saadmahboob / millimeterwavetxarraycomparison Goto Github PK

mmW_Tx_array_architectures

This is Matlab code and excel data sheet for the following paper

H. Yan, S. Ramesh, T. Gallagher, C. Ling and D. Cabric, “Performance, Power, and Area Design Trade-offs in Millimeter-Wave Transmitter Beamforming Architectures,” submitted to IEEE Circuits and System Mag., Mar. 2018. [online available] https://arxiv.org/abs/1807.07201

How to use it

1. The required power in Fig.3 of the paper is simulated by scripts [main_xx_required_power_theo.m] where xx={DA,SA,FH} for three architectures. This value should be verified with [main_xx_DAC_quantization.m] for target spectral efficiency performance, where xx={DA,SA,FH} for three architectures. The figure is plotted by script [plot_allcases_required_tx_power.m] and all data is hard coded.
2. The required quantization precision for DSP/DAC/PS in Fig. 4 and 5 of the paper are generated by scripts [main_xx_DAC_quantization.m] and [main_xx_PS_quantization.m] where xx={DA,SA,FH} for three architectures.
3. The system level power consumption and IC area estimation is generated by spread sheet in [excel_sheets \to Tx_Array_comparison_Jul2018]. When varying antenna array design parameters, the required power should be re-simulated by step 1. The quantization level should be re-simulated by step 2, but excel table uses theoretical value from eq.(11) of the paper and gives good approximations.
4. Fig. 9 to 11 are generated by scripts [plot_casex_power_breakdown.m], where x={1,2,3}. All data are hard coded from excel sheet in step3.
5. Fig. 12 is generated by script [plot_allcases_DSP_FOM.m]. All datas are hard coded from excel sheet in step3.
6. Fig. 13 is generated by script [plot_allcases_area_breakdown.m]. All datas are hard coded from excel sheet in step3.