English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Zendy Plus

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Zendy Tools

Zendy Open

Currently, deep neural networks (DNNs) have achieved a great success invarious applications. Traditional deployment for DNNs in the cloud may incur aprohibitively serious delay in transferring input data from the end devices tothe cloud. To address this problem, the hybrid computing environments,consisting of the cloud, edge and end devices, are adopted to offload DNNlayers by combining the larger layers (more amount of data) in the cloud andthe smaller layers (less amount of data) at the edge and end devices. A keyissue in hybrid computing environments is how to minimize the system cost whileaccomplishing the offloaded layers with their deadline constraints. In thispaper, a self-adaptive discrete particle swarm optimization (PSO) algorithmusing the genetic algorithm (GA) operators was proposed to reduce the systemcost caused by data transmission and layer execution. This approach considersthe characteristics of DNNs partitioning and layers offloading over the cloud,edge and end devices. The mutation operator and crossover operator of GA wereadopted to avert the premature convergence of PSO, which distinctly reduces thesystem cost through enhanced population diversity of PSO. The proposedoffloading strategy is compared with benchmark solutions, and the results showthat our strategy can effectively reduce the cost of offloading for DNN-basedapplications over the cloud, edge and end devices relative to the benchmarks.

Cost-Driven Offloading for DNN-based Applications over Cloud, Edge and End Devices