Table 2: Comparison of the tabu search and greedy algorithms in terms of quality improvement 𝐼 t a b u and execution time over-cost ( 𝑂 𝐶 t a b u )

Benchmark 𝑁 𝑔 𝐼 t a b u 𝑇 g r e e d y (s) 𝑇 t a b u (s) 𝑂 𝐶 t a b u

14 2 . 9 % 4.5 14.6 219%
IIR 18 6 . 5 % 35.1 83.2 137%
36 6 . 6 % 78.3 177.1 126%
8 6 5 . 7 % 26.1 62.8 141%

FFT 12 6 2 . 4 % 57.3 163.4 185%
20 0 . 6 % 57.4 128.8 124%
13 1 2 . 5 % 16.8 37.1 120%

NLMS 25 1 6 . 3 % 76.5 152.4 99%
49 9 . 6 5 % 286.5 579.6 102%