Lei Song

Dr. Lei Song

orcid

Research Focus & Metrics

Academic Career Length

7Years (2020-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
15 October 2020 in Journal of Control Science and Engineering
Verified
Journal Article
Open access
PDF file
CLChuanhong Li
Chuanhong Li
XZXuewen Zeng
Xuewen Zeng
LSLei Song
Lei Song
YJYan Jiang
Yan Jiang
Packet classification algorithms have been the focus of research for the last few years, due to the vital role they play in various services based on packet forwarding. However, as the number of rules in the rule set increases, not only the preprocessing time but also the memory consumption is increasing greatly. In this paper, we first model and analyze the above issue in depth. Then, a fast, smart packet classification algorithm based on decomposition is proposed. By boundary-based rule traversal and smart rule set partitioning, both the preprocessing time and memory consumption are reduced dramatically. Experimental results show that the preprocessing time of our method achieves 8.8-time improvement at maximum compared with the PCIU and achieves about 31.5-time improvement on average compared with CutSplit for large rule sets. Meanwhile, the memory overhead is reduced by 40% at maximum and 27.5% on average compared with the PCIU.

16 September 2022 in IEEE Transactions on Network and Service Management
Verified
Journal Article
Closed access
XTXinxin Tang
Xinxin Tang
XZXuewen Zeng
Xuewen Zeng
LSLei Song
Lei Song
The advances in software-defined networking (SDN) have recently promoted the development of Programmable Data Planes (PDP). PDP is famous in various network environments that need to support new protocols due to its flexibility and protocol independence. Protocol-oblivious forwarding (POF) is an implementation of PDP that uses to describe the match fields and mixes the protocol parsing into the pipeline, so there is no need for a pre-defined packet header parser. Current cache schemes for the data plane, such as Microflow Cache/Megaflow Cache, are primarily implemented based on the pre-defined packet header parser. In this paper, we first analyzed the two difficulties of generating cache tables on POF PDP. Then we analyzed the advantages and disadvantages of the single flow table cache system and multiple flow tables cache system from the perspective of time overhead. Finally, we propose a pipeline-oriented cache system to accelerate POF PDP. Aiming at the pipeline with multiple branches and multiple matching paths, we design algorithms to obtain all cache tables. Then our design leveraged the incremental calculation feature of the hash function to reduce the time overhead of cache lookup. In addition, the sorting mechanism and sleep mechanism for multiple cache tables were proposed according to the locality of traffic. Our design is implemented in a real POF PDP environment. The experimental results show that our pipeline-oriented cache scheme can effectively accelerate the programmable data plane, reducing packet forwarding latency and improving the throughput.

1 October 2024 in Frontiers of Information Technology & Electronic Engineering
Verified
Journal Article
Closed access
PDF file
DRDengyu Ran
Dengyu Ran
XCXiao Chen
Xiao Chen
LSLei Song
Lei Song
Dynamic bandwidth allocation (DBA) is a fundamental challenge in the realm of networking. The rapid, accurate, and fair allocation of bandwidth is crucial

8 March 2024 in Electronics
Verified
Journal Article
Open access
PDF file
DRDengyu Ran
Dengyu Ran
XCXiao Chen
Xiao Chen
LSLei Song
Lei Song
Programmable networks comprise heterogeneous network devices based on both hardware and software. Hardware devices provide superior bandwidth and low latency but encounter challenges in managing large table entries. Conversely, software devices offer abundant flow tables but have a limited forwarding capacity. To overcome this limitation, some commercial switches offer implementations that combine both hardware and software devices. In this context, this paper presents the Composite Pipeline (ComPipe), an algorithm for high-performance and high-precision flow placement and measurement. ComPipe utilizes a multi-level hashing algorithm for the real-time identification of heavy hitters, incorporates a unique flow eviction strategy, and is implemented on commercial programmable hardware. For non-heavy flows, ComPipe employs sketch structures to accomplish a high-performance flow synopsis within limited memory constraints. This design allows to replace flow rules entirely in the data plane, ensuring the timely detection and offloading of heavy-hitter flows, and offering a unified interface to the controller. The ComPipe prototype has been implemented in both testbed and simulation environments. The results indicate that ComPipe is an effective solution for dynamic flow placement in programmable networks, distinguished by its low cost, high performance, and high accuracy.

19 March 2021 in Future Internet
Verified
Journal Article
Open access
PDF file
CLChuanhong Li
Chuanhong Li
LSLei Song
Lei Song
XZXuewen Zeng
Xuewen Zeng
The continuous increase in network traffic has sharply increased the demand for high-performance packet processing systems. For a high-performance packet processing system based on multi-core processors, the packet scheduling algorithm is critical because of the significant role it plays in load distribution, which is related to system throughput, attracting intensive research attention. However, it is not an easy task since the canonical flow-level packet scheduling algorithm is vulnerable to traffic locality, while the packet-level packet scheduling algorithm fails to maintain cache affinity. In this paper, we propose an adaptive throughput-first packet scheduling algorithm for DPDK-based packet processing systems. Combined with the feature of DPDK burst-oriented packet receiving and transmitting, we propose using Subflow as the scheduling unit and the adjustment unit making the proposed algorithm not only maintain the advantages of flow-level packet scheduling algorithms when the adjustment does not happen but also avoid packet loss as much as possible when the target core may be overloaded Experimental results show that the proposed method outperforms Round-Robin, HRW (High Random Weight), and CRC32 on system throughput and packet loss rate.

31 August 2022 in Applied Sciences
Verified
Journal Article
Open access
PDF file
XTXinxin Tang
Xinxin Tang
XZXuewen Zeng
Xuewen Zeng
LSLei Song
Lei Song
Modern software data planes use spin-polling and batch processing mechanisms to significantly improve maximum throughput and forwarding latency. The user-level IO queue-based spin polling mechanism has a higher response speed than the traditional interrupt mechanism. The batch mechanism enables the software data plane to achieve higher throughput by amortizing the IO overhead over multiple packets. However, the software data plane under the spin-polling mechanism keeps running at full speed regardless of the input traffic rate, resulting in significant performance waste. At the same time, we find that the batch processing mechanism does not cope well with different input traffic, mainly reflected in the forwarding latency. The purpose of this paper is to optimize the forwarding latency by leveraging the wasted performance. We propose a forwarding latency optimization scheme for the software data plane based on the spin polling mechanism in this paper. First, we calculate the CPU utilization of the software data plane according to the number of cycles the CPU spends on the valuable task. Then, our scheme controls the Tx queues and dynamically adjusts the output batch size based on the CPU utilization to optimize the forwarding latency of the software data plane. Compared with the original software data plane, the evaluation result shows that the forwarding latency can be reduced by 3.56% to 45% (in a single queue evaluation) and 4.35% to 55.54% (in a multiple queue evaluation).