基于 EASI 自适应盲分离的 I/Q 失衡补偿

2021-10-13 18:02:33 阅读：168 来源： 互联网

标签：EASI factorial b22 b21 失衡补偿 b12 b11 yi1

IQ失衡问题。
1首先用matlab生成输入信号计算b矩阵，达到收敛。
2再导出输入信号作为fpga的输入，在vivado上进行仿真，计算的b矩阵应与matlab计算一致。
3导出真实信号作为输入，在matlab中计算，可能存在直流偏移需要滤除。
4真实信号在vivado上仿真测试。
5上板测试
IQ失衡补偿架构
下面为组合逻辑实现代码

`timescale 1ns / 1ps
//
// Company: 
// Engineer: 
// 
// Create Date: 2021/09/16 20:57:52
// Design Name: 
// Module Name: iq
// Project Name: 
// Target Devices: 
// Tool Versions: 
// Description: 
// 
// Dependencies: 
// 
// Revision:
// Revision 0.01 - File Created
// Additional Comments:
// 
//


module iq(
	input adsp_clk,//10M
	input refclk,//100M
	input refclk2, //10M
	output signed [31:0]I_out,
	output signed [31:0]Q_out,
	input signed [13:0]sin_coe,
	input signed [13:0]cos_coe,
	output signed [31:0]b11_out,
	output signed [31:0]b12_out,
	output signed [31:0]b21_out,
	output signed [31:0]b22_out,
	output signed [31:0]yi1_out,
	output signed [31:0]yq1_out,
	output signed [13:0]yi,
	output signed [13:0]yq
	
    );
	reg signed [31:0] b11=32'sb00010000000000000000000000000000,b12=32'sb00000000000000000000000000000000,b21=32'sb00000000000000000000000000000000,b22=32'sb00010000000000000000000000000000;
	reg signed [31:0] a =32'sb00000000000000000000010000000000;
	reg signed [31:0]I_chanel_reg,Q_chanel_reg;
	reg signed [31:0] yi1,yq1;
	reg signed [31:0] one =32'sb00010000000000000000000000000000;
	reg signed [31:0] yi12,yq12;
	reg signed [31:0] test1=32'sb00000000000000000000000000000000,test2=32'sb00000000000000000000000000000000,test3=32'sb00000000000000000000000000000000,test4=32'sb00000000000000000000000000000000;
	//b11={b11[15],b11[0]<<14};
	//b12={b12[15],b12[0]<<14};
	//b21={b21[15],b21[0]<<14};
	//b22={b22[15],b22[0]<<14};
	reg [17:0]addr=0;
	reg signed[31:0] b11_t,b12_t,b21_t,b22_t;
	wire signed [31:0]sin_wr,cos_wr;
	reg signed [31:0] yi1_2,yi1_4,yq1_2,yq1_4,yi1_31,yq1_31,yiq_1;
	reg signed [31:0]onea=32'sb00010000000000000000010000000000;
	reg signed[31:0] b11_p,b12_p,b21_p,b22_p;
	
	/*
always@(posedge refclk)
begin
if (!rst)
	addr <= 0;                                  //复位地址清零
else 
	addr <= addr + 1;                           //地址自增
end
	*/



	function signed [31:0] factorial;
		input signed [31:0] operand1,operand2; 
		reg signed [63:0] mult1;
		reg signed [35:0] mult2;
		
		begin
				mult1=operand1*operand2;
				mult2=mult1>>28;
				factorial={mult2[35],mult2[30:0]};
		end
		
	endfunction 
 /*
  always@(posedge refclk2	)
  begin
	yi1<=factorial(b11,{{3{cos_coe[19]}},cos_coe[18:0]})+factorial(b12,{{3{sin_coe[19]}},sin_coe[18:0]});
	yq1<=factorial(b21,{{3{cos_coe[19]}},cos_coe[18:0]})+factorial(b22,{{3{sin_coe[19]}},sin_coe[18:0]});


	
  end
 
 */
 
 
  always @(posedge adsp_clk)
begin
#10
	yi1=factorial(b11,{{3{sin_coe[13]}},sin_coe[12:0],16'b0000000000})+factorial(b12,{{3{cos_coe[13]}},cos_coe[12:0],16'b0000000000});
	yq1=factorial(b21,{{3{sin_coe[13]}},sin_coe[12:0],16'b0000000000})+factorial(b22,{{3{cos_coe[13]}},cos_coe[12:0],16'b0000000000});
	
