RECORD
Edge AI as a Service with Coordinated Deep Neural Networks
Networking and Internet Architecture · Submitted 1 Jan 2024
abstract compiled · 2026-09-23 · no language-model review
The authors' main claim Experiments show that CoDE enhances the inference throughput and, achieves higher precision compared to a state-of-the-art existing method.
The compiled record
The paper's own sentences, arranged. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. Numbers in the tables are the authors', checked by code. No sentence in this record was written by a model.
Open the compiled record
The abstract, sentence by sentence
As artificial intelligence (AI) applications continue to expand in next-generation networks, there is a growing need for deep neural network (DNN) models. Although DNN models deployed at the edge are promising for providing AI as a service with low latency, their cooperation is yet to be explored. In this paper, we consider that DNN service providers share their computing resources as well as their models' parameters and allow other DNNs to offload their computations without mirroring. We propose a novel algorithm called coordinated DNNs on edge (\textbf{CoDE}) that facilitates coordination among DNN services by establishing new inference paths. CoDE aims to find the optimal path, which is the path with the highest possible reward, by creating multi-task DNNs from individual models. The reward reflects the inference throughput and model accuracy. With CoDE, DNN models can make new paths for inference by using their own or other models' parameters. We then evaluate the performance of CoDE through numerical experiments. The results demonstrate a $40\%$ increase in the inference throughput while degrading the average accuracy by only $2.3\%$. Experiments show that CoDE enhances the inference throughput and, achieves higher precision compared to a state-of-the-art existing method.
background contribution method result interpretation main claim
| # | sentence | role | the authors' own work | states the main conclusion | strength |
|---|---|---|---|---|---|
| s0 | As artificial intelligence (AI) applications continue to expand in next-generation networks, there is a growing need for deep neural network (DNN) models. | background 1.00 | 0.04 | 0.02 | asserted 0.97 |
| s1 | Although DNN models deployed at the edge are promising for providing AI as a service with low latency, their cooperation is yet to be explored. | gap 0.99 | 0.06 | 0.02 | asserted 0.85 |
| s2 | In this paper, we consider that DNN service providers share their computing resources as well as their models' parameters and allow other DNNs to offload their computations without mirroring. | method 0.47 | 0.24 | 0.04 | asserted 0.78 |
| s3 | We propose a novel algorithm called coordinated DNNs on edge (\textbf{CoDE}) that facilitates coordination among DNN services by establishing new inference paths. | contribution 0.80 | 0.90 | 0.09 | asserted 1.00 |
| s4 | CoDE aims to find the optimal path, which is the path with the highest possible reward, by creating multi-task DNNs from individual models. | method 0.74 | 0.66 | 0.09 | asserted 0.99 |
| s5 | The reward reflects the inference throughput and model accuracy. | method 0.95 | 0.31 | 0.03 | asserted 0.99 |
| s6 | With CoDE, DNN models can make new paths for inference by using their own or other models' parameters. | method 0.89 | 0.42 | 0.06 | asserted 0.62 |
| s7 | We then evaluate the performance of CoDE through numerical experiments. | method 0.88 | 0.92 | 0.03 | asserted 0.83 |
| s8 | The results demonstrate a $40\%$ increase in the inference throughput while degrading the average accuracy by only $2.3\%$. | result 1.00 | 0.96 | 0.54 | asserted 1.00 |
| s9 | Experiments show that CoDE enhances the inference throughput and, achieves higher precision compared to a state-of-the-art existing method. | result 0.96 | 0.93 | 0.45 | asserted 1.00 |
What was examined, and what was not
- arXiv listing abstract v2
- full text
- figures and tables
- references and citing sentences
- relations to other papers
- P abstract
- S sections
- T body sentences
- C citations
- R cross-paper relations
The paper
Original abstract
As artificial intelligence (AI) applications continue to expand in next-generation networks, there is a growing need for deep neural network (DNN) models. Although DNN models deployed at the edge are promising for providing AI as a service with low latency, their cooperation is yet to be explored. In this paper, we consider that DNN service providers share their computing resources as well as their models' parameters and allow other DNNs to offload their computations without mirroring. We propose a novel algorithm called coordinated DNNs on edge (\textbf{CoDE}) that facilitates coordination among DNN services by establishing new inference paths. CoDE aims to find the optimal path, which is the path with the highest possible reward, by creating multi-task DNNs from individual models. The reward reflects the inference throughput and model accuracy. With CoDE, DNN models can make new paths for inference by using their own or other models' parameters. We then evaluate the performance of CoDE through numerical experiments. The results demonstrate a $40\%$ increase in the inference throughput while degrading the average accuracy by only $2.3\%$. Experiments show that CoDE enhances the inference throughput and, achieves higher precision compared to a state-of-the-art existing method.
Human review
How the literature uses it
What this paper does with the arXiv papers it cites
What it cites
Uses the method or tool of 1 paper
-
In this experiment, we use MobileNet [ref] as the host DNN service.
MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications · Experiment 3: AlexNet - MobileNet · 0.94
Cites as background 2 papers
-
BERT models [ref] use adapter modules [ref] to adapt pre-trained knowledge to new tasks and enable task-specificity without modifying the base models' parameters.
BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding · Related Work · 0.89
-
Multi-task learning (MTL) is a technique in which a shared model is used to learn multiple tasks simultaneously [ref] .
Multi-Task Learning with Deep Neural Networks: A Survey · Related Work · 1.00
Discussion
Know something about this paper? Leave a comment, or highlight any passage above to comment on it.