of a dielectric image resonator is discussed based on the electromagnetic field distribution. It is shown that a partial air gap and a dielectric sheet with low permittivity between the dielectric resonator and the shield case reduce both the dielectric loss and the conductor loss. Especially, reduction of the conductor loss is significant, since the magnetic field distribution moves from the conductor to the upper part of resonator. A half-cut image resonator with an air gap and dielectric spacer is simulated and measured. The unloaded Q of the dielectric resonator with low dielectric loss is improved by about two times from that of original image resonator.
Marian CODREANU Djordje TUJCOVIC Matti LATVA-AHO
The knowledge of channel state information (CSI) at the transmitter (TX), which in case of time division duplex (TDD) systems can be easily obtained due to radio channel reciprocity, can dramatically increase the spectral efficiency of a multiple-input multiple-output (MIMO) system. This paper presents a robust link adaptation method for TDD systems employing MIMO-OFDM channel eigenmode based signalling. We propose a rather simple logarithm-free, bit and power loading algorithm which requires low signalling overhead. Proposed method maintains constant frame error rate (FER) by controlling the instantaneous transmitted power in such a way that the average SNR at the equalizer output is kept constant. The scheme takes into account channel estimation errors at TX side by partial compensation of the SNR estimation errors. By using a simple scalar channel encoder, for an unbalanced MIMO system with larger number of transmit than receive antennas, the achieved spectral efficiency at low and medium SNR is significantly higher than the outage MIMO capacity with unknown CSI at the TX. The simulation results show that for a constant FER, the throughput degradation compared to the universally accepted Hughes-Hartogs algorithm is negligible.
Jongkuk PARK Jong-Sik LIM Sangwook NAM
In this Letter, a dumbbell-shaped patch loaded slotline(PLS) is proposed. Like the conventional defected ground structure(DGS) for a microstrip line, we show that the proposed PLS can provide a wide bandstop characteristic in some frequency bands with only one or small number of unit cells. Also, the equivalent circuit model for a unit section is derived from the analysis of the field distributions in the structure and its circuit parameters are determined by means of full wave numerical simulations. This equivalent circuit is shown to be dual to that of the typical DGS in a microstrip line. A broadband microstrip to slotline transition is incorporated in the PLS in order to measure the characteristics of the structure. The experimental results agree well with the simulations and show the validity of the modeling for the proposed PLS.
Kazuo IWAMA Akinori KAWACHI Shigeru YAMASHITA
It is known that the original Grover Search (GS) can be modified to use a general value for the phase θ of the diffusion transform. Then, if the number of answers is relatively large, this modified GS can find one of the answers with probability one in a single iteration. However, such a quick and error-free GS can only be possible if we can initially adjust the value of θ correctly against the number of answers, and this seems very hard in usual occasions. A natural question now arises: Can we enjoy a merit even if GS is used without such an adjustment? In this paper, we give a positive answer using the balls-and-bins game in which the random sampling of bins is replaced by the quantum sampling, i.e., a single round of modified GS. It is shown that by using the quantum sampling: (i) The maximum load can be improved quadratically for the static model of the game and this improvement is optimal. (ii) That is also improved to O(1) for the continuous model if we have a certain knowledge about the total number of balls in the bins after the system becomes stable.
Sangheon PACK Byoungwook LEE Yanghee CHOI
In IP-based mobile networks, a few of mobility agents (e.g., home agent, foreign agent, etc.) are used for mobility management. Recently, Hierarchical Mobile IPv6 (HMIPv6) was proposed to reduce signaling overhead and handoff latency occurred in Mobile IPv6. In HMIPv6, a new mobility agent, called mobility anchor point (MAP), is deployed in order to handle binding update procedures locally. However, the MAP can be a single point of performance bottleneck when there are a lot of mobile node (MNs) performing frequent local movements. This is because the MAP takes binding update procedures as well as data packet tunneling. Therefore, it is required to control the number of MNs serviced by a single MAP. In this paper, we propose a load control scheme at the MAP utilizing an admission control algorithm. We name the proposed load control scheme proactive load control scheme to distinct from the existing load control schemes in cellular networks. In terms of admission control, we use the cutoff priority scheme. We develop Markov chain models for the proactive load control scheme and evaluate the ongoing MN dropping and the new MN blocking probabilities. As a result, the proactive load control scheme can reduce the ongoing MN dropping probability while keeping the new MN blocking probability to a reasonable level.
For an efficient software download in cellular CDMA systems, location dependant session admission control (LDSAC) is presented. In the LDSAC scheme, a mobile that is located near cell center can request software download session, but the mobile that is located far from cell center can request session only after approaching near the cell center. Performance is analyzed in terms of handoff rate, mean channel holding time, session blocking probability and handoff forced termination probability. Numerical results show handoff rate between cells in the proposed scheme is reduced to 30-250% compared to conventional scheme, according to traffic characteristics such as terminal speed, session duration time and the size of the allowable zone area in a cell for the initiation of the session. And new session blocking probability decreases slightly, but handoff session forced termination probability decreases drastically.
