A new interconnection network for massively parallel computing is introduced. This network is called a hierarchal optical ring interconnection (HORN). The HORN consists of a single-hop, scalable, constant-degree, strictly nonblocking, fault-tolerant interconnection topology that uses wavelength-division multiple access to provide better utilization of the terahertz bandwidth offered by optics. ...

ÐIn this paper, we study distributed path reservation protocols for multiplexed all-optical interconnection networks. The path reservation protocols negotiate the reservation and establishment of connections that arrive dynamically to the network. These protocols can be applied to both wavelength division multiplexing (WDM) and time division multiplexing (TDM) networks. Two classes of protocols...

we consider a grid network where nodes contain small buffers. a packet that faces a crowded buffer in its route will get extra latency and may be dropped. in this paper, we propose a novel flow control protocol called rfcc for grid networks. rfcc tries to reroute delayed packets and utilizes network coding to introduce a configurable amount of redundant information in the network, thereby incre...

We investigate the problem of maximizing the number of processors in a massively parallel computer when the degree of the internal nodes and the diameter of the network are physically constrained. The solution we propose for this problem is a fat-tree with processors at the leaves. The basic building block of this fat-tree is a two-level fat-tree. This two-level fat-tree is obtained >from compl...

the structure of a self-routing interconnection network with a recurrent neural network," in Pro-Abstract A modiied Recurrent Neural Network (RNN) is used to learn a Self-Routing Interconnection Network (SRIN) from a set of routing examples. The RNN is modiied so that it has several distinct initial states. This is equivalent to a single RNN learning multiple diierent synchronous sequential mac...

This paper describes the design and use of the Interconnection Network Simulator (ICNS) framework. ICNS is a modular, object-oriented simulation system that has been developed to investigate performance issues in multiprocessor interconnection networks that exploit photonic technology in their design. We describe the ICNS infrastructure, present two distinct photonic interconnection networks th...

The generalized de Bruijn digraph GB(n,d) has good properties as an interconnection network topology. The resource location problem in an interconnection network is one of the facility location problems. Finding absorbants of a digraph corresponds to solving a kind of resource location problem. In this paper, we solve some open problems given in the article ”Erfang Shan, T.C.E.Cheng, Liying Kan...

The Optical Transpose Interconnection System; OTIS for short; optoelectronic interconnection network architecture is an interesting alternative for future-generation parallel processing systems due to its promising and attractive topological properties. Recently, many enhancements were presented on OTIS topologies including the Enhanced OTIS-nCube interconnection networks. Therefore, performanc...

Previous work on network interconnection has tended to overlook that both the sender and receiver of an electronic message take actions, bear costs, and derive benefits from the message exchange. In a simple model with two-sided benefits and fixed network architectures, we find that the socially optimal interconnection charge is independent of the “direction” of the message and is used to induc...