	yi1_2=factorial(yi1,yi1);
	yq1_2=factorial(yq1,yq1);
	yq1_4=factorial(yq1_2,yq1_2);
	yi1_4=factorial(yi1_2,yi1_2);
	yiq_1=factorial(yi1,yq1);
	yi1_31=factorial(yi1_2,yiq_1);
	yq1_31=factorial(yq1_2,yiq_1);
	
	b11_t=factorial(a,yi1_4);
	b12_t=factorial(a,yi1_31);
	b21_t=factorial(a,yq1_4);
	b22_t=factorial(a,yq1_31);
	
	
	b11_p=factorial(onea,b11)-factorial(b11_t,b11)-factorial(b12_t,b21);
	b12_p=factorial(onea,b12)-factorial(b11_t,b12)-factorial(b12_t,b22);
	b21_p=factorial(onea,b21)-factorial(b21_t,b21)-factorial(b22_t,b11);
	b22_p=factorial(onea,b22)-factorial(b21_t,b22)-factorial(b22_t,b12);
	b11=b11_p;
	b12=b12_p;
	b21=b21_p;
	b22=b22_p;
	
/*
	yi1=factorial(b11,I_chanel_reg)+factorial(b12,Q_chanel_reg);
	yq1=factorial(b21,I_chanel_reg)+factorial(b22,Q_chanel_reg);

	test1=factorial(yi1,yi1);
	test1=factorial(test1,yi1);
	test2=factorial(test1,yq1);
	test1=factorial(test1,yi1);
	test3=factorial(yq1,yq1);
	test3=factorial(test3,yq1);
	test4=factorial(test3,yi1);
	test3=factorial(test3,yq1);
	
	b11_t=factorial(a,test1);//yi^4
	b12_t=factorial(a,test2);//yi^3*yq
	b21_t=factorial(a,test3);//yq^4
	b22_t=factorial(a,test4);//yq^3*yi


	b11_p=factorial((one+a),b11)-factorial(b11_t,b11)-factorial(b12_t,b21);
	b12_p=factorial((one+a),b12)-factorial(b11_t,b12)-factorial(b12_t,b22);
	b21_p=factorial((one+a),b21)-factorial(b21_t,b21)-factorial(b22_t,b11);
	b22_p=factorial((one+a),b22)-factorial(b21_t,b22)-factorial(b22_t,b12);
	b11=b11_p;
	b12=b12_p;
	b21=b21_p;
	b22=b22_p;
*/
end 





assign I_out=sin_coe;
assign Q_out=cos_coe;
//assign yi1_out={{3{sin_coe[19]}},sin_coe[18:0]};
//assign yq1_out={{3{cos_coe[19]}},cos_coe[18:0]};
assign yi1_out=yi1;
assign yq1_out=yq1;
assign b11_out=b11;
assign b12_out=b12;
assign b21_out=b21;
assign b22_out=b22;
assign yi={yi1[31],yi1[28:16]};
assign yq={yq1[31],yq1[28:16]};

endmodule

标签：EASI,factorial,b22,b21,失衡,补偿,b12,b11,yi1
来源： https://blog.csdn.net/qq_43424748/article/details/120749177

本站声明： 1. iCode9 技术分享网（下文简称本站）提供的所有内容，仅供技术学习、探讨和分享；
2. 关于本站的所有留言、评论、转载及引用，纯属内容发起人的个人观点，与本站观点和立场无关；
3. 关于本站的所有言论和文字，纯属内容发起人的个人观点，与本站观点和立场无关；
4. 本站文章均是网友提供，不完全保证技术分享内容的完整性、准确性、时效性、风险性和版权归属；如您发现该文章侵犯了您的权益，可联系我们第一时间进行删除；
5. 本站为非盈利性的个人网站，所有内容不会用来进行牟利，也不会利用任何形式的广告来间接获益，纯粹是为了广大技术爱好者提供技术内容和技术思想的分享性交流网站。

ICode9

基于 EASI 自适应盲分离的 I/Q 失衡补偿