Advertisement

Radix Tree for Binary Sequences on GPU

  • Krzysztof KaczmarskiEmail author
  • Albert Wolant
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 10777)

Abstract

In this paper, we present radix tree index structure (R-Trie) able to perform lookup over a set of keys of arbitrary length optimized for GPU processors. We present a fully parallel SIMD organized creation and search strategies. The R-Trie supports configurable bit stride for each level and nodes statistics for optimization purposes. We evaluate the performance using two search strategies and Longest Prefix Match (LPM) problem for computer networks. Unlike dedicated LPM algorithms we do not incorporate knowledge about the data or the network masks statistics into the tree construction or algorithm behavior. Our solution may be used in general purpose indexing structures where a batch search of massive number of keys is needed. (The research was funded by National Science Center, decision DEC-2012/07/D/ST6/02483.)

Keywords

Radix tree Parallel search Longest prefix match GPU 

References

  1. 1.
    Kaczmarski, K.: B+-tree optimized for GPGPU. In: Meersman, R., et al. (eds.) OTM 2012. LNCS, vol. 7566, pp. 843–854. Springer, Heidelberg (2012).  https://doi.org/10.1007/978-3-642-33615-7_27 CrossRefGoogle Scholar
  2. 2.
    Li, Y., Zhang, D., Liu, A.X., Zheng, J.: GAMT: a fast and scalable IP lookup engine for GPU-based software routers. In: ANCS, pp. 1–12. IEEE Computer Society (2013)Google Scholar
  3. 3.
    Karras, T.: Maximizing parallelism in the construction of BVHs, octrees, and k-d trees. In: Dachsbacher, C., Munkberg, J., Pantaleoni, J. (eds.) High Performance Graphics, pp. 33–37. Eurographics Association, Aire-la-Ville (2012)Google Scholar
  4. 4.
    Lim, H., Lim, K., Lee, N., Park, K.-H.: On adding bloom filters to longest prefix matching algorithms. IEEE Trans. Comput. 63(2), 411–423 (2014)MathSciNetCrossRefzbMATHGoogle Scholar
  5. 5.
    Lee, J., Lim, H.: Binary search on trie levels with a bloom filter for longest prefix match. In: HPSR, pp. 38–43. IEEE (2014)Google Scholar
  6. 6.
    Mu, S., Zhang, X., Zhang, N., Lu, J., Deng, Y.S., Zhang, S.: IP routing processing with graphic processors. In: Micheli, G.D., Al-Hashimi, B.M., Müller, W., Macii, E. (eds.) DATE, pp. 93–98. IEEE (2010)Google Scholar
  7. 7.
    Zheng, K., Liu, B.: V6Gene: a scalable IPv6 prefix generator for route lookup algorithm benchmark. In: AINA, vol. 1, pp. 147–152. IEEE Computer Society (2006)Google Scholar
  8. 8.
    Sahni, S., Kim, K.S.: Efficient construction of multibit tries for IP lookup. IEEE/ACM Trans. Netw. 11(4), 650–662 (2003)CrossRefGoogle Scholar

Copyright information

© Springer International Publishing AG, part of Springer Nature 2018

Authors and Affiliations

  1. 1.Warsaw University of TechnologyWarszawaPoland

Personalised recommendations