Abstract
Lattice reduction algorithms are powerful tools for breaking knapsack based cryptosystems. In 2005, Ludwig introduced a new lattice reduction algorithm called Sampling Reduction. In this paper, we explain Sampling Reduction would not be effective for the attack against knapsack based cryptosystems by using our experimental results. Furthermore, we propose a Modified Sampling Reduction, and also show that it is more effective than the original Sampling Reduction by using the experimental results.