The Experts below are selected from a list of 12948 Experts worldwide ranked by ideXlab platform
Atsushi Iwata - One of the best experts on this subject based on the ideXlab platform.
-
Hot Interconnects - High-Speed, Short-Latency Multipath Ethernet Transport for Interconnections
2008 16th IEEE Symposium on High Performance Interconnects, 2008Co-Authors: Nobuyuki Enomoto, Hideyuki Shimonishi, Junichi Higuchi, Takashi Yoshikawa, Atsushi IwataAbstract:In this paper, we propose an Ethernet-based transmission-guaranteed, congestion-controlled network using a simplified multi-path Aggregation scheme. Multi-path Aggregation increases throughput by multiplying the bandwidth of a single path by the number of paths. There are several Aggregation schemes, such as Link Aggregation and Multi-path TCP. However, Link Aggregation is unable to utilize multiple paths to increase throughput because it distributes a single flow only into a single path, The Multi-path TCP scheme requires managing two stages of sequence numbers (SEQ) (i.e., the SEQ assigned to a path and the SEQ assigned to the flow). This complexity requires software-based implementation, which means the method fails to provide high throughput and short latency. We therefore propose a single-stage sequence number scheme with a reorder-buffer-usage-based retransmission-activating algorithm. Hardware implementation of this scheme is very simple. In addition, packet-loss detection is rapid. Our simulation and experimental results showed that our scheme provided a high-throughput, short-latency, and transmission-guaranteed network over conventional, lossy, delay-fluctuated, and best-effort Ethernet.
-
GLOBECOM - High-Speed, Short-Latency Multipath Ethernet for Data Center Area Communications
IEEE GLOBECOM 2008 - 2008 IEEE Global Telecommunications Conference, 2008Co-Authors: Nobuyuki Enomoto, Hideyuki Shimonishi, Junichi Higuchi, Takashi Yoshikawa, Atsushi IwataAbstract:In this paper, we propose a simplified multi-path Aggregation scheme, Ethernet with flow label for multi-path, called EFL-MP. There are several Aggregation schemes, such as Link Aggregation and multi-path TCP, for multi-path load balancing. However, Link Aggregation does not distribute a single flow into multiple paths due to its MAC address based flow dispatch. In addition, the multipath TCP scheme is not suitable for implementing into hardware because of the implementation complexity caused by the management of the two stages of sequence number (SEQ) (i.e., the SEQ assigned to a path and the SEQ assigned to the flow); thus it fails to provide high throughput and short latency. To maximize data transfer throughput and minimize latency, we employed a single-stage sequence number scheme with a reorder-buffer-usage-based retransmission activating algorithm. Simulation results show that EFL-MP provides high throughput and short latency despite its simple scheme, and its throughput and its latency suit to use for interconnection transport. In addition, we demonstrate that our proposed retransmission activating algorithm activates retransmission due to packet losses earlier than the conventional timer-based retransmission activating algorithm.
Nobuyuki Enomoto - One of the best experts on this subject based on the ideXlab platform.
-
Hot Interconnects - High-Speed, Short-Latency Multipath Ethernet Transport for Interconnections
2008 16th IEEE Symposium on High Performance Interconnects, 2008Co-Authors: Nobuyuki Enomoto, Hideyuki Shimonishi, Junichi Higuchi, Takashi Yoshikawa, Atsushi IwataAbstract:In this paper, we propose an Ethernet-based transmission-guaranteed, congestion-controlled network using a simplified multi-path Aggregation scheme. Multi-path Aggregation increases throughput by multiplying the bandwidth of a single path by the number of paths. There are several Aggregation schemes, such as Link Aggregation and Multi-path TCP. However, Link Aggregation is unable to utilize multiple paths to increase throughput because it distributes a single flow only into a single path, The Multi-path TCP scheme requires managing two stages of sequence numbers (SEQ) (i.e., the SEQ assigned to a path and the SEQ assigned to the flow). This complexity requires software-based implementation, which means the method fails to provide high throughput and short latency. We therefore propose a single-stage sequence number scheme with a reorder-buffer-usage-based retransmission-activating algorithm. Hardware implementation of this scheme is very simple. In addition, packet-loss detection is rapid. Our simulation and experimental results showed that our scheme provided a high-throughput, short-latency, and transmission-guaranteed network over conventional, lossy, delay-fluctuated, and best-effort Ethernet.
