PERFORMANCE ANALYSISComputational complexityHere we considered N iterations. So the dot product requires N complex multiplication and N complex addition. Multiplying any scalar by vector requires N complex multiplications. Table #1 below shows the computational complexity of conventional LMS, RLS, Variable Leaky LMS (VLLMS) and our proposed MVSS-LLMS algorithm. And where we observe that the step size modified variable leakage LMS is less complex than the variable leakage LMS. -MVSSLLMS N2+5N+3 6N 1Bit Error Rate (BER)In the 2x2 MIMO-OFDM system, we simulated a modified variable leakage LMS (MVSS-LLMS) for different SNRs and iterations. And after that the result will be compared with standard LMS and RLS algorithms. For the simulation, we must first define the parameters µ, α and w as well as the initial values ​​of µ, ρ and w. In this proposed algorithm, we set these parameters µ, ρ and w to zero. To study the performance of the algorithm in the presence of noise, we use the BPSK signal with Gaussian noise. Here, the performance of LMS and RLS algorithms are plotted in Figure 6 and Figure 7 respectively. SYNTHESIS AND SIMULATION RESULTS The schematic diagram of the modified LMS kernel is shown in Figure 8. Here, the modified LMS kernel is implemented in VHDL and it is simulated in Xilinx ISE 13.1 Design Combination. The pin description is given in Table #2 which represents the input/output signal lists for this design. Figure 6: LMS Performance Graph Figure 7: RLS Performance Graph Figure 8: VLLMS Performance Graph Figure 9 : Performance graph of... ... middle of paper ...... IEEE Transactions On Broadcasting, Vol. 48, no. 3, September 2002.[15] S. Haykin, “Adaptive Filter Theory,” 4th ed. Upper Saddle River, New Jersey: Prentice Hall, 2002.[16] BaoguoYang, Khaled Ben Letaief, Roger S. Cheng, “Channel estimation for OFDM transmission in multipath fading channel based on parametric channel modeling,” IEEE Transactions On Communications, Vol. 49, No. 3, March 2001[17] K.Mayyas and TyseerAboulnasr “Leaky LMS algorithm: Analysis for Gaussion data” IEEE Transaction on Signal Processing Vol.45, No. 4 April 1997[18] Dale A. Lawrance, William A. Sathares and Wei Ren “Parameter Drift Instability in Adaptive Filter” on Decision and Control, CH2917-3/90/0000-323 $1.00 © 1990 IEEE.[19] William Sethares, Dale Richard Johnson, Robert Bitmead. Parameter Drift in LMS Adaptive Filter » IEEE Transactions on Acoustics, Speech and Signal Processing, Vol A August 34 1986