Search

This paper proposed an Integrated Random Early Detection (IRED) method that aims to resolve the problems of the queue-based AQM and loadbased AQM and gain the benefits of both using indicators from both types. The arrival factor (e.g., arrival rate, queue and capacity) and the departure factors are used to estimate the congestion through two integrated indicators. The utilized indicators are mathematically calculated and integrated to gain unified and coherent congestion indicators. Besides, IRED is built based on a new dropping calculation approach that fits the utilized congestion indicators while maintaining the intended buffer management criteria, avoiding global synchronization and enhancing the performance. The results showed that IRED, compared to RED, BLUE, ERED, FLRED, EnRED and DcRED, decreased packet delay and loss under various network status. Specifically, the results showed that in heavy and moderate traffic, the proposed IRED method outperformed the state-of-the-art methods in loss and delay by 18% and 10.6%, respectively. [ABSTRACT FROM AUTHOR]

Recently, large datasets have become increasingly necessary for most deep learning tasks, however, large datasets may bring some problems, such as disk storage and huge computational expense. Dataset distillation is an emerging field that aims to synthesize a small dataset from the original dataset, then a random model trained on the distillation dataset can achieve comparable performances to the same architecture model trained on the original dataset. Matching Training Trajectories (MTT) achieves a leading performance in this field, but it needs to pre-train 200 expert models before the formal distillation process, which is called buffer process. In this paper, we propose a new method to reduce the consumed time of buffer process. Concretely, we use Mixture of Experts (MoE) to train several expert models parallelly in buffer process. The experiments show our method can achieve a speedup of up to approximately 4 ∼ 8 × in buffer process with getting comparable distillation performances. • Apply Mixture of Experts to accelerate dataset distillation. • Our method is widely effective for trajectories-matching distillation methods. • Experiments show that our method gets comparable performances. [ABSTRACT FROM AUTHOR]