TY  - CONF
UR  - http://doi.acm.org/10.1145/2859889.2859896
SP  - 63
ID  - eprints3576
EP  - 68
AV  - none
N1  - SCOPUS ID: 2-s2.0-84969287510
KW  - layered queuing networks
KW  -  ordinary differential equations
KW  -  pepa
CY  - New York, NY, USA
TI  - DiffLQN: Differential Equation Analysis of Layered Queuing Networks
A1  - Waizmann, Tabea
A1  - Tribastone, Mirco
N2  - Layered queuing networks are a popular technique in software performance engineering. In this paper we present DiffLQN, a tool for the analysis of networks using ordinary differential equations. It estimates average performance indices such as throughput, utilization, and response time of software and hardware devices. The complexity of computing the solution is independent of the concurrency levels in the model (i.e., thread multiplicities and processing units) and the estimates are theoretically guaranteed to be asymptotically correct for large enough concurrency levels. DiffLQN is designed having in mind compatibility with other tools that support state-of-the-art methods based on mean value analysis.
M2  - New York, USA
Y1  - 2016///
PB  - ACM
T2  - Companion Publication for ACM/SPEC on International Conference on Performance Engineering
SN  - 978-1-4503-4147-9
T3  - ICPE '16 Companion
ER  -