Efficient GPGPU-Based Parallel Packet Classification

Hung, Che-Lun; Wang, Hsiao-Hsi; Guo, Shih-Wei; Lin, Yaw-Ling; Li, Kuan‐Ching

doi:10.1109/trustcom.2011.186

Cited by 14 publications

(10 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, these approaches take longer to filter a packet set captured off a giga-bit network interface than it took the set to arrive, making them infeasible for realtime traffic reduction. Therefore, we integrate GPU based packet filtering approach [13] into botnet detection tool [30] to enhance the performance of the tool. Figure 2 shows the proposed architecture.…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Parallel botnet detection system by using GPU

Hung

Wang

2014

2014 IEEE/ACIS 13th International Conference on Computer and Information Science (ICIS)

Self Cite

View full text Add to dashboard Cite

In recent years, botnet is one of the major threats to network security. Many approaches have been proposed to detect botnets by comparing bot features. Usually, these approaches adopt traffic reduction strategy as first step to reduce the flow to following strategies by filtering packets. With the rapid development of network hardware and software the network speed has reached to multi-gigabit. However, analyzing header and payload of every packet consumes huge amount of computational resources and is not suitable for many realistic situations. Although signaturebased solutions are accurate, it is not possible to detect bot variants in real-time. In this study, we proposed a GPU-based botnet detection approach. The experimental results show that the network traffic reduction stage on GPU can achieve about 8x times over CPU based botnet detection tool. The proposed algorithm can used to improve the performance of botnet detection tools efficiently.

show abstract

Section: Methodsmentioning

confidence: 99%

“…In recent years, General-Purpose Graphics Processing Unit (GPGPU) has been applied to solve computation intensive problem of various domains [9][10][11][12][13][14][15]. Especially, GPGPU programming is useful for the scientific computing domains that involve a high level of numeric computation.…”

Section: Introductionmentioning

confidence: 99%

Parallel botnet detection system by using GPU

Hung

Wang

2014

2014 IEEE/ACIS 13th International Conference on Computer and Information Science (ICIS)

Self Cite

View full text Add to dashboard Cite

show abstract

“…Chung et al proposed a GPGPU based parallel packet classification method to decrease the huge computing time to filter large number of packets [40]. Efforts in this research work aim to address the performance issues studied in the related endeavours.…”

Section: Related Workmentioning

confidence: 99%

CUDA based Rabin-Karp Pattern Matching for Deep Packet Inspection on a Multicore GPU

Sharma

Singh²

2015

IJCNIS

View full text Add to dashboard Cite

This paper presents a study of the improvement in efficiency of the Rabin-Karp patternmatching algorithm based Deep Packet Inspection. NVIDIA GPU is programmed with the NVIDIA's general purpose parallel computing architecture, CUDA, that leverages the parallel compute engine in NVIDIA GPUs to solve many complex computational problems in a more efficient way than on a CPU. The proposed CUDA based implementation on a multicore GPU outperforms the Intel quadcore processor and runs upto 14 times faster by executing the algorithm in parallel to search for the pattern from the text. The speedup may not sound exorbitant but nonetheless is significant, keeping in view that the experiments have been conducted on real data and not synthetic data and, optimal performance even with the huge increase in traffic was the main expectation, not just an improvement in speed with few test cases.

show abstract

“…However, [17] is based on the assumption that the set of unique criteria is small for large rule set. [18] implements a decomposition-based approach and a decision-tree-based approach on GPU. The design is optimized by exploring a variety of memory architectures of CUDA.…”

Section: Related Workmentioning

confidence: 99%

High-performance packet classification on GPU

Zhou

Singapura

Prasanna

2014

2014 IEEE High Performance Extreme Computing Conference (HPEC)

View full text Add to dashboard Cite

Multi-field packet classification is a network kernel function where packets are classified and routed based on a predefined rule set. Recently, there has been a new trend in exploring Graphics Processing Unit (GPU) for network applications. These applications typically do not perform floating point operations and it is challenging to obtain speedup. This paper presents a high-performance packet classifier on GPU. We investigate GPU's characteristics in parallelism and memory accessing, and implement our packet classifier using Compute Unified Device Architecture (CUDA). The basic operations of our design are binary range-tree search and bitwise AND operation. We optimize our design by storing the range-trees using compact arrays without explicit pointers in shared memory. We evaluate the performance with respect to throughput and processing latency. Experimental results show that our approach scales well across a range of rule set sizes from 512 to 4096. When the size of rule set is 512, our design can achieve the throughput of 85 million packets per second and the average processing latency of 4.9 µs per packet. Compared with the implementation on the state-of-the-art multi-core platform, our design demonstrates 1.9x improvement with respect to throughput.

show abstract

Efficient GPGPU-Based Parallel Packet Classification

Cited by 14 publications

References 17 publications

Parallel botnet detection system by using GPU

Parallel botnet detection system by using GPU

CUDA based Rabin-Karp Pattern Matching for Deep Packet Inspection on a Multicore GPU

High-performance packet classification on GPU

Contact Info

Product

Resources

About