Chao Xu, Jessie Hui Wang, Yipeng Zhou, Yuedong Xu, Changqing An, Jilong Wang, "Offloading Elastic Transfers to Opportunistic Vehicular Networks Based on Imperfect Trajectory Prediction," IEEE/ACM Transactions on Networking. (under revision.)

Due to the high cost of cellular networks, vehicle users would like to offload elastic traffic through vehicular networks as much as possible. This demand prompts researchers to consider how to make the vehicular network system achieve better performance for requests coming online, such as maximizing throughput. The traffic in vehicular networks is transferred through opportunistic contacts between vehicles and infrastructures. When making scheduling decisions, the scheduler must be aware of vehicles' future trajectories. Vehicles' future trajectories are usually predicted by trajectory prediction algorithms when users are not willing to report their future trips. Unfortunately, no trajectory prediction algorithm can be completely accurate, and these inaccurate prediction results will reduce the throughput. In this paper, we focus on overcoming this challenge. Specifically, we first measure the LSTM (Long Short Term Memory) prediction method that has been widely used recently and understand the accuracy of predicted contacts. Then, based on the enlightenment from the measurement, we design a system, i.e., i-Offload, to offload elastic traffic under an imperfect trajectory prediction algorithm. The experimental results show that our method still has good throughput and scheduling efficiency under an imperfect prediction algorithm. Compared with an existing scheduling method, our method outperforms it by 42% and 45% in terms of scheduling throughput and scheduling efficiency, respectively.

AI algorithms have been widely used in various problems, but how to deal with their errors needs more investigation and should be treated seriously. We took a careful analysis on the prediction errors of LSTM-based methods (which is the most popular) and propose a scheduling system which can adapt to imperfect prediction algorithms.

Chenghao Rong, Jessie Hui Wang, Juncai Liu, Jilong Wang, Fenghua Li, and Xiaolei Huang, "Scheduling Massive Camera Streams to Optimize Large-scale Live Video Analytics," IEEE/ACM Transactions on Networking. (minor revision.) paper

In smart cities, a city deploys an infrastructure and departments submit requests to access and analyze videos for their own purposes. The live analytics of massive streams is computation-intensive and the tasks might be latency-critical, which makes scheduling massive streams to optimize all tasks an essential and challenging work. We exploit an end-edge-cloud architecture and propose an adaptive system to schedule the massive camera streams and tasks, which considers all factors affecting the computation and networking resource consumption, e.g., sharing of model computation, video quality, model partition, and task placement. Particularly, the resource consumption of Faster R-CNN + ResNet101 under each partition scheme is profiled for the first time and we notice the partition must be used together with lossless compression techniques to be beneficial. Furthermore, sometimes tasks might be required to migrate because the scheduling decision made by the system changes to adapt to the changing resource supply and demand. In order to avoid the performance degradation during migration, we propose a nondestructive migration scheme and implement it in the system. Simulations demonstrate our system achieves a total utility close to the maximum and our analytics system performs better than state-of-the-art solutions.

The scheduler and the nondestructive migration scheme are welcomed by a well-known edge computing platform KubeEdge and we are working with them to contribute it to the community after the paper is accepted.

[TMC 2021] Yipeng Zhou, Jiawen Chen, Guoqiao Ye, Di Wu, Jessie Hui Wang, and Min Chen, "Collaboratively Replicating Encoded Content on RSUs to Enhance Video Services for Vehicles," IEEE Transactions on Mobile Computing, vol. 20, no. 3, pp. 877-892, March 2021. [paper]

With the development of smart cities, Internet services will be pervasively accessible for moving vehicles. It is envisioned that the video content demand of vehicles will explode in the near future. However, the strategy to efficiently distribute video content in large-scale vehicular networks is still absent due to challenges arising from the huge video population, heavy bandwidth consumption, heterogeneous user devices and vehicles’ mobility. In this work, we propose to collaboratively replicate video content on Roadside Units (RSUs) to enhance video distribution services based on the fact that the contact period between moving vehicles and a single RSU is not long enough to complete video downloading. In our design, a video file is split into multiple chunks. Each RSU replicates a small number of original chunks and chunks encoded by network coding. Replicating encoded chunks can reduce redundancy of chunks on different RSUs so that RSUs can complement each other better, whereas original chunks can be transrated to chunks with lower bitrates flexibly to fit in users’ devices. Therefore, we replicate both original and encoded chunks on RSUs to take advantages of both sides. Stochastic models are employed to analyze chunk download processes and a convex optimization problem is formulated to determine the optimal partition of space allocated to each kind of chunks. Furthermore, we extend our strategy to support video streaming services and empirically prove that the influence caused by limitations of network coding is moderate. In the end, we conduct extensive simulations which not only validate the accuracy of our models but also demonstrate that our strategy can effectively boost video distribution services.

