Synchronizing clocks in the presence of faults
Journal of the ACM (JACM)
Ensuring Fault Tolerance of Phase-Locked Clocks
IEEE Transactions on Computers
Journal of the ACM (JACM)
A new fault-tolerant algorithm for clock synchronization
Information and Computation
Performance Analysis of k-ary n-cube Interconnection Networks
IEEE Transactions on Computers
The network architecture of the Connection Machine CM-5 (extended abstract)
SPAA '92 Proceedings of the fourth annual ACM symposium on Parallel algorithms and architectures
Dynamic fault-tolerant clock synchronization
Journal of the ACM (JACM)
Performance Analysis of Mesh Interconnection Networks with Deterministic Routing
IEEE Transactions on Parallel and Distributed Systems
Interleaved All-to-All Reliable Broadcast on Meshes and Hypercubes
IEEE Transactions on Parallel and Distributed Systems
Probabilistic Clock Synchronization in Distributed Systems
IEEE Transactions on Parallel and Distributed Systems
Real-Time Communication in Multihop Networks
IEEE Transactions on Parallel and Distributed Systems
| Hi-index | 0.00 |
In this paper, we present a new efficient hybrid time synchronization scheme for a mesh or torus interconnection networks, called ROCTS. ROCTS comprises two levels, one for the lower level that is implemented over a special high-speed ring array, one for the mesh or torus network. In ROCTS, the second network we construct is different from the past, which is a ring array with each ring not connected to any other. So, we can implement ROCTS in a distributed and fault-tolerant way. On the other hand, we combine the advantages of hardware and linear-envelope methods so that we can improve the performance of time synchronization even if the work load on a mesh or torus network is not uniform or even the blocked nodes occur. In this paper, we also analyze the performance of ROCTS and present the maximum clock skew and the successful time synchronization probability so that we prove that ROCTS can guarantee a much smaller bound on the clock skew, e.g. 10us for 300*300 mesh or torus networks, than most existing time synchronization schemes.