A Primary-Backup Channel Approach to Dependable Real-Time Communication in Multihop Networks
IEEE Transactions on Computers
A new distributed route selection approach for channel establishment in real-time networks
IEEE/ACM Transactions on Networking (TON)
Distributed computation on graphs: shortest path algorithms
Communications of the ACM
SOSP '01 Proceedings of the eighteenth ACM symposium on Operating systems principles
Distributed Algorithms
IEEE/ACM Transactions on Networking (TON)
An Efficient Resource Sharing Scheme for Dependable Real-Time Communication in Multihop Networks
HiPC '02 Proceedings of the 9th International Conference on High Performance Computing
Hi-index | 0.00 |
Several distributed real-time applications require faulttolerance apart from guaranteed timeliness, at acceptable levels of overhead. These applications require hard guarantees on recovery delays, due to network component failures, which cannot be ensured in traditional datagram services. Several schemes exist which attempt to guarantee failure recovery in a timely and resource efficient manner. These methods center around a priori reservation of network resources called spare resources along a backup channel, in addition to each primary communication channel. This backup channel is usually routed along a path disjoint with the primary channel. In this paper, we propose a distributed method of segmented backups for dependable realtime communication in multihop networks, which improves upon existing methods in terms of network resource utilization, average call acceptance rate, scalability and provides better QoS guarantees on bounded failure recovery time and propagation delays, without any compromise in faulttolerance levels. The distributed algorithm is one of finding a "minimal path" based on flooding with a cut-out mechanism that does not relay messages if they came along longer paths than those known. We further show that the complexity of the distributed algorithm is bounded and acceptable.