Juncai Liu, Jessie Hui Wang, Chenghao Rong, Yuedong Xu, Tao Yu, and Jilong Wang, "FedPA: an Adaptively Partial Model Aggregation Strategy in Federated Learning," Computer Networks. Paper [@ScienceDirect]

Federated Learning has sparked increasing interest as a promising approach to utilize large amounts of data stored on network edge devices. In Federated Averaging, the server keeps waiting for client models to compute the global model in each round unless all client models are received or a pre-configured timer expires, therefore it suffers seriously from participant devices with weak computation and/or communication capability, which is a kind of straggler problem.

In this paper we design FedPA, a framework based on partial model aggregation strategy, in which the server waits for only an appropriate number of device models (referred to as aggregation number) in each round. Our experiment shows that the accuracy loss of the aggregated global model in a single round is not significant if the aggregation number is decided carefully. We propose a waiting strategy to determine the aggregation number for each round dynamically and the aggregation number is adaptive to achieve a tradeoff between single-round training time and the expected number of rounds to reach the target accuracy. Stale models are also included during aggregation when they arrive, and their positive value and negative effect are carefully evaluated and reflected in the aggregation strategy. Experiments show that FedPA outperforms the baseline strategy FedAvg and other three algorithms named FedAsync, FLANP and AD-SG. It can work well in all scenarios with different distributions of data samples (characterized by non-IID ratio) and computation/communication capability (characterized by level of heterogeneity) among devices. Experiments also show that FedPA is robust when a certain amount of noise is added into the input from clients for privacy concerns.

Shengchao Liu, Jianping Weng, Jessie Hui Wang, Changqing An, Yipeng Zhou, and Jilong Wang, "An Adaptive Online Scheme for Scheduling and Resource Enforcement in Storm," IEEE/ACM Transactions on Networking, vol. 27, no. 4, pp. 1373-1386. paper

As more and more applications need to analyze unbounded data streams in a real time manner, data stream processing platforms have drawn attention of many researchers, especially the scheduling problem. However, there are still many challenges unnoticed or unsolved. In this paper, we propose and implement an adaptive online scheme to solve three important challenges of scheduling. First, how to make scaling decision in a real time manner to handle fluctuant load without congestion? Second, how to minimize the number of affected workers during rescheduling while satisfying the resource demand of each instance? We also point out that stateful instances should not be placed on the same worker with stateless instances. Third, currently the application performance cannot be guaranteed because of resource contention even if the computation platform implements an optimal scheduling algorithm. In this paper, we realize resource isolation using Cgroup, and then the performance interference caused by resource contention is mitigated. We implement our scheduling scheme and plug it into Storm, and our experiments demonstrate in some respects our scheme achieves better performance than state-of-the-art solutions.

A few months after our paper was published, Storm announced a new version that supports CGroup Enforcement APIs “to limit the resource usage of workers to guarantee fairness and QoS”. It shows that our proposal on resource isolation using CGroup is reasonable and useful. Storm does not have any adaptive online schedulers till now. We are now working with a commercial cloud network to improve our scheduler and make it practically usable in a production environment.

Jianping Weng, Jessie Hui Wang, Jiahai Yang, and Yang Yang, "Root Cause Analysis of Anomalies of Multitier Services in Public Clouds," IEEE/ACM Transactions on Networking, vol. 26, no. 4, pp. 1646-1659, 2018. paper

Anomalies of multitier services of one tenant running in cloud platform can be caused by the tenant’s own components or performance interference from other tenants. If the performance of a multitier service degrades, we need to find out the root causes precisely to recover the service as soon as possible. In this paper, we argue that the cloud providers are in a better position than the tenants to solve this problem, and the solution should be non-intrusive to tenants’ services or applications. Based on these two considerations, we propose a solution for cloud providers to help tenants to localize root causes of any anomaly. With the help of our solution, cloud operators can find out root causes of any anomaly no matter the root causes are in the same tenant as the anomaly or from other tenants. Particularly, we elaborate a non-intrusive method to capture the dependency relationships of components, which improves the feasibility. During localization, we exploit measurement data of both application layer and underlay infrastructure, and our two-step localization algorithm also includes a random walk procedure to model anomaly propagation probability. These techniques improve the accuracy of our root causes localization. Our small-scale realworld experiments and large-scale simulation experiments show a 15%–71% improvement in mean average precision compared with the current methods in different scenarios.