Nec Usa, Inc.

Bhaskar Sengupta

Duan-Shin Lee

Gopalakrishnan Ramamurthy

Rajiv S Dighe

Subir K Biswas

Qiang Ren

Rauf Izmailov

Shyamal Chowdhury

A technique for reducing ATM call blocking is provided that splits wideband connections into multiple low-bandwidth sub-connections and routing them independently through the network. Fragmented network bandwidth is used for supporting calls which are otherwise blocked by conventional routing. A detailed cell-level design for the split scheduling algorithms is provided. A system for implementing splitting without requiring any protocol changes within the network is provided. Such a system modifies the control plane protocols only within the end-stations.

Connection splitting: an efficient way of reducing call blocking in ATM

Alternative routing schemes in high-speed networks ensure fault tolerance in the event of a link failure. An heuristic algorithm, based on maximal-flow algorithm, is first used to partition the network into parts: one part carries regular or primary traffic and another part is reserved to carry re-routed traffic in the event of a link failure. The result of applying the algorithm is used to find alternative routes for a given call at the time of call setup. The scheme can be modified to accommodate hop-court limitations and loop avoidance.

Link based alternative routing scheme for network restoration under

Methods of traffic smoothing for frames of different video sources where each video source transmits frames at fixed intervals and the network is free to decide the relative temporal spacing of video frames from different sources provide significant performance advantages. The time at which a given source begins to transmit its first frame is under the control of the network; however, thereafter, all frames from the source are transmitted at fixed intervals. Two heuristic and one optimization method control the temporal placement of the video frames from different sources in order to reduce the loss rate of the high priority cells and to minimize the smoothness index of the traffic.

Temporal placement control of video frames in B-ISDN networks

A satellite communication system increases its efficiency by using adaptive pipeline polling as the control protocol. In this protocol, the earth stations are separated into active and inactive sets and a first number of slots (salary) is assigned to each active station. Additionally, periodically active stations are queried as to the length of their queues and assigned salary slots not needed by any active station are reallofted to stations with queues that exceed their assigned salary in proportion to the size of such excess. A contention period is included outside of the normal salary and bonus periods during which stations in the inactive are allowed to contend to be included in the active set.

Satellite communication system

An asynchronous transfer mode switching system uses a parallel array of switching planes, distributors for distributing cells between the switching planes, and resequencers for collecting the cells after they have been switched. To balance the loads in the output buffers of the switching planes, each distributor includes two one-cell buffers and a load matrix. The load matrix stores the state of the output buffers in the switching planes, and this information is used to decide whether to send a cell arriving at the first of the one-cell buffers to its scheduled switching plane or to store the cell in the other of the one-cell buffers until either a switching plane with a lightly loaded output buffer for the cell's destination becomes available or another later arriving cell needs to be stored there.

Threshold-based load balancing in ATM switches with parallel switch

To control congestion in packet switching networks, control of the traffic sent by a given station to each of the downstream nodes to which it is directly connected is effected by control of the traffic that the upstream nodes to which it is directly connected are permitted to send to it. In this regard, a predictive model is used to predict the cross traffic, one round trip delay in advance that the given station can expect. The parameters for the predictive model are obtained by measurements in real time and by the use of moving averages. Using the predicted cross traffic, the amount of controlled traffic that the proximate downstream nodes can accommodate from the given node, and the correct state of the given node, the state of the given node one round trip delay into the future is predicted. This prediction is used to schedule the amount of traffic to be sent by each of its proximate upstream nodes.

Growing community of inventors

Cranbury, NJ, United States of America

Bhaskar Sengupta