Practical Random Linear Network Coding on GPUs

Abstract

Abstract. Recently, random linear network coding has been widely applied in peer-to-peer network applications. Instead of sharing the raw data with each other, peers in the network produce and send encoded data to each other. As a result, the communication protocols have been greatly simplified, and the applications experience higher end-to-end throughput and better robustness to network churns. Since it is difficult to verify the integrity of the encoded data, such systems can suffer from the famous pollution attack, in which a malicious node can send bad encoded blocks that consist of bogus data. Consequently, the bogus data will be propagated into the whole network at an exponential rate. Homomorphic hash functions (HHFs) have been designed to defend systems from such pollution attacks, but with a new challenge: HHFs require that network coding must be performed in GF(q), where q is a very large prime number. This greatly increases the computational cost of network coding, in addition to the already computational expensive HHFs. This paper exploits the potential of the huge computing power of Graphic Processing Units (GPUs) to reduce the computational cost of network coding and homomorphic hashing. With our network coding and HHF implementation on GPU, we observed significant computational speedup in comparison with the best CPU implementation. This implementation can lead to a practical solution for defending against the pollution attacks in distributed systems.

Analysis the Performance of Network Coding for Ad Hoc Networks in Realistic Simulation Scenarios

“…In [7] the impact of several IEEE 802.11based MAC protocols on the performance of network coding for broadcast communications is analyzed. We highlight that to the best of our knowledge only a few works, e.g., [7], [6], exist on network coding for ad hoc networks, referring to practical scenarios or using realistic network simulators. Thus, one of the main goals of our work is to assess the performance of network coding for message broadcasting by using the network simulator ns-2 as an evaluation tool [1].…”

Benchmarking the Memory Hierarchy of Modern GPUs

“…GPUs have become popular parallel computing accelerators; but their further performance enhancement is limited by the sophisticated memory system [1][2][3][4][5][6]. In order to reduce the default memory access consumption, developers usually utilize some specially designed memory spaces empirically [3][4][5].…”

Tsunami: massively parallel homomorphic hashing on many‐core GPUs

“…In CUDA, the GPU becomes a dedicated coprocessor to the host CPU, which works in the principle of single-program multiple data whereby multiple threads based on the same code can run simultaneously. GPUs have recently been proposed to accelerate network coding [13][14][15]. Our previous work also applied GPU computing in accelerating homomorphic hashing [15][16][17].In this paper, we propose the Tsunami system that uses GPUs for homomorphic hashing.…”

“…GPUs have recently been proposed to accelerate network coding [13][14][15]. Our previous work also applied GPU computing in accelerating homomorphic hashing [15][16][17].In this paper, we propose the Tsunami system that uses GPUs for homomorphic hashing. The homomorphic hash function needs to multiply a large number of exponentiations, so the critical design concern of Tsunami is to optimize the modular exponentiation operation for high precision large integers.…”

“…GPUs have recently been proposed to accelerate network coding [13][14][15]. Our previous work also applied GPU computing in accelerating homomorphic hashing [15][16][17].…”