37 results about "Bottleneck bandwidth" patented technology

A novel packet filtering and processing mechanism solves bandwidth estimation problems related to cross-traffic, host limitations, and fabric-specific issues in order to accurately estimate bottleneck bandwidth in fluctuating environments such as home networks. In particular, the process controls the number of packet train experiments, the size of trains and packets within trains, and other factors to eliminate errors due to cross-traffic, network capture, receiver CPU overload, interrupt moderation, and other networking factors. In an embodiment of the invention, a history window is applied to estimates derived from individual trains to ensure that data used to make a bottleneck bandwidth estimate was taken at a time when the network is relatively stable.

The invention discloses a multi-service multi-path routing optimization method in a named data network, which divides the services in the NDN network into delay-sensitive services and bandwidth-sensitive services; edge nodes mark the corresponding differentiated service code point values ​​to the services of interest packets In the type field; route selection and forwarding of the interest packet; after the content publisher receives the interest packet, it generates a corresponding data packet, and updates the core node’s Pheromone concentration. The invention can classify multi-services in the network, and use ant colony optimization to optimize and select routes for different types of services according to their measurement requirements, so as to well support NDN network QoS and realize differentiated services.

This invention relates to measure method and system for bottle wide and remain wide band in the label barrel speed limit network path, wherein, the method comprises the following steps: establishing send end and receive end for data interacting; firstly detecting and judging whether the linkage is label speed limit network path; if yes, then computing linkage each label parameters, bottle and remain wide band; if not adopting traditional method to measure and compute the bottle wide band and remain wide band. The invention is free of label speed limited.

The invention discloses a fully distributed repair method based on an erasure code, and aims to improve the repair efficiency of invalid data. The technical solution is to build a distributed storage system composed of control nodes and storage nodes. Data writing management programs and data repair management programs are installed on the control nodes; data writing programs and data repair programs are installed on the storage nodes; data writing management programs The program divides and encodes the data object into blocks, and writes the original data block and the redundant data block into different storage nodes; the data repair management program is the invalid data block D' i Select a new storage node LN; the data repair management program and the data repair program cooperate with each other to obtain the successfully repaired data sub-blocks after decoding and calculation, and store them in LN; the data repair program of LN merges the data sub-blocks to obtain D' i recovered D. i . The invention can make full use of the upstream bandwidth of each storage node, effectively avoid the bottleneck bandwidth in the network, and improve the recovery efficiency of invalid data.

The invention discloses an integrated routing setting method and an integrated routing setting device. The method comprises the following steps of: converting a physical topology crossing different wavelength division multiplexing networks into a multi-fiber auxiliary chart; dividing a bandwidth of a request service into an integral bandwidth and a decimal bandwidth; determining a cost function of each chain between a source node and a destination node of the request service; seeking k shortest integral bandwidth alternative channels for the integral bandwidth part, and selecting one channel with the maximum bottleneck bandwidth from the k shortest integral bandwidth alternative channels as a label switching path of the integral bandwidth part; and seeking k shortest decimal bandwidth alternative channels for the decimal bandwidth part, and selecting one channel with the minimum path influence function value from the k shortest decimal bandwidth alternative channels as a label switching path of the decimal bandwidth part. By the method and the device, the bandwidth of the service is divided into an integral bandwidth part and a decimal bandwidth part; and the label switching pathsare determined for the two bandwidth parts according to the cost function and a path influence function respectively.

The present invention provides an available bandwidth measurement method and device. The method comprises the following steps: a sending end sends at least two test data packets to a receiving end of a to-be-measured transmission path in sequence after a first packet sending time interval, wherein the first packet sending time interval is used for making the throughput at the transmission bottleneck of the to-be-measured transmission path reaches the bottleneck bandwidth at the transmission bottleneck; the sending end receives response data packets of the at least two test data packets sent by the receiving end, wherein the length of each response data packet is smaller than a preset length threshold; the sending end determines an actual receiving rate of the to-be-measured transmission path on an uplink direction from the sending end to the receiving end according to the response times of the response data packets of the at least two test data packets, wherein the response time of the response data packet of any test data packet is the time from the sending moment of the test data packet to the arrival of the response data packet at the sending end; and the available bandwidth of the to-be-measured transmission path on the uplink direction is determined according to the actual receiving rate.

Embodiments of the application disclose a bottleneck bandwidth measurement method and apparatus and a storage medium. The method comprises: a first data end receives m data packet sets sent successively by a second data end to the first data end, wherein m is greater than or equal to 1; the first data end obtains, from each data packet set, a group of consecutive data packets as a first subset of the data packet set to obtain m first subsets, wherein transmission intervals of the obtained group of consecutive data packets satisfy a preset interval; and the first data end obtains, according to the total number of bytes of each first subset and a time taken for receiving the first subset, a bottleneck bandwidth between the first data end and the second data end. The embodiments of the application solve the technical problem of a large measurement error of a bottleneck bandwidth between two ends.

The invention discloses a multi-path TCP congestion control method. The method comprises the steps that firstly, substream weight factors for describing network environment features of different substreams in a multi-path TCP are introduced into a google BBR congestion control algorithm; secondly, convergence features of the different substreams of different multi-path TCP sessions at a shared bottleneck link when the different substreams deploy the congestion control algorithm are analyzed, and a weight-bandwidth model is built; thirdly, bottleneck bandwidth parameters and background stream parameters for recovering the shared bottleneck link weight-bandwidth model are calculated; and finally, traffic migration is achieved by taking the network fairness as a principle and utilizing a distributed substream weight factor updating iteration algorithm according to bottleneck bandwidth parameters and background stream parameters which are reproduced by the various substreams at different shared bottleneck links, and finally the network fairness guarantee is achieved.