-
GLOBECOM - High-Speed, Short-Latency Multipath Ethernet for Data Center Area Communications
IEEE GLOBECOM 2008 - 2008 IEEE Global Telecommunications Conference, 2008Co-Authors: Nobuyuki Enomoto, Hideyuki Shimonishi, Junichi Higuchi, Takashi Yoshikawa, Atsushi IwataAbstract:In this paper, we propose a simplified multi-path Aggregation scheme, Ethernet with flow label for multi-path, called EFL-MP. There are several Aggregation schemes, such as Link Aggregation and multi-path TCP, for multi-path load balancing. However, Link Aggregation does not distribute a single flow into multiple paths due to its MAC address based flow dispatch. In addition, the multipath TCP scheme is not suitable for implementing into hardware because of the implementation complexity caused by the management of the two stages of sequence number (SEQ) (i.e., the SEQ assigned to a path and the SEQ assigned to the flow); thus it fails to provide high throughput and short latency. To maximize data transfer throughput and minimize latency, we employed a single-stage sequence number scheme with a reorder-buffer-usage-based retransmission activating algorithm. Simulation results show that EFL-MP provides high throughput and short latency despite its simple scheme, and its throughput and its latency suit to use for interconnection transport. In addition, we demonstrate that our proposed retransmission activating algorithm activates retransmission due to packet losses earlier than the conventional timer-based retransmission activating algorithm.
Hideyuki Shimonishi - One of the best experts on this subject based on the ideXlab platform.
-
Hot Interconnects - High-Speed, Short-Latency Multipath Ethernet Transport for Interconnections
2008 16th IEEE Symposium on High Performance Interconnects, 2008Co-Authors: Nobuyuki Enomoto, Hideyuki Shimonishi, Junichi Higuchi, Takashi Yoshikawa, Atsushi IwataAbstract:In this paper, we propose an Ethernet-based transmission-guaranteed, congestion-controlled network using a simplified multi-path Aggregation scheme. Multi-path Aggregation increases throughput by multiplying the bandwidth of a single path by the number of paths. There are several Aggregation schemes, such as Link Aggregation and Multi-path TCP. However, Link Aggregation is unable to utilize multiple paths to increase throughput because it distributes a single flow only into a single path, The Multi-path TCP scheme requires managing two stages of sequence numbers (SEQ) (i.e., the SEQ assigned to a path and the SEQ assigned to the flow). This complexity requires software-based implementation, which means the method fails to provide high throughput and short latency. We therefore propose a single-stage sequence number scheme with a reorder-buffer-usage-based retransmission-activating algorithm. Hardware implementation of this scheme is very simple. In addition, packet-loss detection is rapid. Our simulation and experimental results showed that our scheme provided a high-throughput, short-latency, and transmission-guaranteed network over conventional, lossy, delay-fluctuated, and best-effort Ethernet.
-
GLOBECOM - High-Speed, Short-Latency Multipath Ethernet for Data Center Area Communications
IEEE GLOBECOM 2008 - 2008 IEEE Global Telecommunications Conference, 2008Co-Authors: Nobuyuki Enomoto, Hideyuki Shimonishi, Junichi Higuchi, Takashi Yoshikawa, Atsushi IwataAbstract:In this paper, we propose a simplified multi-path Aggregation scheme, Ethernet with flow label for multi-path, called EFL-MP. There are several Aggregation schemes, such as Link Aggregation and multi-path TCP, for multi-path load balancing. However, Link Aggregation does not distribute a single flow into multiple paths due to its MAC address based flow dispatch. In addition, the multipath TCP scheme is not suitable for implementing into hardware because of the implementation complexity caused by the management of the two stages of sequence number (SEQ) (i.e., the SEQ assigned to a path and the SEQ assigned to the flow); thus it fails to provide high throughput and short latency. To maximize data transfer throughput and minimize latency, we employed a single-stage sequence number scheme with a reorder-buffer-usage-based retransmission activating algorithm. Simulation results show that EFL-MP provides high throughput and short latency despite its simple scheme, and its throughput and its latency suit to use for interconnection transport. In addition, we demonstrate that our proposed retransmission activating algorithm activates retransmission due to packet losses earlier than the conventional timer-based retransmission activating algorithm.
Junichi Higuchi - One of the best experts on this subject based on the ideXlab platform.
-
Hot Interconnects - High-Speed, Short-Latency Multipath Ethernet Transport for Interconnections
2008 16th IEEE Symposium on High Performance Interconnects, 2008Co-Authors: Nobuyuki Enomoto, Hideyuki Shimonishi, Junichi Higuchi, Takashi Yoshikawa, Atsushi IwataAbstract:In this paper, we propose an Ethernet-based transmission-guaranteed, congestion-controlled network using a simplified multi-path Aggregation scheme. Multi-path Aggregation increases throughput by multiplying the bandwidth of a single path by the number of paths. There are several Aggregation schemes, such as Link Aggregation and Multi-path TCP. However, Link Aggregation is unable to utilize multiple paths to increase throughput because it distributes a single flow only into a single path, The Multi-path TCP scheme requires managing two stages of sequence numbers (SEQ) (i.e., the SEQ assigned to a path and the SEQ assigned to the flow). This complexity requires software-based implementation, which means the method fails to provide high throughput and short latency. We therefore propose a single-stage sequence number scheme with a reorder-buffer-usage-based retransmission-activating algorithm. Hardware implementation of this scheme is very simple. In addition, packet-loss detection is rapid. Our simulation and experimental results showed that our scheme provided a high-throughput, short-latency, and transmission-guaranteed network over conventional, lossy, delay-fluctuated, and best-effort Ethernet.
