TY - GEN
T1 - Poster abstract
T2 - 2017 IEEE Conference on Computer Communications Workshops, INFOCOM WKSHPS 2017
AU - Ben Basat, Ran
AU - Einziger, Gil
AU - Friedman, Roy
AU - Kassner, Yaron
N1 - Publisher Copyright:
© 2017 IEEE.
PY - 2017/11/20
Y1 - 2017/11/20
N2 - Bloom filters and their variants support membership or multiplicity queries with a low probabilistic error. For many networking applications, recent data is more significant than older data, motivating the need for sliding window solutions. In this work, we introduce Sliding Window Approximate Membership Protocol (SWAMP), a simple algorithm for membership and multiplicity queries over sliding windows. SWAMP is the first approximate set membership sliding window algorithm that is memory succinct, i.e., up to a factor of (1 + 0(1)) from the information theoretic lower bound, for constant error probabilities. It also operates in constant time and supports multiplicity queries with no additional overheads. Finally, we evaluate the memory consumption of SWAMP on a wide range of parameters and show a 25-40% reduction compared to the state of the art sliding Bloom filters (that cannot count). In summary, SWAMP improves the memory consumption of its competitors and can also count.
AB - Bloom filters and their variants support membership or multiplicity queries with a low probabilistic error. For many networking applications, recent data is more significant than older data, motivating the need for sliding window solutions. In this work, we introduce Sliding Window Approximate Membership Protocol (SWAMP), a simple algorithm for membership and multiplicity queries over sliding windows. SWAMP is the first approximate set membership sliding window algorithm that is memory succinct, i.e., up to a factor of (1 + 0(1)) from the information theoretic lower bound, for constant error probabilities. It also operates in constant time and supports multiplicity queries with no additional overheads. Finally, we evaluate the memory consumption of SWAMP on a wide range of parameters and show a 25-40% reduction compared to the state of the art sliding Bloom filters (that cannot count). In summary, SWAMP improves the memory consumption of its competitors and can also count.
UR - http://www.scopus.com/inward/record.url?scp=85020697577&partnerID=8YFLogxK
U2 - 10.1109/INFCOMW.2017.8116536
DO - 10.1109/INFCOMW.2017.8116536
M3 - Conference contribution
AN - SCOPUS:85020697577
T3 - 2017 IEEE Conference on Computer Communications Workshops, INFOCOM WKSHPS 2017
SP - 1012
EP - 1013
BT - 2017 IEEE Conference on Computer Communications Workshops, INFOCOM WKSHPS 2017
PB - Institute of Electrical and Electronics Engineers
Y2 - 1 May 2017 through 4 May 2017
ER -