| Creator |
|
|
|
|
|
|
|
|
|
| Language |
|
| Publisher |
|
|
|
| Date |
|
| Source Title |
|
| Source Title |
|
| Vol |
|
| Issue |
|
| First Page |
|
| Last Page |
|
| Publication Type |
|
| Access Rights |
|
| Rights |
|
| Related DOI |
|
| Related DOI |
|
|
|
|
|
| Related URI |
|
|
|
|
|
| Related URI |
|
| Related HDL |
|
| Relation |
|
|
|
|
|
| Abstract |
本稿では，通信性能の悪化の主要因である通信の衝突を避けるためのタスク配置最適化の評価を行った．著者らはメッセージのタイミングを制御しつつ，衝突を回避するタスク配置最適化の研究をしている．その中でその最適化をツリートポロジに適用して，その手法が有効であることを示した．一方，メッシュやトーラスなどのネットワークトポロジにはホップ数を評価関数とする通信衝突削減のためのタスク配置最適化が研究なされており，...有効性があるとされている．そこで，ネットワークトポロジが3Dメッシュの際に通信衝突回数を評価関数としたタスク配置最適化を適用した場合とホップ数を評価関数したタスク配置最適化を適用した場合にどのような違いがあるか調べる実験をし，考察を行った．
In this text, we evaluated the optimization of task allocation to avoid contentions that was the key factor of the communication performance degradation. We applied the optimization of task allocation controlling the timing of the message for avoiding contentions to the tree topology, and showed it was effectiveness. On the other hand, there were some optimizations of task allocation for reducing contentions. Those optimizations used the evaluation function which used the number of hops. Those optimizations against the mesh and torus topology were effective. We experimented and investigated what's the difference between the optimization of task allocation which the evaluation function was the number of contentions and the number of hops when the network topology was 3D mesh. We considered about it.show more
|
Export Mendeley
morie07_4