A Relay-Assisted Communication Scheme for Collaborative On-Device CNN Execution Considering Hybrid Parallelism
Deep learning (DL) has gained increasing prominence in latency-critical artificial intelligence (AI) applications. Due to the intensive computational requirements of these applications, cloud-centric approaches have been attempted to address this issue, but they result in intolerable latency, networ...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2023-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10247012/ |
| _version_ | 1797680926070669312 |
|---|---|
| author | Emre Kilcioglu Ivan Stupia Luc Vandendorpe |
| author_facet | Emre Kilcioglu Ivan Stupia Luc Vandendorpe |
| author_sort | Emre Kilcioglu |
| collection | DOAJ |
| description | Deep learning (DL) has gained increasing prominence in latency-critical artificial intelligence (AI) applications. Due to the intensive computational requirements of these applications, cloud-centric approaches have been attempted to address this issue, but they result in intolerable latency, network congestion, and privacy concerns. An alternative concept called edge intelligence, which combines AI and edge computing, has been proposed to perform DL execution at the edge in multiple resource-constrained devices (RCDs) collaboratively. This paper proposes a relay-assisted, distributed, and collaborative on-device convolutional neural network (CNN) execution scheme for latency-critical applications. The system employs hybrid parallelism, combining both data and model parallelism, to optimize collaborative CNN execution on RCDs. A relay-assisted communication technique is used to reduce the input data size per RCD and avoid excessive point-to-point communication between the data owner RCD and collaborative RCDs. The proposed approach reduces communication overhead using two strategies: layer block formation and optimal filter assignment. These strategies are applied to multiple collaborative RCDs, considering their different computing capabilities and network conditions. Finally, a convex optimization problem is formulated to minimize the overall energy consumption by jointly optimizing the workload of each RCD in each layer and communication and computation parameters. |
| first_indexed | 2024-03-11T23:37:21Z |
| format | Article |
| id | doaj.art-d962c6217bb749a98d24e8fcde3bc2a4 |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-03-11T23:37:21Z |
| publishDate | 2023-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-d962c6217bb749a98d24e8fcde3bc2a42023-09-19T23:01:22ZengIEEEIEEE Access2169-35362023-01-0111993979941210.1109/ACCESS.2023.331438110247012A Relay-Assisted Communication Scheme for Collaborative On-Device CNN Execution Considering Hybrid ParallelismEmre Kilcioglu0https://orcid.org/0000-0002-7433-0411Ivan Stupia1Luc Vandendorpe2https://orcid.org/0000-0003-4958-8848ICTEAM/ELEN, Université catholique de Louvain, Ottignies-Louvain-la-Neuve, BelgiumICTEAM/ELEN, Université catholique de Louvain, Ottignies-Louvain-la-Neuve, BelgiumICTEAM/ELEN, Université catholique de Louvain, Ottignies-Louvain-la-Neuve, BelgiumDeep learning (DL) has gained increasing prominence in latency-critical artificial intelligence (AI) applications. Due to the intensive computational requirements of these applications, cloud-centric approaches have been attempted to address this issue, but they result in intolerable latency, network congestion, and privacy concerns. An alternative concept called edge intelligence, which combines AI and edge computing, has been proposed to perform DL execution at the edge in multiple resource-constrained devices (RCDs) collaboratively. This paper proposes a relay-assisted, distributed, and collaborative on-device convolutional neural network (CNN) execution scheme for latency-critical applications. The system employs hybrid parallelism, combining both data and model parallelism, to optimize collaborative CNN execution on RCDs. A relay-assisted communication technique is used to reduce the input data size per RCD and avoid excessive point-to-point communication between the data owner RCD and collaborative RCDs. The proposed approach reduces communication overhead using two strategies: layer block formation and optimal filter assignment. These strategies are applied to multiple collaborative RCDs, considering their different computing capabilities and network conditions. Finally, a convex optimization problem is formulated to minimize the overall energy consumption by jointly optimizing the workload of each RCD in each layer and communication and computation parameters.https://ieeexplore.ieee.org/document/10247012/Convex optimizationconvolutional neural networkdata parallelismDNN partitioningedge inferenceedge intelligence |
| spellingShingle | Emre Kilcioglu Ivan Stupia Luc Vandendorpe A Relay-Assisted Communication Scheme for Collaborative On-Device CNN Execution Considering Hybrid Parallelism IEEE Access Convex optimization convolutional neural network data parallelism DNN partitioning edge inference edge intelligence |
| title | A Relay-Assisted Communication Scheme for Collaborative On-Device CNN Execution Considering Hybrid Parallelism |
| title_full | A Relay-Assisted Communication Scheme for Collaborative On-Device CNN Execution Considering Hybrid Parallelism |
| title_fullStr | A Relay-Assisted Communication Scheme for Collaborative On-Device CNN Execution Considering Hybrid Parallelism |
| title_full_unstemmed | A Relay-Assisted Communication Scheme for Collaborative On-Device CNN Execution Considering Hybrid Parallelism |
| title_short | A Relay-Assisted Communication Scheme for Collaborative On-Device CNN Execution Considering Hybrid Parallelism |
| title_sort | relay assisted communication scheme for collaborative on device cnn execution considering hybrid parallelism |
| topic | Convex optimization convolutional neural network data parallelism DNN partitioning edge inference edge intelligence |
| url | https://ieeexplore.ieee.org/document/10247012/ |
| work_keys_str_mv | AT emrekilcioglu arelayassistedcommunicationschemeforcollaborativeondevicecnnexecutionconsideringhybridparallelism AT ivanstupia arelayassistedcommunicationschemeforcollaborativeondevicecnnexecutionconsideringhybridparallelism AT lucvandendorpe arelayassistedcommunicationschemeforcollaborativeondevicecnnexecutionconsideringhybridparallelism AT emrekilcioglu relayassistedcommunicationschemeforcollaborativeondevicecnnexecutionconsideringhybridparallelism AT ivanstupia relayassistedcommunicationschemeforcollaborativeondevicecnnexecutionconsideringhybridparallelism AT lucvandendorpe relayassistedcommunicationschemeforcollaborativeondevicecnnexecutionconsideringhybridparallelism |