Ming LEI Hiroshi HARADA Hiromitsu WAKANA Ping ZHANG
In this letter, we investigate the performance of using subband adaptive loading for the combination of orthogonal frequency division multiplexing (OFDM) and adaptive antenna array. The frequency-domain adaptive loading is very effective to deal with the frequency-selective fading which is inevitable in broadband wireless communications. However, almost all of the existing adaptive loading algorithms are based on "subcarrier-to-subcarrier" mode which may results in awfully large signaling overhead, since every subcarrier needs its own signaling loop between the transmitter and receiver. We investigate the performance of the combination of OFDM and adaptive antenna array when a subband adaptive loading algorithm is used to decrease the signaling overhead. It is shown by simulation results that at the cost of some tolerable performance loss, the signaling overhead of adaptive loading can be greatly reduced.
Keita KAWANO Kazuhiko KINOSHITA Koso MURAKAMI
Hierarchical Mobile IPv6 (HMIPv6) has been proposed to manage the mobility of Mobile Terminals (MTs) hierarchically to reduce packet losses during local handover. HMIPv6 uses a mobility manageable router in a domain visited by the MTs to manage the micromobility of the MTs. The router is called Mobility Anchor Point (MAP). As a hierarchical mobility management scheme based on HMIPv6, we have already proposed a multilevel hierarchical distributed IP mobility management scheme to manage the mobility of MTs in a decentralized manner using multiple MAPs. Our scheme manages the mobility of an MT using a MAP having a suitable management domain. This usage of MAPs aims to efficiently decentralize the load of mobility management. Our scheme estimates the movement speed of the MT and then estimates the mobility of the MT based on the estimated movement speed of the MT to achieve the objective. However, recent simulation results obtained with more realistic mobility model indicate that our scheme has two problems in estimating the mobility of MTs: One is that our current scheme misestimates the movement speed of an MT in some cases, and the other is that our current scheme does not notice the changes in the mobility of an MT when the MT decelerates and stays in the same access area for a long time. Thus, an enhanced mobility estimation method is proposed in this paper. The enhanced method has an ability to estimate the movement speed of MTs more correctly and an ability to urge decelerated MTs to degrade their MAP quickly. Finally, the performance of the proposed mobility estimation method is evaluated using simulation experiments. The simulation results show that the enhanced method allows our scheme to estimate the mobility of MTs more correctly and so achieve more efficient load sharing.
Rodrigo Fernandes de MELLO Erico C. T. de MATTOS Luis Carlos TREVELIN Maria Stela Veludo de PAIVA Laurence T. YANG
The availability of a low cost hardware has increased the development of distributed systems, by making then more and more accessible. In order to optimize the resources allocation on the distributed systems, some load balancing algorithms have been proposed. These algorithms distribute the application loads over the environment computers, make homogeneous the occupation of the whole environment and increase the application performance. This equal distribution prevents certain computers to get overloaded, to the detriment of the idleness of the other ones. This article proposes and analyzes the TLBAGrid, a load balancing algorithm for Grid computing environments.
Real-world IP networks are heterogeneous in terms of server and link capacities. A sophisticated and comprehensive load balancing method is essential if we are to avoid congestion in the servers and links of heterogeneous networks. If such a method is not available, network throughput is limited by bottleneck servers or links. This paper proposes an anycast technique that achieves load balancing under heterogeneity. The proposed method well suits implementation on active networks. By taking advantage of the processing ability provided by active nodes, the method can decide packet routes flexibly on the basis of various criteria to realize a variety of load balancing schemes. Some of these schemes can successfully prevent the congestion of heterogeneous networks by tackling bottlenecks in both server and link capacities. The method is also advantageous given its light control load even when using many mirrored servers. Computer simulations confirm the effectiveness of these features.
We propose an adaptive load balancing method for multiple paths that makes it possible to achieve high TCP performance on each path. In conventional load balancing methods, link utilization is the main parameter to be balanced among multiple paths that are established between an ingress and egress node pair. However, when we take into account TCP-level performance, balancing the traffic in terms of only link utilization may not always result in balanced TCP performance on each path. Our method utilizes flow statistics such as the number of active flows in each path, which is easy to measure, and can treat TCP performance. By adaptively equalizing the average bandwidth used per active flow in each path, which is calculated by dividing the input rate to the path by the mean number of active flows, our method achieves fair and high TCP performance on each path. Unlike other methods, intermediate nodes between an ingress-egress pair are not required to perform traffic controls or measurements besides normal packet forwarding. We describe a load balancing method for adaptively equalizing the average bandwidth used per active flow on each path and show its effectiveness under heterogeneous conditions through simulation analysis.
Dejiang JIN Sotirios G. ZIAVRAS
The multiplication of large spare matrices is a basic operation in many scientific and engineering applications. There exist some high-performance library routines for this operation. They are often optimized based on the target architecture. For a parallel environment, it is essential to partition the entire operation into well balanced tasks and assign them to individual processing elements. Most of the existing techniques partition the given matrices based on some kind of workload estimation. For irregular sparse matrices on PC clusters, however, the workloads may not be well estimated in advance. Any approach other than run-time dynamic partitioning may degrade performance. In this paper, we apply our super-programming approach to parallel large matrix multiplication on PC clusters. In our approach, tasks are partitioned into super-instructions that are dynamically assigned to member computer nodes. Thus, the load balancing logic is separated from the computing logic; the former is taken over by the runtime environment. Our super-programming approach facilitates ease of program development and targets high efficiency in dynamic load balancing. Workloads can be balanced effectively and the optimization overhead is small. The results prove the viability of our approach.