摘要:
According to one exemplary embodiment, a method for determining the forward and reverse available capacity or tight link capacity of an IP path from a single endpoint includes the steps of: transmitting, from a source IP endpoint node toward a destination IP endpoint node, a forward packet train including a first plurality of IP test packets over the forward IP path; and receiving, at the source IP endpoint node, a corresponding reverse packet train from the destination IP endpoint node, the reverse packet train including a second plurality of IP test packets over the reverse IP path each of which correspond to a respective one of the first plurality of IP test packets. For those IP path capacity measurements embodiments which are TWAMP-based, no changes are needed to the TWAMP control protocol, e.g., since exemplary embodiments make use of padding octets to transfer additional information which can be used for available IP path capacity and tight IP link capacity calculations.
摘要:
According to one exemplary embodiment, a method for determining the forward and reverse available capacity or tight link capacity of an IP path from a single endpoint includes the steps of: transmitting, from a source IP endpoint node toward a destination IP endpoint node, a forward packet train including a first plurality of IP test packets over the forward IP path; and receiving, at the source IP endpoint node, a corresponding reverse packet train from the destination IP endpoint node, the reverse packet train including a second plurality of IP test packets over the reverse IP path each of which correspond to a respective one of the first plurality of IP test packets. For those IP path capacity measurements embodiments which are TWAMP-based, no changes are needed to the TWAMP control protocol, e.g., since exemplary embodiments make use of padding octets to transfer additional information which can be used for available IP path capacity and tight IP link capacity calculations.
摘要:
A method for queuing data packets is described. Upon receipt of a data packet, e.g., at an egress buffer of a node, a position within an output queue at which to place the data packet is determined. The queue position for the data packet can be determined based on various criteria, e.g., the type of the data packet and a time associated with its subsequent transmission, and timing aspects associated with the data packets which are already stored in the queue.
摘要:
A probing method, arrangement, and system for estimating a condition, such as available bandwidth, of a data path in a communication network including multiple nodes are described. A probe-packet sender node transmits probe packets with a sent inter-packet separation and a time-to-live towards an intermediate node. Control message packets are generated by the intermediate node when the time-to-live expires. The probe-packet sender node receives the control message packets and calculates a received inter-packet separation. An estimate of the data path condition can be generated based on the sent inter-packet separation and the received inter-packet separation.
摘要:
A system and method for determining end-to-end available bandwidth of a path in a packet-switched communication network. A multi-rate packet chirp injector injects probe traffic packets at different probing rates, u, into the network utilizing a probe chirp. A receiver samples time separation of the probe traffic packets to generate time-separation samples, which are used to calculate relative inter-packet separation strain, ε, samples at different probing rates. An Expectation-Maximization, EM, analyzer utilizes an EM algorithm to identify useful strain ε samples and to estimate parameters of a line l2 representing a direct linear relationship between the probing rate and the inter-packet separation strain ε at probing rates greater than a break point. A filter such as, for example, a Kalman filter then determines the available bandwidth of the network path based on the estimated parameters of the line l2.
摘要:
A method for queuing data packets is described. Upon receipt of a data packet, e.g., at an egress buffer of a node, a position within an output queue at which to place the data packet is determined. The queue position for the data packet can be determined based on various criteria, e.g., the type of the data packet and a time associated with its subsequent transmission, and timing aspects associated with the data packets which are already stored in the queue.
摘要:
A probing method, arrangement, and system for estimating a condition, such as available bandwidth, of a data path in a communication network including multiple nodes are described. A probe-packet sender node transmits probe packets with a sent inter-packet separation and a time-to-live towards an intermediate node. Control message packets are generated by the intermediate node when the time-to-live expires. The probe-packet sender node receives the control message packets and calculates a received inter-packet separation. An estimate of the data path condition can be generated based on the sent inter-packet separation and the received inter-packet separation.
摘要:
The present invention relates to a method and arrangement for estimating available bandwidth of sub-path of a data path. The solution is based on the sending of probe packet trains comprising inner and outer packets from a start node to an end node using different TTL values for the outer and inner packets of these probe trains. By comparing measurements using different TTL values of the inner packets, it is also possible to determine which link are the local tight link and what available bandwidth it has.
摘要:
A system and method for determining end-to-end available bandwidth of a path in a packet-switched communication network. A multi-rate packet chirp injector injects probe traffic packets at different probing rates, u, into the network utilizing a probe chirp. A receiver samples time separation of the probe traffic packets to generate time-separation samples, which are used to calculate relative inter-packet separation strain, ε, samples at different probing rates. An Expectation-Maximization, EM, analyzer utilizes an EM algorithm to identify useful strain ε samples and to estimate parameters of a line l2 representing a direct linear relationship between the probing rate and the inter-packet separation strain ε at probing rates greater than a break point. A filter such as, for example, a Kalman filter then determines the available bandwidth of the network path based on the estimated parameters of the line l2.
摘要:
A method of measuring network performance is disclosed. A payload packet is transmitted from a sending node to a receiving node via a communication network. A reference packet is transmitted to the receiving node in connection with sending of the payload packet. Receive information related to the payload packet and to the reference packet is generated, at the receiving node. Send information related to the payload packet and to the reference packet is received, by the receiving node. Network performance is estimated by utilizing the send and receive information.