-
GLOBECOM - High-Speed, Short-Latency Multipath Ethernet for Data Center Area Communications
IEEE GLOBECOM 2008 - 2008 IEEE Global Telecommunications Conference, 2008Co-Authors: Nobuyuki Enomoto, Hideyuki Shimonishi, Junichi Higuchi, Takashi Yoshikawa, Atsushi IwataAbstract:In this paper, we propose a simplified multi-path Aggregation scheme, Ethernet with flow label for multi-path, called EFL-MP. There are several Aggregation schemes, such as Link Aggregation and multi-path TCP, for multi-path load balancing. However, Link Aggregation does not distribute a single flow into multiple paths due to its MAC address based flow dispatch. In addition, the multipath TCP scheme is not suitable for implementing into hardware because of the implementation complexity caused by the management of the two stages of sequence number (SEQ) (i.e., the SEQ assigned to a path and the SEQ assigned to the flow); thus it fails to provide high throughput and short latency. To maximize data transfer throughput and minimize latency, we employed a single-stage sequence number scheme with a reorder-buffer-usage-based retransmission activating algorithm. Simulation results show that EFL-MP provides high throughput and short latency despite its simple scheme, and its throughput and its latency suit to use for interconnection transport. In addition, we demonstrate that our proposed retransmission activating algorithm activates retransmission due to packet losses earlier than the conventional timer-based retransmission activating algorithm.
Takashi Yoshikawa - One of the best experts on this subject based on the ideXlab platform.
-
Hot Interconnects - High-Speed, Short-Latency Multipath Ethernet Transport for Interconnections
2008 16th IEEE Symposium on High Performance Interconnects, 2008Co-Authors: Nobuyuki Enomoto, Hideyuki Shimonishi, Junichi Higuchi, Takashi Yoshikawa, Atsushi IwataAbstract:In this paper, we propose an Ethernet-based transmission-guaranteed, congestion-controlled network using a simplified multi-path Aggregation scheme. Multi-path Aggregation increases throughput by multiplying the bandwidth of a single path by the number of paths. There are several Aggregation schemes, such as Link Aggregation and Multi-path TCP. However, Link Aggregation is unable to utilize multiple paths to increase throughput because it distributes a single flow only into a single path, The Multi-path TCP scheme requires managing two stages of sequence numbers (SEQ) (i.e., the SEQ assigned to a path and the SEQ assigned to the flow). This complexity requires software-based implementation, which means the method fails to provide high throughput and short latency. We therefore propose a single-stage sequence number scheme with a reorder-buffer-usage-based retransmission-activating algorithm. Hardware implementation of this scheme is very simple. In addition, packet-loss detection is rapid. Our simulation and experimental results showed that our scheme provided a high-throughput, short-latency, and transmission-guaranteed network over conventional, lossy, delay-fluctuated, and best-effort Ethernet.
-
GLOBECOM - High-Speed, Short-Latency Multipath Ethernet for Data Center Area Communications
IEEE GLOBECOM 2008 - 2008 IEEE Global Telecommunications Conference, 2008Co-Authors: Nobuyuki Enomoto, Hideyuki Shimonishi, Junichi Higuchi, Takashi Yoshikawa, Atsushi IwataAbstract:In this paper, we propose a simplified multi-path Aggregation scheme, Ethernet with flow label for multi-path, called EFL-MP. There are several Aggregation schemes, such as Link Aggregation and multi-path TCP, for multi-path load balancing. However, Link Aggregation does not distribute a single flow into multiple paths due to its MAC address based flow dispatch. In addition, the multipath TCP scheme is not suitable for implementing into hardware because of the implementation complexity caused by the management of the two stages of sequence number (SEQ) (i.e., the SEQ assigned to a path and the SEQ assigned to the flow); thus it fails to provide high throughput and short latency. To maximize data transfer throughput and minimize latency, we employed a single-stage sequence number scheme with a reorder-buffer-usage-based retransmission activating algorithm. Simulation results show that EFL-MP provides high throughput and short latency despite its simple scheme, and its throughput and its latency suit to use for interconnection transport. In addition, we demonstrate that our proposed retransmission activating algorithm activates retransmission due to packet losses earlier than the conventional timer-based retransmission activating algorithm.
